Changelog

Reports now collapse obvious competing-brand variants like domain-shaped names without damaging real multi-word brand names.

Reports now avoid telling you to earn mentions on a rival's own site when your website already shows that company as a competing brand.

Project-ready and reminder emails now use a simpler, more readable format with clearer calls to action.

Reports now separate confusing domains from the recommended next step, making brand-confusion warnings easier to read.

Free audit projects now have a clear 30-day retention window unless the account upgrades, keeping trial data temporary by default.

If AI mentions your brand but mixes it with similarly named domains, the report now treats that as brand confusion instead of making direct recognition look healthier than it is.

Project setup now shows the market category and focused buying angle before the first snapshot, so you can catch a bad category choice before it turns into a bad AI visibility report.

Competing-brand checks now stay closer to the buyer's real market, so reports are less likely to mistake ordinary words for actual competitors.

New projects now pause for a review step before monitoring starts. Confirm the brand, buyer, owned domains, and competing brands so bad autofill data does not turn into a bad AI visibility report.

Google AI Overviews results are now more reliable inside ongoing reports, giving teams a clearer read on how Google answers buyer questions.

Auto fill now turns messy website copy into cleaner buyer language before the first snapshot runs: sharper audience, clearer category, better problem framing, and fewer weird competitor suggestions.

Reports now look beyond a single rival, giving a broader view of which competing brands are taking visibility from you.

When AI answers confuse your brand with a similarly named company, domain, or generic category, the report now calls that out instead of counting it as a win.

The project history view now leads with up-is-good movement: visibility, coverage, buyer visibility, and action progress. Negative signals still appear, but they no longer make an improving project look like it is falling apart.

Daily monitoring now gives projects a fresher sense of movement without overwhelming the report with noisy changes.

Paste any URL at /llms-txt-generator and get a complete, ready-to-deploy llms.txt written from the site's real crawled pages — actual URLs and descriptions, not a fill-in template. Free, no signup.

Plain-language guides now explain what it means to be visible across ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews, and Grok.

Side-by-side comparisons against PromptMonitor, Otterly, Rankscale, Peec, Profound, and Scrunch — verified pricing, what they genuinely do better, and where we win. Find them at /vs/<name>.

Enter any URL at /check and see whether AI assistants mention the brand when buyers ask. Fast, free, and no signup required.

One click creates a public, read-only link to your latest report — diagnosis, scores, sources, and the actual AI answers. Revoke it anytime.

Every score now opens to the raw stored answer behind it, per question per platform, with honest mention tags. No black box.

Brand checks and reports now return faster while preserving the evidence behind each answer.

Reports now avoid giving credit for weak, confused, or non-committal AI mentions. If an assistant does not really recognize the brand, the score says so.

For llms.txt, robots.txt, schema, sitemap, and metadata findings, the workshop now generates the complete, ready-to-deploy file from your real site data. Download, upload, and the next daily site check verifies it.

Archive a project to free its plan slot without losing history. Re-creating it later picks up right where it left off.

Starter, Pro, and Scale are live with localized flat pricing and self-serve checkout. Every new account starts with a full free audit — no card, no sales call.

Scale plans now check Grok with live web search — the sixth platform, alongside ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews.

Your buyer questions now run as real Google queries too. When Google shows no AI Overview for a question, that's recorded as a finding — not an error.

Projects now update continuously instead of feeling like a one-time audit, making AI visibility easier to track as a living signal.

SearchByAI generates buyer questions for each project, and paid plans can add their own questions with the same evidence-backed reporting.

Recommended actions now open into practical workshops with steps, examples, and a clear checklist for what to improve.

Dark console interface across the whole product: score gauges, trend lines, source donuts, and verdict-first reports in plain language.

Reports started turning missed mentions, weak sources, and competitor pressure into a prioritized fix plan instead of a pile of generic SEO advice.

Monitoring became easier to follow day by day, with cleaner history and fewer stale report moments.

Setup began capturing the brand, buyer, problem, market, owned domains, and competing brands before the first AI visibility snapshot runs.

Paste a homepage and SearchByAI drafts the brand name, buyer, offer, problem, category, and competing brands from the real page instead of making you start with a blank form.

Projects can now separate official brand domains from confusing lookalikes, making reports clearer when similarly named websites appear.

The report now checks whether the site gives AI systems the pages they expect: product, pricing, alternatives, use cases, docs, FAQs, and proof.

Snapshots started stacking into a history view, making it possible to see visibility, coverage, competitor pressure, and site readiness change over time.

Reports began showing which competing brands appear when yours does not, turning invisible AI demand into something you can actually fight.

SearchByAI started generating the core question set every project needs: brand recognition, category search, use-case search, problem search, and alternatives.

Robots.txt, sitemap, llms.txt, metadata, schema, canonical URL, crawlable buyer pages, and answer coverage became part of every visibility report.

AI citations started getting classified into owned domains, competitor domains, third-party authorities, and brand-confusion domains.

Every score was tied back to stored AI answers and cited sources, so the report could show receipts instead of asking you to trust a black-box grade.

The first version asked buyer-style questions across major AI assistants and measured whether a brand was mentioned, cited, recommended, ignored, or confused.