The proof you do not write yourself
Reviews are the only proof an engine did not have to take your word for
Everything else on your site is you describing yourself. A review is somebody else doing it, on a domain you do not control, which is exactly why engines lean on it. Read the honest part first: getting reviews is your job, not ours. What we do is tell you which platform matters for your industry, which sources the engines actually cited for your questions, and where a competitor is being cited and you are not.
Every plan queries all 6. See how each engine picks who it names
What we measured
firms listed on Clutch across the 15 regions of our managed IT study were named by no AI engine at all. Being in the directory is not the same as being cited from it, and one profession is the whole scope of that number.
Engines queried
ChatGPT, Claude, Perplexity, Gemini, Google AI Mode, Microsoft Copilot. Every plan, including the free one.
Answers kept
Every answer is stored word for word, so any number on your report can be checked against the text it came from.
The gap this closes
- You are on the directories already and it has changed nothing. In our managed IT study 139 of 228 Clutch-listed firms were named by no engine.
- The platform that matters is not the same for everyone, and most advice picks one and calls it universal.
- Engines disagree wildly about who to trust: on the same study Perplexity corroborated 39% of the firms it named and Claude 2%.
- A five-star average nobody quotes is worth less than three reviews that say what you actually do.
What you end up with
- The review and directory sources the engines cited in your category, from your own scan.
- Which of those were cited in answers that left you out.
- The right platform for your industry, rather than the one an article recommended.
- Plain guidance on getting there. We do not write, solicit or place reviews.
How we work it
Three steps, and you only do the first one.
The scan records who was cited
Every scan, including the free one, stores the sources cited across the answers, and separately the ones cited in answers where you were not named. That second list is the target list. Two honest limits: the sources are pooled across engines rather than attributed to one, and the list tells you a rival was named, not which rival was cited where.
You get the platform, not a lecture
Which platform matters is an industry question. Software buyers are on G2 and Capterra, services buyers on Clutch, local buyers on Google Business Profile, consumer-facing brands on Trustpilot. The scan tells you which ones the engines actually cited for your category.
You get the reviews yourself
This is the part we will not dress up. Asking your customers is your job. We can tell you which platform, which gap, and in what order, and the next scan tells you whether it moved an answer. We do not write, solicit or place reviews, and anyone offering to should worry you.
What you get on each plan
Every plan surfaces the citation gaps, including the free one. What the paid plans add is the loop that tells you whether closing one changed an answer.
Track
Free- The review and directory sources the engines cited
- Which of them were cited in answers that skipped you
- Ranked under authority in the Fixes list
- No card, and the gaps are the hard half to find
Fix
$99/mo- Weekly scans turn the list into a loop, not a snapshot
- A listing you claim shows up in the next crawl
- Verified rather than ticked off
- Alerts when a rival appears in a source that ignores you
Dominate
$499/mo- The wider question budget widens the citation evidence
- More questions means more answers that left you out
- Sequenced, so the source that decides your shortlist comes first
- A person reads the list with you
Rule
$1999/mo- Portfolio level, where the same directories serve several brands
- Sequencing becomes which brand gets the placement
- Human strategy on top of the same production
Which engine actually leans on this
Your row is Reviews and directories. The rest of the grid is on the engines page.
| Channel | ChatGPT | Claude | Perplexity | Gemini | Google AI Mode | Microsoft Copilot |
|---|---|---|---|---|---|---|
| Your own site Schema, answer pages, llms.txt, crawlability | Heavy Vendor | Heavy Vendor | Heavy Measured | Heavy Our read | Heavy Vendor | Heavy Our read |
| Reviews and directories G2, Clutch, Capterra, Trustpilot, Google Business | Moderate Our read | Light Measured | Heavy Our read | Moderate Our read | Moderate Measured | Moderate Measured |
| Reddit and forums Unpaid opinion, where buyers go for the unvarnished view | Light Our read | Light Measured | Heavy Our read | Moderate Our read | Heavy Measured | Moderate Measured |
| Video YouTube. Only counts once there is a transcript | Moderate Our read | Light Our read | Moderate Our read | Heavy Our read | Heavy Measured | Light Measured |
| Newsletters LinkedIn, Substack, Beehiiv. Published under a named person | Moderate Our read | Light Our read | Moderate Our read | Heavy Our read | Heavy Measured | Light Our read |
| News and PR Earned coverage, wire pickup, journalist mentions | Heavy Vendor | Heavy Vendor | Heavy Our read | Moderate Our read | Moderate Our read | Moderate Our read |
Your own site. Your own site counts most for ChatGPT, Claude, Perplexity, Gemini, Google AI Mode, Microsoft Copilot.
