# Changelog > What we shipped, when we shipped it. No fluff, just features. URL: https://brandcrux.ai/changelog ## 28 Jun 2026 - **Clearer cookie choices, analytics off until you opt in**. A new consent banner lets you accept or reject non-essential analytics cookies, and analytics now stays completely off until you choose to accept. You can change your choice anytime from the "Cookie preferences" link in the footer or the button on the Cookie Policy page. We also refreshed our Privacy and Cookie policies to spell out exactly what we collect, the choices you have, and your rights. ## 10 Jun 2026 - **Shared Search Console, Analytics, GEO, and Web Search reports now match the full dashboard**. The shareable versions of these reports went from a single summary to the same depth you see inside the app: Search Console adds decliners, topic clusters, cannibalisation, device and country breakdowns, and intent gaps; Analytics adds its channel and engagement panels; the GEO report adds a full category breakdown, the top recurring issues with recommended fixes, per-page detail, and a competitor comparison; and Web Search adds your ranking keywords and per-engine presence. The BrandCrux overview also gained Search Console and Analytics tiles and a cross-channel competitor leaderboard. - **Invite teammates to specific workspaces**. When you invite a member or viewer, you can now limit them to the workspaces you choose, and they will only see and work in those. Leave "All workspaces" selected for full access. You can change any member's workspace access at any time from the members list. Owners and admins always see every workspace. - **The Board Report is now the Brand Authority Report**. Same deep, board-room-ready assessment of how your brand shows up across every channel, with a name that says what it measures. Nothing about the report or how you order it changed. - **Shared report leaderboards show your brand at its real rank**. Across AI authority topics, web search, socials, citations, and backlinks, the "You" row now sits at the brand's actual position by score instead of being pinned to the top, so the competitive standing reads at a glance. It stays visible even when the brand ranks far down the list. - **One link to share your whole BrandCrux picture**. There is now a single overview link per brand that shows the headline BrandCrux Score plus the top number for every channel (AI Visibility, Web Search, Socials, Citations, Backlinks, GEO, and more), with a tap-through to each full report. Help tips were also added throughout every shared report, so anyone reading it can hover any number to see what it means. - **Shared reports now match what you see inside the app**. Every shareable report (Web Search, Socials, Directory Citations, Competitor Intelligence, Recommendations, and Authority Topics) now renders the same rich panels as your in-app view: KPI strips, full competitor leaderboards, heatmaps, scorecards, and feeds. Anyone you send a link to sees the complete picture, not a trimmed summary. - **Citation URLs are now validated before they land on your reports**. Every AI Visibility scan now probes each cited URL: dead links get dropped, and pages that resolve but never actually mention any of your tracked brands are dropped too. Stops the AI engines from filling your Citations panel with plausible-looking links to pages that do not exist or that have nothing to do with the brand the AI claimed to be citing. The dropped URLs are still recorded in the raw response for audit, so nothing is lost. ## 06 Jun 2026 - **ChatGPT citations are no longer undercounted**. ChatGPT started returning overflow citations in a second array a few days ago and we were silently dropping them. Now both arrays feed the citation set, so AI Visibility scans capture every source ChatGPT shows behind an answer. No action needed on existing scans; new runs pick up the additional rows automatically. - **Reposts no longer inflate the Social leaderboard**. The X (Twitter) feed now distinguishes original posts from reposts and the engagement leaderboard hides reposts by default, with a one-click toggle if you want to count them. Stops viral reshared content from being credited to whoever reposted it. - **The Board Report has a brand new guided setup**. Ordering a report is now a focused step-by-step flow. The moment you enter your brand and domain, it pre-fills your closest competitors, your authority topics, and a set of buyer-intent prompts grouped under each topic, all from your website. Everything is editable in place, recommended prompts are pre-selected, and a "Suggest more with AI" button adds fresh ideas. You can tap any step in the progress bar to jump straight to it. - **Clearer, all-in pricing**. The Board Report is now $250 and includes 5 competitors, 5 authority topics, 5 AI engines, and 25 prompts. Anything beyond that is shown live as you build the order: $25 per extra competitor or topic, $5 per extra prompt, and $0.50 per engine for each prompt it runs. The total updates as you go, with no surprises at checkout. - **Every order now creates a free account with your brand saved**. When your report is delivered you get a free BrandCrux account with 10,000 credits, and your brand profile is stored (with your permission) so future reports stay up to date without starting over. Returning customers can pick an existing brand to re-run, review and update the setup, and order again in a couple of clicks. - **Pick up where you left off**. If you close the tab mid-setup, the report flow now resumes your in-progress order instead of starting blank. ## 02 Jun 2026 - **Saved social handles are now re-verified, and the list is half the height**. The handle verifier used to only run on new AI suggestions; older saved handles never got checked again, so a deleted profile sat in your account silently breaking the leaderboard. The Socials page now has a Verify accounts button, auto-runs it once on page load for any handle that has never been verified, and flags any profile the network reports as missing with a red dot, a "Likely dead" badge, and a one-click Remove. The handle panel itself collapses to a 2-column grid so 8 handles fit in 4 rows instead of 8. - **Run discovery now tells you what happened**. Clicking Run discovery on the New Scan banner used to fire silently. Now it shows an inline status the moment it dispatches, then 30 seconds later either clears (results landed) or surfaces a clear note that no ranked keywords were found for your domain yet, with a one-click path to seed positive keywords manually. The mapping action gets the same treatment. No more clicking a button that looks like it did nothing. - **"Channels" is now "Signals"**. Half the things on the old Channels page were not channels in the marketing sense (citations, Search Console and GA4 insights, backlinks). The umbrella is now called Signals and the per-tile noun is "surface" where appropriate. The old Channels URL still works so existing bookmarks do not break, and Web Search and Socials keep being called channels where that meaning fits. - **Mission Control is no longer a separate page**. The same checklist now lives at the top of the BrandCrux dashboard, auto-expanding while there is anything to do and collapsing once everything is on track. The standalone URL still works (it redirects to the dashboard) so existing links keep flowing. ## 01 Jun 2026 - **Pillar tiles now explain the foundation-vs-outcome story**. Hover the info icon on AI Visibility, Web Search, Web Content, or GEO Audit and you get the diagnostic: which pillars are foundations (GEO and Web Content), which are outcomes (AI and SERP), and the three common reasons a high foundation score sits next to a low outcome score. The new GEO Audit feature page also walks through the framing in long-form for visitors. - **Tracked-keyword set now respects supplier vs competitor**. The brand filter on keyword classification used to drop every brand on the workspace's competitor list, which meant a hearing-aid retailer could never track queries containing the manufacturer brands it actually sells. Each competitor is now classified by its real commercial relationship (competitor, supplier, partner, reseller, parent, unrelated). Only true competitors gate the brand filter; suppliers and partners stay in the competitor list but their brand-prefixed queries flow into your authority topics where they belong. - **Volume and Search Console hint on Keywords**. A sortable Volume column now sits next to Impressions on the Keywords page, and an amber banner explains why Impressions sit at zero when Search Console is not connected. - **Tracked keywords now include high-volume competitor terms**. Keywords your competitors rank for that map to one of your authority topics now land in the Pending Review tab with their search volume and topic already assigned, so you can decide which to track without going hunting. ## 31 May 2026 - **Onboarding checklist is now embedded in the Topic Authority dashboard**. The "what to do next" list that used to live on its own Mission Control page now sits at the top of the BrandCrux dashboard. It is expanded while any onboarding step is still open and collapses automatically once everything is complete, so a workspace in steady-state no longer has to scroll past it. Clicking any unfinished step still routes you to the Mission Control drawer that hosts the per-step form. - **Returning users now land on the Topic Authority dashboard after login**. Once you have completed onboarding for at least one workspace, signing in takes you straight to the BrandCrux composite view so you can read the day's score without an extra click. Users still finishing their first workspace continue to land on Mission Control where the wizard lives. - **Mission Control graph cards no longer squeeze on narrower windows**. The pipeline tiles were forcing four or five cards into whatever horizontal space was available, which on smaller windows compressed each card so much that brand names broke a character at a time and titles ellipsed. The grid now reflows automatically: fewer columns at narrower widths so each card keeps a usable width, the body wraps on word boundaries, and titles stay readable. - **Refreshed authority topics across active workspaces**. Low-coverage and duplicated topics were removed and new high-volume gaps were added based on the keyword universe across each workspace and its competitors. Each touched workspace now has a cleaner taxonomy that better matches what its audience is actually searching for. ## 29 May 2026 - **Keywords are now classified by intent, not by lexical overlap**. The previous topic mapper stamped a seed topic onto every keyword the keyword-ideas data returned in its neighbourhood, which is how queries like "imf loan to pakistan", "india finance ministers", and "education loan" ended up tagged on a vehicle-finance topic. Each suggested keyword is now run through a vertical-aware classifier that uses your workspace's full topic taxonomy (with descriptions) plus your business context (name, domain, categories) to decide intent, and explicitly skips off-topic news, politics, and unrelated brand queries instead of forcing them onto the nearest-word topic. The same classifier now also covers your competitors' ranked keywords so the union view across self and competitors is consistent. Existing keyword-to-topic mappings have been refreshed. - **AI-suggested social handles are now verified before they reach the UI**. When you ask BrandCrux to suggest the official handles for a brand or competitor on Instagram, Facebook, YouTube, TikTok, X, or LinkedIn, each suggestion is now checked against the network in parallel and any handle that points to a deleted or non-existent profile is dropped before it lands on the page. Live profiles and handles the network is slow to confirm both stay, so a transient hiccup never silently deletes a real account. ## 28 May 2026 - **Top-up credit packs are now cheaper across the board**. The 1,500-credit pack drops from $25 to $20, the 4,000-credit pack from $50 to $45, the 9,000-credit pack from $100 to $90, and the 30,000-credit pack from $250 to $200. Subscriptions still carry the best per-credit rate for sustained usage; top-ups are now substantially closer to subscription parity so buying overage to finish a project no longer feels like a punishment. See the pricing page for the full ladder. - **Backlinks and Traffic moved to its own URL**. The route name now matches the page title. Old bookmarks redirect automatically. - **Backlinks and Traffic now covers every tracked competitor end-to-end**. Workspaces that previously saw a single brand row in the Backlinks and Traffic boards now pull a full snapshot for each competitor automatically. Estimated monthly organic visits, traffic change, bounce rate, pages per visit, average visit duration, and top-country mix all populate per competitor on the next page load. No action needed for existing workspaces. ## 27 May 2026 - **BrandCrux composite now favours visibility over infrastructure**. Topic Authority and the BrandCrux dashboard hero used to give site-health signals (content depth, GEO audit) almost as much weight as AI mentions or SERP placements, so a brand with a 90/100 GEO score and zero AI presence could still score middling. The composite is now weighted 30% AI, 25% SERP, 15% citations, 12% socials, with the remaining 18% across content depth, crawl coverage, and GEO infrastructure. A perfect GEO audit no longer compensates for being invisible in AI answers or the top-3 of search results. - **SERP column reads as a 0-100 score**. The per-topic SERP cell used to show a raw top-10 placement count, which collided with every other pillar's 0-100 scale. It now reads as a normalised score (10 top-10 placements caps at 100). Colour bands match the rest of the table: green at or above 70, amber 40 to 70, red below. - **Empty competitor cells now show 0 instead of a dash**. When a competitor returned nothing on SERP, citations, socials, website, or AI for a topic, the table rendered "no value" cells that looked like missing data. They now show 0 explicitly, since 0 signal IS the signal. GEO stays blank when there's no site-wide audit on file (genuinely-missing data, not a 0-score). ## 26 May 2026 - **Discovered Competitors now classifies industry match per domain**. The Discovered list used to surface every brand that appeared alongside yours, even when the candidate was a hospital, a regulator, a news outlet, or an unrelated industry. Each newly surfaced domain is now classified for industry fit