Straight answers about AI search visibility

SEO gets you ranked in a list; AI search names three or four sources and ignores the rest. Ranking #1 on Google does not make ChatGPT mention you. Each engine selects sources with its own logic, and that logic rewards different work: extractable answers, citable evidence, third-party mentions. SEO is still necessary. It is no longer sufficient.

We never suggest dropping SEO. Google AI Overviews build from Googlebot's normal index, so classic crawlability still matters. The work we add — answer-first structure, evidence density, earned mentions — sits on top of a sound SEO base, not instead of it.

Because ranking and being cited are different mechanisms. Google orders pages; ChatGPT composes an answer and names a handful of sources chosen by its own logic — extractable passages, corroborating mentions across the web, machine-readable facts. Your page can top the list and still offer nothing an engine can lift. That gap is fixable, and it's the exact gap an audit measures.

Use one if you have someone to act on what it shows. Profound, Peec and Otterly are good dashboards: they tell you where you're missing from AI answers. They don't rewrite your pages, add schema, or earn the third-party citations that move the numbers. We run the same kind of tracking, then do the work it points to.

Real prices, so you can compare: Profound starts around $82.50/mo, Peec at €89/mo, Otterly at $29/mo. If you have in-house capacity to execute, a tool plus your team is a legitimate setup — we'll say so on the call if it fits you better.

The acronyms are new; the behavior isn't going back. Generative AI platforms drew 9.5 billion visits a month this past year, up 70% year over year (Similarweb), and 68% of Google searches already end without a click (SparkToro). And the underlying work — clear answers, structured facts, third-party evidence — is durable content quality that keeps paying even in classic search.

It matters most if your buyers research before they buy. Generative AI platforms drew 9.5 billion visits a month this past year (Similarweb), and AI answers name only a few businesses per question. AI referrals are still a small share of most sites' traffic — the case for acting now is who these visitors are and how fast the behavior is growing.

Start where your buyers are and where movement comes fastest: Google AI Overviews (reaching ~2 billion monthly users (DemandSage) and built from Google's normal index) and ChatGPT search. Perplexity rewards the same on-site work quickly. Then broaden — engines agree on almost nothing, so coverage across six engines is what monitoring is for.

Our prices are published: the fixed-scope audit runs $1,500–3,000, implementation projects run $5,000–15,000 depending on the roadmap, and monthly monitoring runs $2,000–10,000 depending on prompt-set size and engine coverage. For context, mid-market agency retainers typically land between $2,000 and $10,000 a month, usually unpublished. You'll always see the number before any work starts.

The market context comes from published pricing research: Digital Elevator's 2026 AEO/GEO pricing guide puts credible mid-market retainers at $2,000–10,000/mo, with premium agencies at $8,000–20,000/mo and enterprise programs higher.

First movement typically shows in four to eight weeks on retrieval-based surfaces like Perplexity, ChatGPT search and Google AI Overviews, because on-site answer fixes register on the next crawl. Authority-driven visibility, where engines recommend you from synthesis, builds over three to six months. We report both monthly, so you watch the trajectory rather than wait on a promise.

Usually both, weighted toward what you have. Most sites already own the right pages; they just don't open with extractable answers or carry citable evidence, so restructuring beats rewriting. New content enters where the prompt set exposes a question you answer nowhere. The audit tells you the split before you commit to either.

Yes, as one lever among several. Structured data states your facts — services, prices, locations, authorship — in a machine-readable form, which supports entity clarity and extraction. It is not a magic switch: schema makes good content easier to lift; it does not rescue pages that never answer the buyer's question. We ship it as standard implementation work, not as the strategy.

llms.txt is a proposed root file that points AI systems to your best content. Honestly: about 97% of llms.txt files are never fetched by AI crawlers (OrganiKPI), and Google has confirmed Search ignores the file. We ship it anyway because it's cheap and useful for coding agents and agentic browsers — but we will never sell it as a visibility lever.

Separate the jobs. Search and user-fetch bots — OAI-SearchBot, PerplexityBot, Claude-SearchBot — control whether you can be cited in AI answers; blocking them kills visibility. Training bots like GPTBot and ClaudeBot only affect model training, and allowing or blocking them is your policy call. We configure the split deliberately in every engagement rather than blanket-blocking everything.

Only if its crawlers can reach you. ChatGPT search uses OAI-SearchBot; Perplexity uses PerplexityBot; Google's AI Overviews build from Googlebot's normal index. If your robots.txt blocks the search bots — a common leftover from blanket AI blocking — you are invisible regardless of content quality. Checking crawler access is the first item in our audit and in the free checklist.