against your workspace categories. Out-of-scope domains are hidden by default with a one-click "Show them" toggle, pending classifications are labelled "Pending review", and you can override any verdict with "Mark as competitor" so a false negative never stays hidden again. ## 25 May 2026 - **Every public page now ships a complete social-share card and SEO meta set**. New shared composable emits title, description, canonical, robots, social preview (with image, dimensions, alt, site name), summary card, and per-page-type structured data (WebPage, Article, FAQPage, Product, BreadcrumbList). Sitewide Organisation and WebSite structured data lands on every route. Sitemap now includes cookies, help, and status pages. - **Your own social activity now sits next to competitors on every Socials view**. Every Socials scan pulls posts from the brand's own LinkedIn, X, Instagram, Facebook, YouTube, and TikTok accounts alongside every tracked competitor. The engagement leaderboard and Authority Topics scorecard pin "You" to row 1, the per-topic expansion shows your engagement against each competitor side-by-side, and a new "What to post next" panel produces 5-10 specific post ideas where competitors are outpacing you. New brand-self handles card at the top of the page handles add, remove, and AI-suggest with one click. Onboarding auto-suggests handles when you create a business, so the first scan has full self vs competitor data without any manual setup. - **Mission Control flowchart cards stop chopping words mid-character**. Long-form tier names and summaries used to wrap one letter per line on narrow viewports. Tightened the wrapping rule to break on whitespace only, and titles now span up to 2 lines instead of ellipsis-truncating. - **AI Visibility insight row is balanced again**. The four cards (Personas, Top citations, Competitors, Insights) sit in a 2x2 layout instead of leaving Insights orphaned in row 2. - **Web Content Pages table no longer pushes columns off-screen**. Long URLs in the Page column used to truncate the Words, Score, and Last seen columns into invisibility. The grid layout now reserves min-widths so every column stays readable. - **Help info on every Web Content tab**. Each tab (Overview, Pages, Topics, Gaps) now has a collapsible help section explaining what's shown, how the data is built, and what to do with it. - **Web Search deep-dive opens as a closable tab**. Clicking a keyword on the Web Search Queries page now opens a tab next to Overview, Queries, and Pages, so you can keep multiple keyword deep dives open and switch between them without losing the listing. Close any tab to dismiss it. - **Socials Off-topic section opens by default**. The "what else competitors are posting about" view used to start collapsed. Now it expands automatically with the top 10 rows in view. ## 22 May 2026 - **Public pricing page**. A new pricing page lists every plan side-by-side with a monthly and annual toggle, top-up packs, a feature-matrix table, and pricing FAQs. Annual savings call out automatically when you flip the toggle. - **Annual billing back for new sign-ups**. The per-charge limit was raised, so the monthly and annual toggle in the in-app billing panel works for new subscribers too. Annual saves up to 15% across Starter, Growth, and Scale. - **Settings, run the onboarding wizard**. A new card in account Settings opens the guided onboarding flow on demand, so dormant users (or anyone who didn't finish the wizard on signup) can re-enter at any time without going through Mission Control or signup again. - **Clearer GDPR and CCPA story**. The Privacy Policy now spells out California Consumer Privacy Act (CCPA / CPRA) rights alongside the existing GDPR coverage, and the in-app "Delete my account" flow explicitly references the right-to-erasure laws under which it operates. - **Admin nav now exposes Leaderboards and Social drafts**. The content-marketing engine's surfaces (industry-leaderboard jobs, topic, persona, and prompt libraries, social-post draft queue) now have entries in the admin sidebar instead of being reachable only by direct URL. - **Cleaned up legacy comparison pages**. The five per-competitor head-to-head landing pages were retired during the rebrand. The main compare hub stays: same matrix, same decision tree, all five competitors in one place. - **Every dashboard tile is now a 0-100 score that matches its channel page**. SERP, Citations, and Web Content each render a vs-leader score (your count as a share of the leading competitor's, capped at 100) as the headline number on both the BrandCrux tile and the corresponding channel page. The supporting counts sit underneath. AI Visibility and GEO Audit continue to show their native 0-100 scores. Socials renders blank until brand-side posting tracking comes online. Click a tile, land on the page, see the same score. - **Sharpened the Growth-tier story**. The marketing page now leads with what's live today: AI Assistant, Content Engine briefs, CMS publishing across WordPress, Shopify, Webflow, Wix, and Squarespace, REST API and MCP server, and on-demand Authority Cards. Tighter copy, no maybes. ## 21 May 2026 - **Guided onboarding wizard, every workspace lands fully wired**. Adding a new business now walks through a five-step flow: brand basics, competitors (AI-suggested and manual, name and website required), authority topics, tracked prompts per topic, confirm. Each step auto-generates suggestions on first arrival and lets you regenerate if you change earlier inputs. On confirm we kick off keyword discovery, backlinks and traffic refresh, social handle suggestions, and the GEO audit, and optionally fire your first AI Visibility scan. Dashboards start populating with real signal within minutes of finishing instead of after a manual sequence of channel clicks. - **Authority Topics page now loads reliably for older workspaces**. A regression earlier today caused the Authority Topics list to show as empty for workspaces that had topics defined, while the Topic Map below kept showing data. Both panels now render against the same source so the page is consistent. - **BrandCrux dashboard polish**. The KPI tile strip is now a balanced 4+3 layout instead of leaving a lonely tile alone on row 2. Tiles also share a common baseline across titles, scores, and footer descriptions, so the row scans cleanly even when one card is still waiting for data. The Share button now lives inside the Topic Authority hero card, top-right, instead of floating above it. - **Mission Control, removed the noisy status strip**. The strip above the flowchart polled the pipeline state every few seconds but rarely produced anything new. The flowchart and per-step drawers already surface every next action with a real call to action, so the strip was just visual noise above them. - **Keyword Discovery, one click, every keyword you and your competitors rank for**. A new Keyword Discovery page pulls every keyword your brand ranks for, plus the same for every active competitor, with search volume, difficulty, AI Overview flag, and ranking URL. Discovery seeds the topic clusterer, the prompt-to-topic backfill, and the competitor-gap reports automatically, so brands that onboard without Search Console finally have something for the clusterer to work with. Adding a new competitor now also fires a discovery scan in the background, so authority topics light up on its first appearance. 15 credits per domain scanned. - **Auto-discover competitor social handles on add**. Adding a new competitor used to leave the socials surface blank until you opened the suggestions modal. Now we look for the handles upfront, pulling them from the crawl footer when one is available, and from a confident model lookup when it is not. High-confidence handles auto-attach so the Socials page lights up on first scan instead of after a second round-trip. - **Suggest socials is more resilient**. When the primary suggestion model is rate-limited or unavailable, Suggest with AI now retries through a second model instead of failing back to a generic error. Same cost ceiling, far fewer "couldn't generate suggestions" messages. - **Dashboards always surface scan data, no more "No signal yet" after a successful scan**. Fresh accounts without keyword coverage used to land in a silent gap where BrandCrux and Authority Topics showed empty cards even after a completed AI scan, because the per-topic aggregation needed a topic-to-keyword-to-prompt chain that didn't exist yet. We now auto-fire keyword discovery on business creation, on every newly-added authority topic, and again as a safety net when a scan completes without topic mapping. The topic-snapshot writer also writes a fallback distribution across your authority topics so the dashboards never sit empty after a successful scan. - **Mapping readiness banner on Mission Control and the scan launcher**. Before scans fire, we now check that authority topics, keywords, prompts, and topic-mapping all exist. If anything is missing, a banner surfaces the exact next step (Add topics, Run discovery, Add prompts, Map prompts) with a one-click action. No more spending credits on scans that can't aggregate. - **Email when a scan finishes**. Every scan type now sends one completion email: AI Visibility, Web Search, Citations, Socials, Search Console, GA4, and Competitor Crawl. The email has the report name, two or three headline numbers from the run, and a direct link to the report. Toggle it off in Settings under Notifications. The topbar "scans running" pill continues to work whether the email is on or off. - **Authority Topics Scorecard is now full-width on the Socials page**. The per-topic scorecard was previously sharing the row with the engagement leaderboard, which truncated topic names and clipped the per-channel breakdown. It now spans the full width of the page so every topic's posts, engagement, and top competitor are visible without squeezing. The Off-Topic panel also got a Show all and Show top 10 control so long lists no longer drown out the rest of the page. - **Every channel is now shareable**. Public shareable reports rolled out to every channel: Web Search, Directory Citations, Socials, Authority Topics, BrandCrux Summary, GEO Audit, and Search Console Insights. Each page has a Share button that generates a stable public link, re-clicking returns the same URL rather than fanning out duplicates. Each report is read-only with no edit affordances, sized for stakeholder skim rather than deep ops. Starter and above. - **Search Console and GA4 scan triggers are now gated behind their connectors**. The Launch a scan page used to let you queue a Search Console or GA4 deep scan even when no connection existed; the run would silently fail downstream. Those channel cards now show a dashed "Not connected" treatment with a direct "Connect Search Console" / "Connect GA4" link so you set it up first. - **Competitor "Extract from Crawl" socials actually finds handles now**. A key-rename in the crawl pipeline left every competitor crawl writing an empty page body, so Extract-from-Crawl returned zero handles every time even when the footer clearly listed them. The crawler now reads the right field; re-crawling a competitor with social-link footers populates handles as expected. Suggest-with-AI also surfaces real error detail instead of a generic "failed" message when the model errors out. - **"Scans running" indicator in the topbar**. Whenever a scan is in flight for the active workspace (AI, Web Search, Citations, Search Console / GA4, Crawl, GEO Audit) a small pulsing pill appears in the topbar showing the count. Tap it to jump straight to the Scans page. When all scans complete you also get a toast nudging you to view the output. - **Add a business actually adds a business again**. The "Add business" item in the workspace switcher used to redirect to a stub page that bounced back to Mission Control for the current workspace. It now opens a real new-business form (name, website, category, products and services, auto-fill from website) and drops you into Mission Control for the freshly created workspace. - **Content brief drafts: honest progress instead of a premature "ready" toast**. Tapping Generate Now on a brief used to flash "is ready for review" within seconds even though the writer was still composing. The success banner now says the draft is being generated, offers a direct "View draft" link to monitor it, and the draft page itself shows a live "Generating your draft" loader with auto-refresh. The editor populates the moment the writer finishes. - **Persona scans now route through the unified scan launcher**. The Personas page previously opened its own dedicated "Run persona scan" modal. It now links straight into the main Launch a scan page with the AI Visibility section already in view, so persona-scoped scans share the same channel picker, cost estimate, and approval flow as every other scan. - **Mission Control cards render at full width again**. A regression yesterday squeezed each tier card's title and summary into a narrow left column when the corner page-link arrow was added. Cards now fill their grid cell so titles no longer truncate. - **Adversarial watch cost is now clearly disclosed**. The Competitors page tooltip for Adversarial watch previously quoted "about 1 credit per day". Updated to the accurate "about 50 credits per day for up to 50 pages crawled" so the cost of toggling it on per competitor is upfront. The feature is opt-in, off by default, and toggled per competitor. - **GEO Audit listing repaired**. The GEO Audit page was rendering an empty list for both your own audits and competitor audits, even when completed audits existed. Fixed the underlying response shape so every audit you've ever run shows up again. Also patched the deep link used by the unified Scans history view so clicking a GEO Audit row opens the right page. - **Mission Control cards now jump to where you configure them**. Every tier card on Mission Control, in both the graph view and the new list view, is now a link to the matching configuration page (Authority Topics, Competitors, Connectors, Personas, Prompts, AI Pulse, Web Search, Socials, etc.). Title click in the list, corner arrow on the graph card. - **Mission Control reads bigger**. Bumped the whole page up 15% so the tier names, status pills, and progress numbers are easier to scan at desk-distance. - **Anomaly and competitor-move cards on your dashboard are now actionable**. Each row in the activity feed on the BrandCrux dashboard now navigates somewhere: anomalies open in Trend Intelligence, competitor moves open in Competitor Watch, and published content rows open the brief. ## 20 May 2026 - **Topic Setup Wizard**. A new wizard for every authority topic. Click Setup wizard on any topic card and BrandCrux proposes 8 buyer-intent prompts plus 20 SERP keywords for that topic, surfaces any Search Console queries that look related so you can pull them in with one click, and lets you edit or add your own before applying. Adding a new authority topic auto-opens the wizard. Apply writes the prompts and keywords atomically; the next pipeline run picks them up. - **Topic leaderboards now show every tracked competitor**. Previously a competitor only appeared on a topic leaderboard if a scan had returned data for them. They now appear on every topic, with blank chips in the channel cells where no data exists yet and a compact "Run pipeline" link on rows that have nothing at all. Clicking the link deep-links straight into the Run Pipeline modal on Competitors with that competitor pre-selected, so backfilling is one screen and one confirm. - **GEO score is now per-topic, not site-wide**. The GEO chip on every topic card was previously the same site-level score copied across every topic. It now averages the actual GEO Audit scores of pages mapped to that topic (relevance-weighted), so a brand with great schema on their pricing pages but missing schema on their guides shows the difference. Topics with no page-level GEO data show blank rather than a misleading number. - **Run Pipeline modal, centering and checkbox fix**. The Run Pipeline modal on the Competitors page now centers in the viewport regardless of scroll position. Toggling a stage checkbox no longer removes the row from the list. Total estimate updates only based on what's checked. - **Web Search Queries now shows positions**. The Queries view was rendering "Not ranked" for every keyword regardless of actual data. Fixed the response-shape mismatch so real positions and competitor-rank columns populate again. - **In-progress pages tucked out of production**. Automations, Plays, and Attribution were appearing in the sidebar on the live site while still under active development. They now stay visible on localhost and hide from the production sidebar until they ship. - **BrandCrux is now the Brand Authority Tracker across channels**. Sharper category claim, same product. The homepage hero, every footer signature, page titles, social previews, and the public llms.txt now describe BrandCrux as the Brand Authority Tracker across channels: AI, Web, SERP, Socials, citations, and your own analytics. The seven-channel breakdown stays on every deep page, the Topic Authority composite stays the headline metric, the closed-loop measurement stays the differentiator. - **1,000 free credits on signup, fatter referral rewards**. Every new account now starts with 1,000 free scan credits (up from 300), enough to run a full scan across all seven channels for one business and see the composite light up before you decide on a paid plan. The Refer and Earn rewards are also bigger: 500 credits each (up from 100) when an invitee signs up, and 1,000 credits each (up from 500) when they upgrade to a paid plan. Existing balances stay intact; new grants apply to all signups from today onwards. ## 19 May 2026 - **Homepage rewrite, straight to the question that matters**. The landing page hero now asks the actual question and runs a free scan inline. Drop your domain, hit enter, get the truth. The Without and With comparison moved up the page so the value lands before the deeper sections, every channel card now leads with the outcome you get (not the mechanism behind it), and every step of the Loop carries a concrete example of what that step produces. A slim sticky call to action now follows you down the page so the free scan is always one tap away. - **FAQ now opens to substantive answers**. The first two FAQs on the homepage and the AI Visibility page expand by default. Visitors see real answers without having to click first. - **AI Visibility pricing comparison matches the live plans**. The desktop pricing table on the AI Visibility page previously showed the pre-rebrand credit pools and a stale 100-credit Free tier. It now reflects the real ladder: Starter 10,000 credits, Growth 50,000, Scale 100,000, Free 300 one-time credits, plus team-member counts and an accurate REST API and MCP split. - **Recent Scans hides the zero-score noise**. The recent-scans strip on the AI Visibility page no longer shows scans whose overall score never came back above zero. Indexed, scored businesses only. - **Source citations on the Uncomfortable Math**. Every headline stat on the AI Visibility page now carries a one-line source attribution underneath. - **Bigger top-up packs to match the new credit appetite**. Top-up packs rebuilt at clean $25 / $50 / $100 / $250 price points for 1,500 / 4,000 / 9,000 / 30,000 credits respectively. Top-up credits still never expire, and the per-credit cost is calibrated so a sustained-usage customer always gets a better rate by subscribing to the matching plan. - **New pricing, in USD across the board**. Plans rebalanced and now billed in USD on every checkout. Starter moves to $99 a month and gains REST API access plus the AI Assistant (previously gated to Growth) and 10,000 credits a month. Growth moves to $299 with 50,000 credits, MCP server, content engine, and CMS publishing. Scale moves to $499 with 100,000 credits, white-label reports, priority support, and a dedicated CSM. Annual billing offers a 15% discount across every paid tier. Scale is now visible again in the plan picker. Top-up credit packs unchanged. - **Citation gap list no longer flags a competitor's own properties as "gaps"**. The gap list on Directory Citations and on every topic page now filters out citations on the competitor's own website or their own LinkedIn, Facebook, or X profile. Only third-party publisher and directory citations remain. - **Web Search competitor leaderboard skips non-competitor platforms**. The leaderboard on Web Search no longer surfaces social platforms, reference sites, big news outlets, consulting firms, code hosts, or app stores as "competitors". The Domains Observed KPI still counts every root domain seen in top-10. - **Engagement and geography backfills**. The bounce rate, pages-per-visit, average duration, and top-country fields on Backlinks and Traffic now populate for every tracked competitor after the next traffic refresh. Existing rows can be backfilled in place without re-pulling the rest of the snapshot. - **Topic Map now reflects competitive reality**. Major competitors get their full property rollups crawled. Crawl page budgets scale per-competitor. Hard-walled sites that used to silently return zero pages now route via proxy after an SSL fix so the bot-walled retries actually work. - **Keyword detail page**. Clicking any keyword on Web Search now opens a per-keyword detail view with position-over-time sparkline, latest top-10 SERP, per-competitor board (best position each domain has held), and People-Also-Ask. - **Pages tab shows every ranking URL**. Web Search Pages previously only returned featured-snippet rows, so the page looked empty. It now shows every URL that turned up in SERP for your tracked keywords: organic, People-Also-Ask, and featured snippets together. - **Backlinks and Traffic upgrade**. Competitor Intelligence is now titled Backlinks and Traffic to match the Channels card. Every table gets click-to-sort column headers. Two new widgets: DoFollow link quality with a strength bar, and Engagement and geography (bounce rate, pages per visit, average visit duration, top country) for every tracked domain. A "How to read this" panel explains each widget. - **Competitor social handles, three ways**. Open any row on Competitors and tap "Socials" to manage handles. Add manually, ask AI to suggest the brand's known accounts, or extract them from the footer of an already-crawled site. Every completed competitor crawl auto-extracts LinkedIn, X, Instagram, Facebook, YouTube, and TikTok URLs going forward. - **New scan: GEO Audit channel**. The new-scan picker gains a GEO Audit channel. Toggle Self and Competitors, set the max-pages slider, and the run dispatches an audit per target. GEO audits show as their own rows on the Scans page with status, score, and target domain. - **Socials pre-flight**. Selecting the Socials channel now warns when no competitor has any social handle configured, with a one-click path to Competitors to fix it. - **Personas: Suggest with AI fixed**. The Personas page's Suggest button previously returned a generic error for some brands because the output sometimes contained bare numeric ranges that strict parsing rejected. Now repairs malformed output automatically and generates 4 to 6 buyer personas reliably. - **Bug sweep**. Channels page tiles for Citations, Search Console Insights, and GA4 Insights showed empty values for businesses that had data; now populate. Authority Topics cards get visible borders. Directory Citations queries cap explicitly to avoid silent truncation. ## 18 May 2026 - **Topic Authority now scores each topic distinctly**. Every priority topic gets its own composite score based on the actual content depth and relevance of the pages covering it, instead of echoing one site-wide audit number. Multi-topic pages count toward each topic they cover. - **New: Content Authority**. Filterable per-page score table covering your site and every tracked competitor. Filter by topic, by source (you vs them), sort by authority, technical, freshness, or relevance. Click any page for the 90-day score history per topic and the full signal breakdown. - **Two grading modes you can pick per scan**. At the start of a Discovery run, choose Technical-only (no extra cost on top of the crawl: structural signals, schema, freshness, link depth) or Technical plus Content (5 credits per page for AI-graded depth and per-topic authority). Your last pick is remembered. - **Shared page-score index, Force-fresh override**. When a competitor URL is already in your platform's index and fresh, the crawl step is automatic. If you know a competitor just shipped something significant, tick "Force fresh crawl" to skip the index and re-pull live. - **Board Report Section 7**. The report now includes top brand pages per priority topic alongside competitor top pages, with the authority score and grading mode for each. - **New scan picker shows Web GUI engines**. The engine list is now split into two tiers: API tier (1 credit per cell) and Web GUI tier (the real product UI scraped in real time, 5 credits per cell: ChatGPT, Gemini, Perplexity, Copilot, Grok, DeepSeek, and Google AI Overview). Mix the two to balance cost against what a real user would actually see right now. - **Live progress on the Scans page**. Every running scan now shows a pulsing status pill, a live elapsed timer, and a thin progress shimmer along the row. The page auto-refreshes every 20 seconds while anything is in flight. - **Backlinks and Traffic now on Channels**. The Channels hub gains a Backlinks and Traffic card showing your brand's referring-domain count, the 7-day delta, and a 14-snapshot sparkline. - **Competitor crawls fail honestly**. When every page request gets blocked and zero pages come back, the scan now reports as Failed (not Completed). When more than one competitor is requested in the same kick-off, all of them get dispatched. The crawler also retries via proxy on a hard block instead of giving up after a single attempt. - **Each scan row shows its parameters**. Every row on Scans gets a small "params" chip. Hover or focus to see exactly which engines, country, page budget, and depth the scan ran with. - **Crawl-time and AI-suggested competitor socials**. Finished competitor crawls now extract LinkedIn, X, Instagram, Facebook, YouTube, and TikTok handles from the page automatically. When the crawl came up empty, a new "Suggest with AI" path proposes the brand's known handles for the user to confirm. - **Assistant history is easier to find**. On Assistant, empty "New conversation" stubs are hidden, and the mobile toggle now reads "History" with a count badge. - **Personas: AI suggestion fixed**. Suggest with AI on Personas previously returned a generic error for some brands; now generates 4 to 6 buyer personas reliably. ## 17 May 2026 - **Competitor Intelligence, backlinks, traffic, keyword gaps**. A new channel arrives behind your login. For your brand and every tracked competitor: total referring domains and domain rank, estimated monthly organic visits with month-over-month change, and the keyword gap list ranked by opportunity score (search volume times ease times competitor rank tier). - **New and lost backlinks feed**. Every week, the system pulls fresh backlink data and surfaces what changed: new referring domains gained, links lost (with the referring domain rank so reclaim outreach is prioritised). The feed rolls into the weekly admin email so you don't have to check. - **Board Report adds a Backlinks, Traffic, Keyword Gaps section**. Section 6 of every Board Report now includes the full competitive board: brand vs each competitor on referring domains, monthly visits with delta, the top 15 keyword gaps by opportunity, and the latest 30 days of new and lost links. - **Built for cost**. Every data point is cached with a time-to-live (backlinks 7 days, keyword portfolio 14, traffic 30). The "Refresh now" button is rate-limited to once per business per 24 hours. The weekly auto-refresh respects the cache, so it costs about 5 credits per business on a steady-state week. ## 16 May 2026 - **One brand, one home**. The product you knew as LLMSight is now BrandCrux, live at brandcrux.ai. Every bookmark, link, and email click-through still works. Older addresses redirect to the same page on the new domain. Your login, scans, credits, billing, API keys, and team members carry over unchanged. ## 15 May 2026 - **Board Report landing page rebuilt**. A clearer first impression: a copy-led hero with a centred "Starting at $99" gradient block, then eight alternating feature rows. Each row pairs benefit-led copy with an actual widget preview pulled from the product: composite scorecard, Topic Authority leaderboard, AI engine sources grid, diagnosis feed, 90-day Plays board, citation gap list, competitor tracking grid, and CMO memo. The sign-in card is now a full-width "Start your Board Report" panel with no step indicator on first view. - **Email sign-in switches to a 6-digit code**. Instead of waiting for a magic link to land and bouncing back to a new tab, the Board Report sign-in now sends