Two mechanisms. When it searches the live web, it retrieves and extracts passages that directly answer the question — clear, quotable, well-structured pages win. When it answers from what it already knows, it synthesizes from sources the wider web treats as authoritative — reviews, references, communities, coverage. Being recommendable means working both: extractable pages and a real third-party footprint.

Yes. Analyses of large citation samples put Reddit, Wikipedia-class references, YouTube and LinkedIn among the most-cited domains across engines, alongside vertical review sites like G2, Avvo and Healthgrades. That's why off-site work matters: engines synthesize from where the web already talks. We earn presence in those sources; we never fake it — engines and readers both punish that.

We build a fixed set of buyer-intent prompts for your market, run it on a schedule across ChatGPT, Gemini, Perplexity, Claude, Google AI Overviews and Meta AI, and log two numbers per engine: mention rate, how often you're named, and citation share, your share of cited sources versus competitors. You see both trend monthly.

Both metrics are defined in plain language in the glossary, and the monthly report format is shown — with clearly labeled demonstration data — on the how it works page.

A prompt set is the fixed basket of questions we track your visibility against — the questions your buyers actually ask an AI, like 'best [your service] in [your city]' or 'is [your product] worth it.' We build it from your sales conversations, search data and market language, agree it with you, then freeze it so month-over-month numbers stay comparable.

Mention rate is how often your business is named in AI answers across your tracked prompt set. Citation share is your slice of the cited sources versus competitors. There is no universal good number — baselines vary wildly by market — so we benchmark you against named competitors on your own prompts and report the trend monthly.

One page of numbers you can act on: mention rate and citation share per engine, month over month; the competitor comparison on the same prompts; which prompts moved, which didn't, and why we think so; and next month's prioritized fixes. A sample report with clearly labeled demonstration data is on the how-it-works page.

See the sample on how it works — the figures there are demonstration data and stamped as such, but the shape is exactly what clients receive.

The evidence says yes: ChatGPT ecommerce traffic converts 31% higher than non-branded organic (Visibility Labs via Search Engine Land), and Semrush estimates LLM referral sessions are worth about 4.4x an organic session. The honest caveat: AI referrals are still a small share of most sites' traffic. The case is who these visitors are and how fast this is growing.

ChatGPT passed 900 million weekly active users (DemandSage, reporting OpenAI figures), Google's AI Overviews reach roughly 2 billion monthly users (DemandSage, from Google I/O disclosures), and generative AI platforms overall drew 9.5 billion visits a month, up 70% year over year (Similarweb). Usage is mainstream; the open question is which businesses the answers name.

No, and you should walk away from anyone who says yes. Engines change source selection without notice, and only 11% of domains cited by ChatGPT are also cited by Perplexity — nobody controls that. What we guarantee: a defined prompt set, measured monthly, the scoped work delivered, and honest reporting of what moved and what didn't.

The 11% overlap figure is from AuthorityTech's 2026 per-engine citation audit. It's also why we track six engines separately rather than selling a single blended score.

Often, yes. Engines repeat what the web tells them, so wrong facts usually trace to stale or conflicting information on directories, review sites and your own pages. The fix is entity consistency: the same correct facts everywhere engines look, structured data that states them machine-readably, and time for recrawls. No one can force a correction, but consistency reliably shifts what gets repeated.

Possibly, and you wouldn't know: there is no alert when an AI answer names your competitors and skips you. The only way to find out is to ask the engines the questions your buyers ask and log the results. That is exactly what an audit does — your prompt set, run across six engines, with every named business recorded.

Engines default to what the web documents best, and national brands have deeper footprints: more reviews, more mentions, more consistent data. Local businesses win by being specific — pages that answer 'near me' questions directly, a clean Google Business Profile, consistent name-address-phone data, and real reviews. Specific and local beats big and generic on local-intent prompts.

The work is content and accuracy work, so it lives comfortably inside advertising rules: truthful practice-area pages, verifiable credentials, real reviews on platforms like Avvo, consistent facts everywhere. Nothing involves fake reviews, misleading claims or guarantees of outcomes — which the rules prohibit and we don't do anyway. Your existing marketing-review process applies unchanged; we work within it.

Yes, because nothing we do touches patient data. The work is public-facing: service pages that answer patient questions accurately, provider credentials stated consistently, structured data, and presence on the review sources engines read, like Healthgrades. No patient information is used in prompts, content or reporting. Health content is held to a stricter accuracy bar — which is the point.

Yes, with your compliance process in the loop. Everything we produce is public marketing content, so it goes through your existing review and approval workflow before publication — we plan for that lead time. The substance is accuracy work: correct product facts stated consistently everywhere engines look, which typically reduces compliance risk from AI answers rather than adding to it.