a numeric one-time code that you can paste straight into the page. Continue with Google is still the one-click option above it. - **Delivery window: 24-48 hours**. Board Report turnaround tightened from "within 72 hours" to "24 to 48 hours" across the landing page, order confirmation, and emails. - **New: $99 self-serve Board Report flow**. The Board Report is now a guided product, not an email handoff. Sign in, walk through brand, competitors, topics, optional Search Console, prompts, engines, channels, pay. Each step persists as you go. The first 5 competitors, 5 AI engines, and 15 prompts are included in the $99 base; meters on every step show what you're choosing and any overages. Final order summary itemizes the breakdown and you can jump back to any step. A real BrandCrux account is provisioned at sign-in and activated with 1,000 credits on payment. The report is delivered in your account and emailed within 24 to 48 hours of admin approval. - **Coupons and targeted invites**. Admins can issue percent-off or flat-off coupons, global or scoped to a single email, with optional expiry and usage caps. 100%-off coupons skip the payment step entirely. Invite composer sends a personalized email to one or more recipients with the coupon attached and the recipient's brand context prefilled into the flow. - **Monthly nudge**. Customers with a Board Report older than 30 days get a one-line monthly reminder with a prefilled link to order a fresh one. ## 14 May 2026 - **(superseded, see 15 May 2026 entry) The BrandCrux Board Report, $499 one-time**. A comprehensive, board-room-ready assessment of your brand's visibility across every channel BrandCrux tracks, delivered as a signed ten-section PDF in 72 hours. Composite Topic Authority with a category baseline. Pillar contribution. Competitive battlefield, topic by topic, every named competitor benchmarked. Typed diagnoses ranked critical, warning, info. A 90-day Plays roadmap with forecasted lift and verification signal per Play. Citation gap list. Owned-channel synthesis from Search Console and GA4 written by a senior analyst. GEO scorecard. Flywheel state. And a three-page CMO memo. Every order includes 14 days of live workspace access, a 30-minute walkthrough call, and a full refund any time before delivery. No subscription required. ## 12 May 2026 - **Smarter credit pricing**. Credit cost per call now reflects the actual model your scan or content brief runs on. Heavy reasoning models cost a few more credits per call, light models stay the same, and free models stay free. Total credit usage for a typical scan is unchanged or slightly better. The credit approval card always shows the upfront estimate before you run anything. - **Automations, Watchers, Playbooks, and one-click push to your team tools**. A new Automations entry opens a three-tab control plane. Watchers are if/then rules: pick a signal (competitor content burst, composite score drop, topic share drop, search rank drop, citation went stale), set a threshold and a time window, and pick an action (spawn a Play, post to your team channel, file an issue, or send an email). Hit Test on any rule to dry-run it against current data without firing. Playbooks are pre-built multi-step sequences: two come stocked, "Close a citation gap on a top directory" and "Counter a competitor content burst". Run any Playbook on demand, or let it auto-fire when its trigger diagnosis lands. Destinations connects your team chat or issue tracker with credentials encrypted at rest. Every Play detail page gets a "Push to" button that drops the Play title, hypothesis, forecast, effort, confidence, verification signal, and a deep-link into a message or issue with one click. - **One Channels hub replaces seven separate sidebar entries**. AI Visibility, Web Search, Web Content, Citations, Socials, Search Console Insights, and GA4 Insights have collapsed into a single Channels entry. Clicking it opens a unified channel grid where each tile shows the channel's current KPI, last activity, and a 14-day trend sparkline. The legacy URLs all still work and every channel name still appears in the command palette. - **In-product schema editor lives inside every schema-verified Play**. Open any Play whose verification signal is "Schema present" and a side-by-side schema editor appears under Generate schema. Pick from ten schema types (Product, FAQ, Organization, Article, VideoObject, HowTo, Review, LocalBusiness, Service, BreadcrumbList), fill in the required fields with live validation, and watch the structured data render in real time. Copy it, download it as a file or a ready-to-paste script tag, or click Apply to Play to save the block to the verification target so the verifier compares the live page against the exact structured data you drafted. - **Content briefs can be generated straight from a Play**. When a Play's verification signal is "Content published", the detail page now exposes a "Generate brief from this Play" shortcut that drops you into Content Hub with the topic, hypothesis, and target URL pre-filled. - **Plays close the loop on every recommendation**. A new Plays page in the Act section consolidates your atomic recommendations into outcome-shaped units of work. Each Play rolls up one or more recommendations under a single hypothesis with a forecasted visibility lift, and binds a verification signal that proves the work shipped: a published URL with measured lift, a structured-data block appearing on a target page, a citation showing up on a target domain, or a sustained click delta on a target query. Ship a Play and an automated worker keeps polling the chosen signal until evidence lands, only then does the Play flip to Verified, and the realized lift is captured next to the forecast. - **Design system pass: consistent dates, statuses, and self-row markers across every page**. Recommendation categories now read as Title Case instead of all-caps slugs. Every status indicator renders through a single pill component with one consistent set of variants (success, warn, error, neutral). Every leaderboard row that points back at your own brand now uses the same green-dot "You" badge. Dates auto-flip from relative ("2h ago", "Yesterday") to absolute at the 7-day mark, and "Select a business" empty states share one short sentence plus a "Or create one" link. Persona KPI cards offer a "Run your first persona scan" call-to-action inside the empty cards, the Web Search SERP leaderboard adds a position-distribution column, and engagement-per-post numbers keep one decimal place. - **Recommendations and content briefs no longer pile up duplicates**. Near-duplicate recommendations now collapse into a single canonical row with a small "+N similar" badge you can click to see what was rolled up. The headline "estimated visibility points available" uses a diminishing-returns curve calibrated against your current score, so it stops claiming impossible numbers. Content briefs are deduped the same way. - **Content Hub now owns the full Brief, Draft, Published, Lift loop**. The separate Content Actions page has been folded in as the new Lift tab on Content. Every brief gets a "Mark as published" button that captures a baseline visibility score, schedules a 14-day post-scan, and surfaces the measured lift in the Lift tab once the post-scan completes. Old links to the content-actions page redirect to the Lift tab. - **Entity names are now deep-linked across every report**. When a recommendation mentions a competitor, topic, or keyword you already track, the name becomes an inline link straight to that entity's canonical page. Authority Topics competitor rows and topic-map columns open the matching competitor page. The Discovered Competitors panel drills into the channel that surfaced each candidate. Top citations on AI Visibility deep-link by domain. Scan History rows open AI Visibility focused on that scan. - **Intuitive URLs now work everywhere**. Typing the short form of ai-visibility, Search Console, GA4, citations, web, or topics now lands on the right page instead of an error. Bookmarks and shared links survive even when the canonical route uses a longer slug. - **Consistent browser-tab titles**. Every app page now reads "Page Name, BrandCrux" in the tab. - **New Scan button shows a credit estimate at a glance**. The topbar New Scan button now reads "New scan, about N credits" based on your active prompts and the default 3-engine AI run. Hover for the formula. The launch page still recosts precisely once you pick channels and tune parameters. - **Sidebar scan-status indicator is no longer mysterious**. The small clock next to AI Visibility is a scan-status badge, never a deprecation marker. Hover now explains exactly what it means ("A scan is currently running for this business" or "Auto-scan scheduled, daily at 09:00 UTC"). - **Connector failures now tell you what went wrong**. A failed Analytics or Search Console sync used to show a red "Errored" pill with no detail. Now every errored connector surfaces a plain-language hint, a "Show details" toggle that reveals the raw error for support, an "Errored 2h ago" timestamp, a one-click Retry sync button, and a Reconnect button that re-runs authorisation while keeping your historical data. Mission Control's "Sync owned analytics" step also turns red when any underlying connector is errored. - **Duplicate workspaces no longer count against your limit**. If two workspaces share the same name and domain, only one counts toward your plan's business limit. The workspace switcher now picks the most recently updated row as canonical and the dropdown ends with an always-visible "n / m workspaces used" footer. At the cap, the footer adds a tier-aware upgrade prompt and the "Add business" row goes muted with a tooltip. Workspaces never get deleted automatically. - **Brand names no longer split across pages**. Variants of the same brand name now resolve to the same row everywhere. A new canonical-entity layer keys every brand on its root domain and uses the name you saved in Tracked Competitors as the display label across Authority Topics topic breakdowns, the Web Search leaderboard, Citation Gap proofs, the Competitor Leaderboard, and the entity leaderboard. Tracked Competitor rows without a domain now show a "Domain needed" chip with an inline input. - **Web Search "Competitors tracked" now matches Tracked Competitors**. The KPI at the top of Web Search used to count every distinct domain seen in a top-10 result. The KPI now reads the tracker directly; the bigger number lives next to it as "Domains observed". - **Scan status is now accurate at the source level**. A scan where every source returned data is marked Completed; a scan where some sources came back empty is marked Partial; a scan where every source failed is marked Failed. Scan History also has a dedicated Partial tab. - **Failed and cancelled scans no longer pollute the dashboard average**. The headline visibility score and detection rate ignore scans that finished without data. The trend sparkline matches. - **Scan results no longer go missing on busy workspaces**. On workspaces with several big scans, older scans would sometimes show empty in the detail view even though the data was there. We now load each scan's results separately so every scan keeps its full row count. - **Mission Control header now matches the steps below it**. When every discovery step is already Synced, the wave banner now says "Discovery complete" instead of "Ready to start discovery", and the call to action becomes "Generate refined topics". The same applies to the scoring wave. If the credit estimate ever fails to load, the error block now offers both Retry and "Start anyway". - **Workspace selection now sticks across every page**. The active workspace, your plan badge, and the keywords count no longer flash back to defaults when you navigate to BrandCrux, Citations, Socials, Search Console insights, and other dashboards. We persist the workspace in a long-lived cookie. - **GEO Audit is now a sixth pillar in your Topic Authority score**. Your site's GEO Audit overall score now feeds the unified composite alongside AI mentions, web search, citations, socials, and website coverage. Better schema, semantic HTML, and entity clarity lift your authority on every tracked topic. Default weight is 10 of 100, adjustable per business. ## 11 May 2026 - **Keywords now opens sorted by impressions**. The Keywords tables land with the highest-demand keywords on top by default. Click any other column header to re-sort. - **Bell drawer "View" buttons now land on the right widget**. Every notification's View button now opens the exact page and widget the alert is about, not the catch-all alerts list. Toast clicks follow the same routing. - **Search Console insights, easier to scan, clearer when empty**. Quick-win queries, declining CTR, intent gaps, CTR vs. benchmark, site sections, and refresh candidates now render as a card grid that wraps cleanly at any width. The empty state distinguishes "Search Console isn't connected for this business" from "connected, no analysis yet". - **Run AI Visibility scans across multiple personas at once**. The New Scan launcher now lets you toggle any combination of personas as chips. Selected prompts run once per persona, with each persona's voice prepended, and the credit estimate updates live with an explicit "times N personas" multiplier. Skip all personas to run an aggregate pass. - **Real-time progress for every action you approve**. The topbar bell and a bottom-right toast now light up the moment a queued action starts running, again when it finishes with the result, and once more if it fails so you can retry in one click. Notifications arrive within a second of the event. - **Content brief suggestions from every GEO audit**. When a GEO Audit finishes, BrandCrux now picks the three highest-impact recommendations and offers to turn each one into a full content brief. The suggestions land in the topbar bell drawer with a credit estimate. No briefs ever get written without your say-so. - **Dashboards refresh themselves the instant you change something**. Add an authority topic, reassign a keyword, or update a persona and the dashboard widgets pick up the change on the next view without waiting for a cache to expire. - **Lifecycle Events page for support and engineering**. A new page shows the audit log of every change that travels through BrandCrux internally: topic edits, keyword reassignments, persona updates, competitor adds, GEO audit completions. Filter chips narrow by event family, and clicking any row expands the full before-and-after payload. - **Discover competitors with AI, proactively**. Competitors now has a "Discover competitors" button that asks BrandCrux's discovery engine to surface other companies you should be benchmarking against. It anchors on your authority topics, your products and services, your existing tracked set, and optionally a country, state, or city. Each suggestion comes with a one-line rationale, a confidence label, and a one-click Add to tracker. - **Scan History, every run, every channel, one page**. A new Scan History entry shows every scan you have ever run, across AI Visibility, Web Search, Citations, Socials, Search Console, GA4, and competitor crawls. Filter by channel, sort by started time, duration, status, or credits used, and click any row to drill into the full report. Four KPI cards summarise the last 30 days. - **New Scan is now a multi-channel launcher**. Pick any combination of AI Visibility, Web Search, Citations, Socials, Competitor Crawl, Search Console, and GA4 from one page. Each channel has its own parameter section that re-costs live, a sticky summary panel shows the total against your balance, and one click queues every selected channel and optionally sets a weekly or monthly cadence. - **Competitor social posts now cover every network**. LinkedIn, Instagram, X, TikTok, Facebook, and YouTube. Posts from each network land on the Socials page with the right platform icon, and the new Topic Authority scorecard breaks them down by topic, channel, and competitor. - **Free GEO Audit on every new brand, no setup needed**. The moment you add a business, BrandCrux now crawls the first 50 pages of your site breadth-first and scores them for Generative Engine Optimization on us. The free first run never touches your monthly credits. - **Personas, score your visibility per buyer archetype**. Define the buyer personas you sell to (or let BrandCrux draft them from your business profile). A new persona-scoped scan runs every tracked prompt once per persona, with the persona's voice and pain points framing the question, then tags every result with the persona it came from. - **Keywords and Prompts now share one home**. They're the same loop (demand to questions), so they live behind one sidebar entry with a tab strip at the top. - **Topic dashboards now populate after every scan**. A long-standing gap meant some scans completed successfully but the per-topic rollup that drives the BrandCrux composite score and the Authority Topics leaderboard didn't get written. New scans now populate both surfaces the moment each engine finishes. - **Search Console keywords auto-map to your topics on pull**. When you sync Search Console queries, the new keywords now cluster into your authority topics in the same request. - **Every channel now reads as one unit**. AI Visibility, Web Search, Citations, Socials, and Owned Analytics each have a sub-tab strip at the top of every page so you can move between Overview, Competitors, Citations, Posts, Queries, Pages, and Sources without leaving the channel context. New per-channel competitor leaderboards land for AI Visibility, Web Search, Citations, and Socials. - **Competitor share of voice now exists per channel**. AI Visibility shows who appears alongside you in AI answers. Web Search shows who outranks you. Citations shows who earns directory, review, and editorial coverage you do not. Socials shows who posts on your topics and how much engagement they earn. - **Socials Topic Authority scorecard**. A new sub-tab in the Socials channel rolls up the last 90 days of competitor posts into a Topic, Channel, Competitor grid. Authority topics lead the table; an "Off-topic" panel surfaces the other subjects competitors are creating content for. - **Search Console insights now read like a report**. The page now leads with a top-line snapshot (clicks, impressions, CTR, average position vs. the previous 30 days), then surfaces every analysis as its own card: quick-win queries on page two, queries losing CTR, pages serving conflicting intents, keyword cannibalisation, branded vs. non-branded split, plus country and device breakdowns. - **GA4 insights got the same treatment**. Engagement summary as a top KPI strip, channel mix shift and landing-page decline as side-by-side cards, executive summary and recommendations panels carry through. - **Authority Cards, live previews of every variant, in one click**. The page now shows live previews of all three variants (full card, compact strip, badge) for every card, side by side. New help guide explains where each variant fits. - **Share an Authority Card as an image**. From the embed modal you can now generate a PNG of the selected variant, share it straight to X, LinkedIn, or Facebook, or download it for slide decks. The image is hosted at a permanent URL. - **Track competitors across every domain they own**. A competitor can now have multiple domains (main site, blog, regional TLDs). Open the Competitors page, expand any row, and add as many domains as you like. Every crawl follows all of them. - **Stronger competitor crawls when sites push back**. If a competitor's site blocks our direct crawl, BrandCrux automatically retries the same domain through a proxy network (capped at 100 pages per domain). The crawl history records which domains needed the fallback. ## 10 May 2026 - **The Growth-tier intelligence suite is now reachable from your MCP client**. Anyone using BrandCrux through an MCP-aware tool can now ask for the proprietary intelligence layer in plain English. Fifteen new tools cover daily trend snapshots and anomaly hypotheses, classified competitor moves with recommended counters, end-to-end content actions, the attribution graph that explains why a metric changed, and embeddable Authority Cards rendered on demand. Same Growth-or-higher gate as the dashboard. - **Trend Intelligence works on day one for new brands**. Triggering an on-demand snapshot the first time you opened the page used to fail for newly onboarded brands with no history yet. Fixed; the page now lights up instantly. - **Mission Control, see your whole brand pipeline at a glance**. A new top-level page that visualises the entire brand-authority pipeline as a top-down flowchart of 13 stages across two waves. Wave 1 ingests the evidence (crawl your site, crawl competitors, pull socials, sync owned data). Then we re-recommend authority topics with the new evidence and surface the diff. You confirm refined topics and tracked prompts, then Wave 2 scores everything in parallel (AI visibility, web SERP, directory citations, social authority, owned analytics) and the composite Topic Authority score auto-recomputes. - **Embedded onboarding**. New users land on Mission Control directly. Tier 0 ("about your brand") renders as a prominent inline form with a grouped category typeahead (50+ standard categories, plus free-text for anything else). As you fill each prerequisite, the next stage unlocks. - **One-click bulk pipeline run**. A single approval card surfaces the per-thread credit breakdown for the whole wave, with checkboxes to opt out of any specific channel. Approve and Wave 1 fires every discovery thread in parallel; once you've reviewed refined topics and prompts, a second approval card fires Wave 2 the same way. - **Stage-completion notifications**. A new bell drawer in the topbar surfaces in-app alerts for every pipeline milestone. Email alerts fire too. Per-user opt-out from settings. - **Partial data flows in real-time**. Channel pages show a banner during active runs surfacing what's complete vs. still in flight. Composite scores recompute as each channel finishes. - **Resumable end-to-end**. Pipeline runs survive any backend interruption. A watchdog checks every 5 minutes for stuck stages and re-fires them safely. The state of every wave persists, so a crash mid-flight picks up exactly where it left off. ## 09 May 2026 - **Public site restructured around five channels and the closed loop**. The home page is now a top-level introduction to BrandCrux as a multi-channel brand-presence platform, leading with The Loop (declare, measure, see, act, re-measure) and the five-pillar grid. Each pillar lives at its own top-level URL. Each pillar page has a "Connected to BrandCrux" section that cross-links to the other channels and back to the composite Topic Authority score. - **Pillars dropdown replaces Features dropdown in the marketing nav**. The top-level nav now exposes the five channels first, then the AI Visibility sub-features under them. Mobile menu mirrors the same hierarchy. Footer adds a dedicated "Pillars" link section. - **BrandCrux, the crux of your brand across channels**. The product is now BrandCrux. Same closed loop, broader vision: AI Visibility (the original surface) is now one of five channels under a single brand-presence platform. The other four (Web Search, Authority Signals, Social Authority, and Owned Analytics) sit alongside it, with a composite Topic Authority score per topic that blends every channel into one ranked leaderboard against every competitor we observe. - **New landing page: five channels and Topic Authority callout**. The home page leads with "The crux of your brand across every channel," shows a five-pillar grid, then a Topic Authority composite band, then deeper coverage of the AI Visibility channel. The FAQ is rewritten to address the broader scope. - **Four new feature pages**. Web Search (Google, Bing, DuckDuckGo, Yandex rankings with featured snippet, knowledge panel, and PAA capture), Authority Signals (directory and publisher citations across G2, Capterra, Crunchbase, Trustpilot, editorial publishers, forums; Gap List for missed directories), Social Authority (last-90-days competitor posts across Instagram, LinkedIn, X, TikTok, YouTube, Facebook with engagement scoring and content pack generator), and Owned Analytics (Search Console and GA4 deep insights with pre-built pivots and 5 specific recommendations per run). - **Marketing nav reorganised by pillar**. The Features dropdown now groups by pillar: AI Visibility (the existing 8 sub-features), Beyond AI (the 4 new pillars), and Cross-cutting (Integrations). The 9 existing AI Visibility feature pages each get a "Channel 01, AI Visibility" eyebrow above the hero. - **Favicon, social image, and meta updated**. Browser-tab icon is the BrandCrux teal radar mark in SVG form. Social previews reflect the new positioning. Sitemap and llms.txt rewritten. ## 08 May 2026 - **Sidebar streamlined to four-or-fewer items per group**. Navigation rebalanced so no group exceeds four entries: Strategy (Topic Authority, Keywords, Prompts, Competitors), AI Visibility (Dashboard, Scans, Citations, Trends), Search Visibility (Web Search, Search Console Insights, GA4 Insights, Traffic), Authority Signals (Directory Citations, Social Authority), Improve, Setup, Account. The page title shows once in the topbar instead of being repeated inside the page. - **Discover Authority moves into Topic Authority**. The discover-from-competitor-crawls wizard is now an inline panel on the Topic Authority page rather than its own separate page. The panel surfaces a count of pending discovered topic suggestions inline. - **Help walkthroughs on every visibility page**. Web Search, Directory Citations, Search Console Insights, GA4 Insights, and Social Authority each gained a Help launcher that opens a step-by-step walkthrough of how the page works, why each signal matters, and what to do with the data. - **Sidebar reorganised around the closed loop**. Navigation groups now read top-to-bottom as Strategy (Topic Authority, Keywords, Prompts), Measure (Dashboard, Scans, Citations, Competitors, Trends, Traffic), Improve (Alerts, Recommendations, Content, GEO Audit), Setup (Connectors, Settings, Team), Account (Billing, API, Refer and Earn). Connectors moves into Setup so the data foundation is more discoverable. - **Onboarding now connects Search Console and confirms authority topics before the first scan**. Two new optional steps: Demand (connect Search Console with one click, paste keywords manually, or skip) and Authority (accept AI-suggested authority topics for your brand). When you reach the prompt-selection step, the suggestions already reflect your real demand signal and the topics you want to own. ## 08 May 2026 - **Visibility expansion, five new dimensions of brand presence, all on one Topic Authority page**. Topic Authority now spans every channel that shapes how people find a brand: AI search engines (where it started), competitor website crawl depth, social engagement on competitor handles across Instagram, LinkedIn, X, TikTok, YouTube, Facebook, directory and publisher citations across G2, Capterra, Yelp, Crunchbase, editorial publishers, and standard web search rank on Google, Bing, DuckDuckGo, Yandex. Each dimension also gets its own dedicated page. A composite Authority Score per topic blends all five into one number, with weights you can tune in settings. - **Discover Authority Topics from competitor websites**. A new Discover Authority wizard lets you pick competitors, set a per-site page budget, and we crawl their websites breadth-first and propose authority topics they're winning that you should consider tracking. Pages are classified into a topic taxonomy in real time; the wizard's last step lists each proposed topic with the supporting URLs and a one-click Accept button. - **Directory and publisher citation tracking**. A new Directory Citations page surfaces where you and your competitors are listed across directories (G2, Capterra, Yelp, Crunchbase, etc.) and editorial publishers. Per-target citation leaderboard plus a gap list. New scans run on demand and can be scheduled weekly or monthly. - **Standard web search visibility tracker**. A new Web Search page tracks where you and competitors rank on Google, Bing, DuckDuckGo, and Yandex for any keyword you choose. Per-keyword position history with sparklines, featured-snippet ownership tracking, knowledge-panel ownership, and a per-competitor SERP-share leaderboard. - **Social authority via competitor post analytics**. A new Social Authority page connects to your social data token, pulls the last 90 days of posts from every tracked competitor's social handles across six networks, classifies each caption into your topic taxonomy, and ranks the per-topic engagement leaderboard. A built-in caption-pack generator drafts new posts for the brand on any topic, modelled on the highest-performing competitor posts on that topic. - **Search Console and GA4 deep AI analysis**. Two new pages turn your existing Search Console and Analytics connections into full analytics surfaces. Pre-computed pivots feed into an AI narrator that produces an executive summary plus 5 specific recommendations on each run. Recommendations land directly in your Recommendations queue. - **Pre-approved credit budgets for every scheduled scan**. Every scan type now flows through a single pre-approval card: see the credit estimate broken down line-by-line, your balance after the scan, and approve before anything runs. Recurring schedules auto-pause if your balance dips below one cycle's worth, with email alerts. - **Plan credit pools bumped to absorb the new dimensions**. Starter is now 6,000 credits/mo (was 3k), Growth is 30,000 credits/mo (was 12k), and Scale is 80,000 credits/mo (was 25k). Free plan stays at 300 credits. Existing subscriptions auto-receive the bigger budget on the next billing cycle. ## 07 May 2026 - **Mobile-friendly onboarding, full-page scrolling, and a desktop-best banner on shared reports**. The 6-step onboarding flow now collapses cleanly on narrow screens: the step strip drops the connector lines, hides labels for steps you're not on, and keeps the active step labelled. Form fields stack to a single column, and onboarding adds a bottom buffer so the Continue and Back action row is always reachable. Shared reports now also lead with a clearer top banner on phones explaining that some data tables won't fit a small screen. - **Topic Authority, a new page that ranks every topic against every competitor we observe**. A new Topic Authority page lets you declare which topics your brand wants to be known for, then ranks you against every competitor we extracted from AI responses. Each topic carries a recommendation badge (Lead, Defend, Compete, Exit) based on your share of voice plus the topic's traffic potential. One-click Generate brief on Compete-flagged topics queues a content brief anchored on that whole cluster. - **Recommend authority topics from your data**. On Keywords, Manage topics, a new Suggest authority topics button drafts 8-12 candidate topics based on your business profile, top Search Console keywords, and the competitors we already track. Accept the ones you want; the marker on each topic flips them into the Topic Authority page automatically. - **Page-category and topic badges on every cited URL**. The Citations page now classifies each cited URL into a content category (API Docs, Blog, Pricing, Case Study, etc.) and maps it into your topic taxonomy. Classification runs automatically after every scan and caches per page. - **Content-coverage badge on each topic**. Each topic header on the Keywords by-topic view now shows how many of your own pages have been cited under that topic. A "No own content" badge flips on for topics where AI engines never reach for your content yet. - **Citations page no longer truncates the long tail**. Previously the page capped cited pages at 100 across all domains. Now every own-domain page we ever tracked is surfaced, and external domains get a more generous 250-page cap. - **Curated topics, define your own taxonomy and have AI map keywords into it**. A new Manage topics action lets you upload your own list of topic names. When curated topics exist, the AI clustering switches to a constrained-classification mode and routes every keyword into one of your labels. Leave the list empty and the workspace continues to free-form cluster as before. - **Topic visibility trends, sparkline plus 90-day chart per cluster**. Each topic header now carries a 90-day visibility sparkline. Expand the topic and you get a chart panel showing how AI visibility for that cluster has moved scan to scan, plus a delta callout. - **Brief impact, did this content actually move the needle?**. Every brief on the Content page now carries an Impact pill once an artifact ships and at least one scan runs after the publish date. Hover for the before-and-after sample sizes and publish date. - **Top competitor pill on every topic**. Each topic header surfaces who is winning AI mentions on that cluster. Hover to see the top three competitors and their share. - **Topic-level alerts, visibility drops, competitor breaches, content gaps**. Three new alert types fire after every scan: Topic visibility drop (cluster fell more than 8 points), Topic competitor breach (a competitor crossed your share for the first time), and Topic position loss. All land on the existing alerts page. - **Keywords now group by topic, and act as a group**. The Keywords page has a new By topic view that AI-clusters your positive keywords into semantic groups. Each topic header carries the cluster's total impressions, weighted AI visibility, and prompt count, plus two one-click actions: Generate brief and Auto-map to prompts. Topics refresh every Monday automatically and can be recomputed any time. - **Generate prompts in your customer's language**. The prompt generator now accepts a language picker. Pick from 20+ languages (Spanish, German, Portuguese, Japanese, Hindi, Arabic, Mandarin, and more) and we'll produce buyer-intent prompts in that language. - **Postal codes on Scan Location for tighter local targeting**. Add a ZIP, postcode, or PIN to your scan and we inject it into the prompt context for both web and API sources. - **Bulk-import competitors from a CSV**. On Settings, paste an Import CSV file with one competitor per row (optional second column for domain) and we'll merge them into your competitor list, skipping any that are already there. - **By Platform headline cards on the dashboard**. A new tile strip rolls up web and API visibility into one number per AI platform (ChatGPT, Perplexity, Gemini, Claude, Copilot, Grok, Google AI). - **Visibility trend pill on the headline score**. The dashboard now shows a percentage change pill next to your visibility score so the direction of travel reads at a glance. - **Downloadable visibility report PDF, white-labelled where enabled**. The PDF button on the dashboard now produces a real downloadable report (KPIs, by-platform rollup, top prompts, citations, competitors). White-label customers get it rendered with their own brand name, logo URL, and accent color. - **Keyword opportunities now land on your dashboard**. The dashboard surfaces the top 5 high-impression search queries where AI engines aren't picking you up yet. - **Opportunities tab on Keywords, high-impression terms where you're losing**. A new Opportunities filter surfaces positive keywords with at least 100 impressions in the last 90 days where you're either sitting in low search positions, AI engines aren't mentioning you, or you're not yet tracking them. Each opportunity row gets a Generate brief button. Bulk-select to queue several at once. - **Promote any keyword to a tracked AI scan prompt, one click**. On the Keywords page, every positive keyword now has an Add as prompt button. Click it and we create a tracked prompt anchored on that keyword. Bulk-select multiple keywords and add them all at once. The keyword row then shows live AI visibility. - **Keyword-to-prompt cross-linking**. Expand any prompt on the Prompts page and you'll see chips listing the keyword(s) that drove it. Click a chip to jump straight to that keyword's row. - **Cleaner Keywords and Content surfaces**. The duplicated stat-cards-plus-tiny-chips pattern on both pages is gone. Counts now live inline on a single prominent tab strip. ## 06 May 2026 - **Downloadable PDF invoices on every successful payment**. Click Invoice on any paid row in your billing history to get a professional PDF with full BrandCrux legal details, your billing address, line items, GST breakdown for Indian customers, and the payment reference. Replaces the previous plain-text receipt. - **Billing details required before checkout**. Before you subscribe, top up, or authorize auto-recharge, we now ask for your billing address (one-time, then remembered) and an optional tax ID. The form is country-aware: India validates your GSTIN; the EU asks for your VAT number; the US asks for an EIN; and so on for 12+ countries. Tax IDs land directly on every invoice. - **Keyword workspace, bring your existing SEO keywords into prompt generation**. A new Keywords page lets you add the words your customers actually search. Positive keywords feed customer vocabulary into AI prompt generation; negative keywords are excluded from every generated prompt. Manually paste them, upload a CSV, or click "Suggest from profile" to have AI propose 25 seed keywords. On Starter and above, click "Pull from Search Console" to auto-import your top queries from the last 90 days. Newly pulled keywords land in a "Pending review" queue. A weekly auto-refresh keeps the workspace in sync. - **No more "no data yet" flicker on slow networks**. Pages like Scans, Prompts, Competitors, Citations, and Trends used to flash a "run your first scan" empty state for half a second on slow connections before the real data appeared. The app shell now does a single tiny pre-flight call when you switch businesses to learn upfront whether each page has data. ## 05 May 2026 - **Partner Program, earn 25% revenue share for the lifetime of every customer you refer**. Marketing agencies, consultants, and creators can now apply to the BrandCrux Partner Program. Approved partners get a unique link, a dedicated portal to track signups and conversions, and earn 25% commission on every paid invoice for the lifetime of the customer. Settlements run monthly; balances above $100 pay out within 7 business days via bank ACH, PayPal, wire, or card network. Partner Program and Refer and Earn run independently. - **Refer and Earn, get 100 free credits when a friend signs up, 500 more when they upgrade**. Every account now has a personal invite link in the new Refer and Earn sidebar entry. The moment someone signs up via your link, both of you get 100 bonus credits in your top-up balance. If they pick a paid plan, both of you get another 500 credits each. We'll also nudge both sides 72 hours after signup if the referee hasn't upgraded. Bonus credits never expire. Self-referrals and obvious abuse patterns are blocked. ## 04 May 2026 - **Authority Score now floors to zero for uncited domains**. Domains with no own citations were getting an Authority Score in the 30s because the gap and trend sub-scores happily returned high values when there was nothing to compare against. Now if no AI engine has cited you, your score is 0 across both the dashboard widget and the Citations page. - **Auto-recharge low-balance alerts fire as soon as you dip**. Previously a single alert per 24 hours. Now you get notified the first time your balance drops below your configured threshold, then a one-hour cooldown before the next reminder. - **Cleaner upgrade prompts on plan-locked pages**. The API Playground and Traffic Analytics pages no longer load a half-empty UI for plans that don't include them. Free and Starter accounts now see a single, honest upgrade card that says exactly which plan unlocks the feature and what it costs. - **Free scan no longer infers the name "Www" when you paste a www URL**. Submitting a www address on the free scanner used to produce a report titled "Www" because we'd split the hostname on the first dot. Now we strip the leading www first. - **Signup notifications added to the internal ops feed**. Same lightweight admin email system as scans, subscriptions, and top-ups, now also fires on every new account creation, with the new account's name, email, signup method, IP, and geo. ## 03 May 2026 - **Internal ops notifications**. Lightweight admin-only emails now land for every free scan, every subscription event (new signup, upgrade, downgrade, renewal, cancellation, payment failure), and every top-up (manual or auto-recharge). Lets the team spot oddities without staring at the dashboard. Configurable per category. - **One scan-complete email instead of two, with the full dashboard report inside**. Every scan used to send a separate "alert" email next to the scan-summary email. Consolidated to one email that opens with a dashboard-style KPI strip (mentions, detection rate, competitors, engines), keeps the per-engine breakdown table, and now adds two real comparison sections vs your previous scan: "Who's Showing Up Instead" shows top competitors with delta arrows and a callout when a brand-new competitor appears, and "Citation Sources" lists the top domains AI engines cited along with how many citations were gained or lost. - **MCP server is now auto-discoverable**. Agent platforms can now find BrandCrux automatically through a discovery card. As more clients adopt the convention you'll see "Add BrandCrux" in their MCP browsers without needing to paste the URL by hand. - **Comparison pages**. Side-by-side breakdowns of BrandCrux vs Profound, Otterly, Peec AI, Rankscale, and Scrunch are live at the compare hub. Pricing, feature matrices, and an honest "when to pick them instead" callout for each. ## 02 May 2026 - **Bigger plans, simpler structure, fresh credits every month**. Free goes from 100 to 300 credits (still API sources only). Starter from 300 to 3,000. Growth from 1,200 to 12,000. Scale drops to $179 (was $299) and goes from 5,000 to 25,000 credits. Rollover is gone, every cycle starts with fresh credits, and top-up credits never expire if you want to bank capacity. Top-up packs were also rebuilt: 500 / 1,500 / 3,500 / 6,000 credits at $10 / $25 / $50 / $75. - **Enterprise tier is here**. Need higher limits, SSO and SAML, audit logs, custom data residency, dedicated success manager with 99.9% SLA, custom integrations, or white-glove onboarding? Reach out to sales@brandcrux.ai. - **API and MCP can now manage alerts, billing, and auto-recharge**. The public REST API and the MCP server gained a dozen new tools: list and mark alerts read, read your plan and credit balance, list payments and the credit ledger, and inspect or update auto-recharge settings. First-time payment-method authorization still happens in the dashboard; everything after that is scriptable. - **API keys now use the llms_ prefix**. New keys are issued with the llms_ prefix. Existing keys continue to work as-is. - **AI responses now render as clean Markdown**. When you expand a prompt result on the Scans page or open a shared scan, the AI's answer now renders from its Markdown form instead of the raw scraped HTML. Same content, in your app's typography, with cleaner headings, lists, and links. - **Per-run picker for high-confidence scans**. When a prompt is sampled multiple times against the same source, the Scans page and the share report now show a small row of run pills above the response, each pill is one run, color-coded by what the AI said about you and tagged with the run's timestamp. Click a pill to load that run's response. - **Competitors, Ranking Breakdown shows every prompt**. The widget was capping the table at 10 rows. It now lists every prompt you track. - **Prompts, expanded view now reconciles top cards with the heatmap**. On a high-confidence scan, the source cards above and the historical heatmap below could disagree. They now both use the run that actually scored. ## 01 May 2026 - **Time-remaining and notification note on running scans**. While a scan is in flight, the running banner now shows about how much time is left and a quiet line telling you a notification will appear when the scan finishes so you can close the tab. Updates every 5 seconds. - **Cancelled scans now show what they collected**. Cancelling a scan mid-flight used to leave the Scans page blank for that run. It now displays every prompt and source result that completed before the cancel, with a Cancelled chip on the row. - **Scans no longer double-bill on infrastructure retry**. When a scan was interrupted by a transient hiccup, the recovery path was re-running every cell that had already finished, charging credits a second time. The orchestrator now resumes from where it left off. - **Auto-recharge that actually charges**. Authorize a payment method once on the Billing page and we'll top your credits back up automatically the moment your balance crosses your threshold. Works with cards (domestic and international) and UPI Autopay; you choose USD or INR. If a charge ever fails we email you, retry once a day later, and pause auto-recharge after a second failure. - **Error monitoring on the way**. We now have automatic error capture across the app. When something breaks we get notified in real time so we can fix it without waiting for someone to report it. - **Confirm-before-flip on prompt status**. The Active and Paused control on a prompt's expanded view is now an explicit "Mark Active" or "Pause prompt" button with a confirmation dialog. - **Weekly auto-scans on Starter**. Starter plans now run weekly auto-scans every Monday. Active prompts carry a small calendar icon next to the query so you can see at a glance which prompts run on the schedule. - **Country-aware location field on New Scan**. The optional second field next to Country now adapts its label to the selected country: "State" for US, India, Mexico, Brazil, Australia, "Province" for Canada, South Africa, Indonesia, "County" for UK, Ireland, "Prefecture" for Japan, "Canton" for Switzerland, "Region" for France, Italy, etc. - **Watchdog for upstream provider credit**. Every 15 minutes we now check the credit balance with our scraping and AI providers. If a provider runs dry, the affected sources are paused immediately so new scans skip them with a clear note. As soon as credit is replenished, the sources come back online. - **Backup scraping path for paid plans**. When the primary scraping route can't deliver a cell, Starter, Growth, and Scale scans now have a second path that opens a real browser session against the AI source and reads the response directly. Free plans continue to use the primary route only. Off by default. - **Big scans on the same source go out as one batch**. When a scan has 4 or more prompts on the same web GUI source, we now bundle them into one upstream job instead of sending each one separately. Off by default while we validate. - **One scan per user, faster lane for clicks**. Each account is capped at 5 simultaneous scans, and user-clicked scans now jump ahead of background work like weekly auto-scans. - **Web GUI scans now keep their distance from each provider's queue**. Each AI source has its own concurrency cap so a heavy scan burst on ChatGPT no longer queues up behind itself. Faster sources can still run wide; slower or rate-limited ones hold back. - **Per-source patience is now realistic**. Some AI sources legitimately take 7 to 13 minutes when they're under load, others wrap up in 30 seconds. Every source now has its own deadline tuned to that range. - **Quieter ops alerts when the provider is just slow**. Our hourly scan-health alert now differentiates: soft outcomes (timeouts and rate limits) are tracked but no longer page support; hard outcomes (captcha, blocked, malformed, content filter) still trigger an alert. - **Every scan attempt now carries the upstream scrape ID**. When something does go wrong on a web GUI source, the row in our scan-attempts log carries the provider's snapshot ID. ## April 30, 2026 - **Source cards on the Prompts detail now match the heatmap**. On a prompt's expanded view the source cards at the top would say "Missing" for nearly every engine even when the score table below clearly showed scores. The cards were picking an arbitrary older scan instead of the most recent one. They now use the same latest scan the heatmap does. - **Google Search source temporarily off**. Google Search has been disabled across scans while we finish the underlying scraping path. Existing scans show their historical Google Search data; new scans will skip it without consuming credits. - **Removed Citation Volume Trend from the Citations page**. The trendline at the top of the page wasn't adding signal beyond what the Citations Over Time and Citation Sources widgets already cover. - **Slow ChatGPT and Gemini scans no longer get marked failed**. Web GUI scans now use an async trigger and poll workflow. A single long job (ChatGPT runs can go 5 to 13 minutes when the model is slow) used to time out our connection at 3 minutes and report a failure even though the scrape was finishing successfully. The connection no longer stays held open for the full job, per-call timeouts are tight, and the overall budget is 10 minutes. - **Admin can actually grant credits and change plans now**. Plan switches and credit grants from the admin Users page used to fail silently. They now route through a privileged server endpoint that updates both the legacy plan column and the canonical subscription plan column, sets the right monthly credit base, business cap, and resets the cycle counter. Every change is written to the audit log. - **Grant up to 100,000 credits in one shot**. The admin credit dialog now goes up to 100,000 per action with one-click presets at 100, 500, 1,000, 5,000, 25,000, and 100,000. Negative numbers revoke. - **Optional email when an admin changes a plan or credits**. Both admin actions now have an "Email this user" checkbox with an optional note. If you tick it, the user gets a clean email from billing@brandcrux.ai summarizing the change. - **Refine drafts with one credit**. The article editor has a new Refine panel. Type what you want changed and the article is re-written by the same model that produced it, then reviewed by the same critic if one was used. Costs 1 credit per refinement and saves a new version every time. - **Drafts editor now actually shows the draft**. Opening a generated article (Markdown body, Meta Title, Meta Description, slug, focus keyword) was loading blank because the editor was looking at the wrong field. The split editor, preview pane, SEO panel, and version selector now populate correctly. - **Cleaner slugs and full meta titles**. Article slugs are now proper URL slugs (no colons, question marks, or stray punctuation; collapsed hyphens; lowercase). Meta titles are no longer hard-truncated mid-word at 60 characters; you see the full title in the editor with a live counter. - **Generate Image and Export now work in the editor**. The Generate Image button now correctly fires the image generation flow with a prompt derived from the article title. Markdown, HTML, and JSON export all work. - **Removed the Competitor Mentions Over Time chart**. The trend lines on the Competitors page often showed a single dot or noisy data. The page now goes straight from Share of Voice into the per-competitor leaderboard with threat scores and head-to-head breakdowns. - **Per-sample dots and selectable Confidence on every scan**. Each prompt and source row in the scan detail now shows tiny green and grey dots (one per sample run). The scan setup form has a new Confidence selector (Standard, High, Maximum equals 1x, 2x, 3x samples per cell). The credit estimate updates live with the multiplier. Higher tiers tighten the confidence interval. - **Scan detail now shows every prompt for every engine, even failed cells**. Previously, prompts where every sample retry failed silently disappeared from the per-engine table. The view now fills in the missing cells with a "Failed, all retries failed" row. - **Web GUI scraping failures now retry properly**. When the browser-capture path fails for a query, the error string used to be classified as "unknown" and skipped. Now it's classified as an upstream error, which is in the retryable list. - **"What Changed" banner is gone when there's nothing to say**. The dashboard now hides the banner unless the narrative is a substantive sentence; the email digest does the same. - **Scan detail no longer repeats the same prompt 10 times**. Multi-sample scanning (we run up to 10 attempts per cell) was leaking through the UI as one row per sample. The expanded scan view now aggregates samples into a single row per prompt, with a small "found / total" sample badge. - **Prompts page splits Active and Paused**. A new Active, Paused, All filter makes it clear which prompts are part of ongoing scans versus generated-but-not-selected or manually disabled. Defaults to Active. - **Competitors page now aggregates across the last 5 scans**. Latest-scan-only under-reported competitors for businesses with mature scan history. Threat scores, win and loss counts, and the per-competitor leaderboard now use up to 5 recent scans. - **Competitors and All Entities are separate tabs again**. Competitors is the default tab. All Entities is the secondary tab. Onboarding-listed names get a marker in both. - **Restored the original Competitors KPIs and per-competitor leaderboard**. Brought back the "Competitors Found / You Win / You Lose / Both Beat" KPIs and the per-competitor leaderboard with threat scores and per-query head-to-head breakdown. - **Phantom self-mention bug fixed on the Competitors page**. A scan would sometimes show "Your Share of Voice" as 1% even when your brand wasn't mentioned at all. Both the dashboard widget and the analytics endpoint now skip these placeholders. - **Headline percentages no longer trail with long decimals**. The "Your Share" KPI and the per-entity share percentages are now rounded to 2 decimal places. - **Optional location injection for Web GUI prompts**. A new checkbox lets you prepend "I'm based in [city, state, country]." to the query typed into web AI engines. Off by default. - **Partial scans now show up everywhere**. Dashboard, Prompts, Competitors, Traffic, Trends, Citations, Scans, and shared report pages now treat a partial scan as a regular completed scan for display purposes. - **AI Assistant icon visible to every plan**. Free and Starter users now see the sparkle chat icon in the top bar. Opening it shows a locked card with example questions and a clear path to Growth. - **Hourly scan health check and daily digest emails**. A new background check runs at the top of every hour and emails support@brandcrux.ai a per-engine table only when something actually failed. A separate daily digest covers totals, status breakdown, and per-engine success rates over the previous 24 hours. - **API-tier scans now retry every cell up to 3 times**. Bumped the API tier to match the Web GUI tier so transient gateway timeouts have one more attempt before a cell is dropped. - **Scan setup form polish**. The City input no longer overflows the card on narrow viewports, the "Run Scan" button no longer silently does nothing for paid users with more than 5 prompts, and Website is now required for both online and local businesses. - **Competitors page is now one unified leaderboard**. Dropped the "Competitors / All Entities" tab split. Every brand AI sources actually mentioned for your prompts shows up by default. Brands you also listed as competitors during onboarding are marked. - **Competitors page now opens on All Entities**. The default tab now shows every brand AI engines mentioned for your prompts, not just the small list you typed in during onboarding. - **"Citations by Source" widget tells the truth when you're uncited**. When no engine cites your domain yet, the subtitle now reads "Engine citation activity (none cite your domain yet)" instead of the misleading "Which AI engines cite you most". - **Authority Score Trend stays neutral when there's nothing to trend on**. A new business with two zero-citation scans is no longer flagged as "declining"; we report 50 (neutral) until there's actual citation history. - **Citation Rank now reads zero when you aren't cited**. Previously, an uncited domain produced a perfect Citation Rank of 100. Authority Score now correctly reflects an uncited state across all five components. - **Claude API scans no longer drop slow cells**. Slow Claude responses (sometimes 100 to 180 seconds on long prompts) used to fail and weren't retried. We now classify these correctly as timeouts and gave Claude a 180-second per-attempt budget. - **City and state targeting for API sources**. Scans now accept City and State alongside Country. We expand country codes to full names before sending them to AI sources. - **Scan summary emails now show competitors and next steps**. The post-scan email surfaces the competitors AI sources mentioned ("Who's Showing Up Instead") and links you straight to the three actions that move the score: review recommendations, find content gaps, and run a GEO audit. - **Cleaner engine names in scan emails**. Engine rows now read "ChatGPT", "Claude", "AI Overview", and "Google Search" instead of raw API IDs. - **Web GUI sources now require Starter or higher**. Free plans run scans against the 8 API sources. Web GUI sources (ChatGPT Web, Perplexity Web, Gemini Web, Copilot Web, Grok Web, AI Overview, Google Search) unlock with Starter. - **Smarter onboarding scan setup**. The first scan now defaults to API sources (1 credit each) for Free plans so your first scan always fits in the credit allowance. You can toggle individual sources on or off and see live credit math. - **Fixed onboarding showing both "Scan failed" and "Scan running"**. The error state now hides the queued message so the result is unambiguous. - **Removed Google Maps as a tracked source**. Google Maps is no longer offered as an AI source. We focused our coverage on the AI assistants and SERP surfaces where mentions actually move recommendations. ## April 29, 2026 - **Publish to 5 CMS platforms from Content Hub**. You can now push AI-generated content directly to WordPress, Shopify, Webflow, Wix, and Squarespace. Connect your CMS from the Connectors page, then use the new "Publish to CMS" dropdown on any content artifact. Each platform uses API token authentication. - **CMS integration guides**. Step-by-step setup guides for each CMS connector are now available. Each guide covers prerequisites, credential generation, connecting in BrandCrux, and troubleshooting. - **Team members now inherit the business owner's plan features**. Previously, invited team members were blocked from Growth and Scale features like Content Hub, Connectors, and AI Assistant because the system checked their own plan. Team members now see and use all features available on the business they are accessing. ## April 29, 2026 - **API and MCP coverage for the new sampling fields**. The public REST API and the MCP server now expose every confidence-aware field added this session, including presence rate, confidence interval bounds, sample counts, quality flag, mention rank, score standard deviation, and per-sample scores per cell. New endpoints surface per-pull attempt diagnostics and a confidence block with degraded, suspect, and insufficient cell counts. MCP gains tools so AI assistants can reason about measurement confidence. - **Confidence-aware Ranking Breakdown**. The Dashboard and Competitors page rank widgets now show per-cell sample counts, data-quality dots, and a 95% confidence interval on the aggregate "Avg Rank when listed" KPI. Insufficient cells are excluded from averages. - **Methodology article and landing-page explainer**. A new "How BrandCrux Measures AI Visibility" article walks through multi-sample probing, confidence intervals, decomposed judging, and quality flags. Linked from the landing page and the AI Visibility feature page. - **Citation Gap Analysis is now the canonical "earn coverage" view**. The aggregated Source Recommendations card has been removed from the Citations page. The Citation Gap Analysis section already surfaced the same signal in a more useful form: for each domain, query, and engine where AI cited a third-party domain and named your competitors but not you, you see exactly which publisher to chase and why. - **Confidence-aware visibility scoring**. Every visibility cell is now sampled multiple times per scan and rolled up into a presence rate with a 95% confidence interval. Higher tiers get more samples for tighter bands. - **Adaptive deeper sampling on uncertain queries**. When a query is on the edge (mentioned roughly half the time), we automatically run more samples to lock in a tighter result. - **Data quality flags**. Every cell is tagged ok, degraded, suspect, or insufficient based on how many attempts succeeded and what failed. Suspect cells are flagged rather than silently averaged in. - **Source Recommendations**. A new section on the Citations page ranks the third-party domains AI sources cite most in your category, with multi-engine consensus and breadth signals. - **Two-stage judge for sharper classification**. The visibility judge is now split into a structural pass (entities and mention type) and a sentiment pass that only fires when something was mentioned. - **Versioned extractions**. Each scan result now stores the verbatim engine output alongside a judge fingerprint, so when we improve the judge we can replay it on history and compare apples to apples. - **Per-attempt diagnostics**. Every fetch attempt, successful or not, is logged with a categorised outcome (timeout, captcha, content filter, network error, etc.). - **Query topic and intent grouping**. Queries can now be tagged with a topic and an intent (informational, comparison, recommendation, problem-led, transactional, local), with paraphrases linked back to a canonical query. ## April 29, 2026 - **AI Assistant actions**. The AI Assistant can now take actions on your behalf: run scans, add, pause, or delete tracked prompts, generate content briefs, create content, and update business details. It always asks for confirmation before executing. All actions are security-scoped to your account only. - **Shared Share of Voice component**. The Dashboard, Competitors page, and shared reports now all pull from a single backend endpoint, eliminating ranking mismatches between pages. - **Smarter traffic comparisons**. Period-over-period deltas on the Traffic page are suppressed when the comparison period has insufficient data, preventing misleading huge changes. Traffic page now defaults to 14 days. - **Send Feedback**. Added a "Send Feedback" option to the user menu in the sidebar for quick access. - **Upgrade modal light mode fix**. The upgrade prompt is now properly themed in light mode with readable text and visible buttons. - **AI Assistant side panel**. Open the AI Assistant from any page using the floating chat button in the bottom-right corner. The assistant knows which page you are viewing and tailors its responses. Available on Growth and Scale plans. - **Login page layout fix**. The sign-in page no longer appears squished on narrow browser windows or when accessed via invite links. - **Fortnightly GEO audits**. GEO audits now run automatically every other Wednesday night. You'll receive an email Thursday morning with your updated scores, category breakdowns, and changes since the last audit. - **Weekly auto-briefs**. Every Tuesday, BrandCrux automatically generates fresh content briefs based on your latest scans and newly discovered visibility gaps. Available on Growth and Scale plans. - **Content Hub guide and focus areas**. The Content Hub now has a detailed info tooltip explaining the full workflow. Creating a brief lets you pick a focus area (Product, Content Gap, Topic Authority, FAQ, Comparison, Case Study). - **Model picker for content generation**. Choose which AI model generates your content. Pick from free models for quick drafts, standard models for polished content, or premium models for deep research and critique. Each shows its credit cost upfront. - **Current-year content recommendations**. Content advisory and generated briefs now always reference the current year and date. - **Traffic by property breakdown**. When multiple Analytics or Search Console properties are connected, the Traffic page now shows a per-property comparison table at the bottom. - **Detailed data tooltips**. Every data widget on the Dashboard and Traffic pages now has a rich info tooltip explaining what the data means and how each column is calculated. - **AI Assistant**. Ask questions about your visibility data in natural language. The assistant looks up your scores, trends, competitors, citations, and content briefs in real time. Available on Growth and Scale plans (1 credit per 10 messages). - **Shareable conversations**. Share any AI Assistant conversation as a read-only link. Recipients can view the full thread without logging in. - **MCP connection guide**. The assistant page now promotes connecting BrandCrux to your AI tools via the MCP server for direct data access. - **Inline visualizations in AI Assistant**. The assistant now renders score badges, bar charts, line charts, pie charts, and data tables directly in chat responses. - **Dark and light theme toggle on website**. All public pages now support dark and light themes with a toggle in the navigation bar. Defaults to your system preference. - **Mobile responsive overhaul**. Every app page now adapts to mobile screens. Compact topbar, collapsible grids, scrollable tables, and touch-friendly controls across all 22 pages. - **GEO Audit default view**. The GEO Audit page now opens with your most recent audit results displayed by default, with audit history shown below. - **AI Assistant feature page**. New dedicated landing page showcasing example queries, visualization types, MCP integration, and credit pricing. ## April 28, 2026 - **AI traffic tracking**. See exactly how many visitors ChatGPT, Perplexity, Claude, Gemini, and other AI engines send to your website. Per-engine breakdown, per-page heatmap, and week-over-week trends. - **Analytics and Search Console connectors**. Connect GA4 and Search Console with a property picker. 90-day historical backfill on first connect, daily auto-sync, and email notification when data is ready. - **AI visibility vs traffic correlation**. Side-by-side view of your AI visibility score and actual referral traffic per engine. See which engines recommend you AND send visitors. - **Content advisory enrichment**. Recommendations now factor in AI traffic patterns. "This page gets ChatGPT traffic but has high bounce rate" type insights. - **Cross-referencing insights engine**. Automated insights comparing AI visibility with organic traffic. Detects opportunities, validates strategies, and flags drops. ## April 27, 2026 - **Design system overhaul**. Completely new UI built on oklch color tokens. Refined sidebar with active indicator bar, breadcrumbs in topbar, density modes, and consistent component styles across all 22 pages. - **Executive Report page**. New CXO-ready report with revenue opportunity calculator, competitive threat matrix, topic authority analysis, ranking breakdown, and 3 Quick Wins. - **Shared report enhancements**. Added topic authority, ranking breakdown per source, AI traffic section with blurred mockup call to action, and "more recommendations" prompt. Mobile-friendly layout with logical section ordering. - **Scans page rewrite**. Scan history view with expandable detail, cancel, re-run, and delete actions, animated progress for running scans, and auto-reload on navigation. - **Prompts heatmap**. Score history shown as a color-coded 2D grid (sources by dates) with inline sparklines per engine. Replaces the old flat chip list. - **Scan queued modal**. Beautiful overlay modal when a scan is queued, with "View scan status" action. - **Business context sync**. Selected business persists across page navigation. Switch transition overlay shows "Switching to [Business]" during business change. - **Entity deduplication**. Variant spellings of the same brand now merge into one entity with combined mentions. - **Smooth trend lines**. All trend charts now use catmull-rom bezier curves instead of straight polylines. Hover tooltips on data points show date, engine, and value. ## April 26, 2026 - **Date range selector**. Dashboard now has a preset dropdown (7d, 14d, 28d, 30d, 90d, this month, last month, quarter to date) plus custom date inputs. Selection persists across pages. - **PDF export**. Export any report page as PDF via print. Added to dashboard, prompts, trends, competitors, entities, citations, scans, and executive report. - **Toast notifications**. Real-time toast popups in bottom-right corner for new alerts. Polls every 30 seconds, auto-dismiss after 8 seconds, click to navigate. - **Delete business and account**. Delete a business (with all data cascade) or your entire account from Settings. Name-confirmation required for safety. - **Team invitation improvements**. Resend pending invitations, non-BrandCrux users see Sign In and Sign Up on the invite page, auth redirect preserves invite token. - **API documentation portal**. Comprehensive docs with sidebar navigation, 5 use cases with code examples, "Copy for LLM" button, per-endpoint copy, rate limits, and error codes reference. - **Connect call-to-action prompts**. Pages that benefit from Analytics and Search Console data show dismissible banners prompting users to connect. ## April 25, 2026 - **Team collaboration**. Invite team members with roles (Owner, Admin, Member, Viewer). Share businesses across your team. - **Onboarding email sequence**. New users receive a 4-step guide: welcome, score explanation, competitor insights, and usage tips. - **API visibility blog**. New article: "Why Tracking Your Visibility in LLM APIs Is More Important Than Web Chat". ## April 25, 2026 - **New models tracked continuously**. BrandCrux now continuously adds new language models from 10 major providers. Each model includes pricing, capability tags, and a "why track this" explanation. - **Provider pages**. Dedicated pages for each AI provider showing all tracked models and integrations. - **Integrations page**. Browse all 45+ integrations across AI sources, analytics, content publishing, notifications, and developer tools. - **Why API Visibility section**. New landing page section explaining why API-channel tracking matters for the agent economy. ## April 24, 2026 - **BrandCrux rebrand**. Formerly Prism. New name, new domain (brandcrux.ai), new icon. - **Scan completion emails**. Get a summary email after every scan with per-engine breakdown and comparison to previous scan. - **GEO audit email notifications**. Receive detailed audit results by email with category scores and page-level issues. - **Scan frequency control**. Choose daily or weekly automated scans per business. - **Light mode overhaul**. Improved contrast, depth, and visual clarity for light theme users. ## April 24, 2026 - **AI scorer improvements**. More accurate implicit mention detection with confidence thresholds, reducing false positives. - **Free scan query selection**. Choose which prompts to scan instead of auto-selected queries. Add your own custom prompts. - **Device-based rate limiting**. Fair usage controls for free scans and signups. - **Browser error logging**. Development tool for faster bug detection. ## Earlier - **18 AI source support**. 9 Web GUI plus 9 API engines tracked simultaneously. - **Competitor tracking and share of voice**. See which brands AI recommends instead of you. - **Citation and entity analysis**. Track which domains are cited and which entities are mentioned. - **GEO Audit for site optimization**. Crawl your site and score every page for AI citation readiness. - **Content Engine**. AI-powered briefs, writing assistance, and publishing workflows. - **Analytics and Search Console integration**. Connect GA4 and Search Console for holistic visibility data. - **WordPress publishing**. Publish optimized content directly to WordPress from BrandCrux. - **REST API and MCP Server**. Programmatic access to all scan data and platform features. - **Weekly digest emails**. Automated weekly summary of visibility changes and competitor movements. - **Credit-based pricing with auto-recharge**. Flexible pay-as-you-go credits that never expire, with optional auto-recharge. --- BrandCrux is the brand authority tracker across channels. The crux of your brand across channels. https://brandcrux.ai