Reviews and directories. Reviews and directories counts most for Perplexity. It is a light signal for Claude.
Reddit and forums. Reddit and forums counts most for Perplexity, Google AI Mode. It is a light signal for ChatGPT, Claude.
Video. Video counts most for Gemini, Google AI Mode. It is a light signal for Claude, Microsoft Copilot.
Newsletters. Newsletters counts most for Gemini, Google AI Mode. It is a light signal for Claude, Microsoft Copilot.
News and PR. News and PR counts most for ChatGPT, Claude, Perplexity.
Where this comes from
- OpenAI: publishers and developers FAQ States that a site must allow OAI-SearchBot to be eligible for ChatGPT search at all.
- OpenAI: introducing ChatGPT search The named publisher partnerships behind the news and PR row.
- Anthropic: web search tool documentation Citations are always enabled, and Claude only searches when a question needs current information.
- Google Search Central: AI features and your website AI Overviews and AI Mode report inside the ordinary Search Console Performance report.
- Search Console Help: impressions, position and clicks How AI Mode counts a click, and why a follow-up question is treated as a new query.
- Dan Petrovic: citation mining of OpenAI grounding metadata Reddit retrieved in 76 percent of OpenAI searches and selected 0.61 percent of the time.
- Our production scan database, read 12 August 2026 46 completed reports, 34 brands, 419 buying questions, 2,421 answered buying cells. Where the per-engine mention rates come from.
- Our 15-region managed services study 743 records, 566 named by an engine, and the per-engine corroboration rates.
- What our crawler finds on real sites 39 readable crawls: 38 with no llms.txt, 26 with no FAQ schema, 16 with no Organization schema.
- How we check an AI visibility report The method behind every first-party number on this page, including what it cannot show.
Engine behaviour moves month to month. This table is maintained, not published once.
The number that should change how you read your own directory listings
Across 15 regions of our managed IT study, four engines named 566 companies between them. An independent check could corroborate 115 of them.
Of the 228 firms listed on Clutch in those same regions, 139 were named by no engine at all. Listed and cited are different states, and the gap between them is the work.
Read the full managed services study →Which platform, for whom, and why an engine reads it
The general shape of a question that gets answered badly everywhere else. Read the row for your business, not the column, and treat this as the illustrative case rather than yours: your own scan is what settles which of these the engines actually cited for your category.
Platform by buyer, and where it lands
| Platform | Whose buyers are there | Why an engine reads it |
|---|---|---|
| G2 | Software and SaaS | Deep category taxonomy and structured comparison pages |
| Capterra | Software, SMB-weighted | Category listings that rank for buying phrasings |
| Clutch | Agencies, MSPs, professional services | Verified project detail, which is rare corroboration |
| Google Business Profile | Local and near-me buyers | Google's own surface, so Gemini and AI Mode see it natively |
| Trustpilot | Consumer-facing and D2C | Volume and recency, weaker on B2B category context |
What the scan tells you that this table cannot
- Which of these the engines actually cited for your category, rather than which should matter in theory
- Which cited a competitor in an answer where you were absent, which is the ranked target list
- Whether the listing you claimed last month has been picked up by the next crawl
The table is the general case. Your scan is the specific one, and where they disagree, trust the scan.
How it actually works
The mechanism, for anyone who wants it. Open a card to read the detail.
Why an engine trusts a review page more than your page
Every page on your own domain is you describing yourself, and an engine knows it. A review sits on a domain you do not control, written by someone with no stake in your marketing, which makes it the cheapest corroboration available to a model deciding whether to name you.
Read the detail
That is also why the effect is not linear. One more five-star rating on an average nobody reads changes nothing. A review that names the problem you solved, in the words a buyer would use, is a sentence an engine can lift into an answer.
The practical consequence is that review quality beats review volume for this purpose, which is the opposite of what most review-generation advice optimises for.
Recency compounds the same effect. A retrieval-first engine reaching for current pages will surface a profile updated last month over one that stopped in 2024, regardless of which has the higher average. That is why a dormant five-star profile quietly stops working and nobody notices, and it is also why this channel needs a cadence rather than a launch.
The platform depends on your industry, and nobody else will say so
Most advice on this topic picks one platform and generalises. That is what makes it useless: the platform that carries a software buyer is not the one that carries a plumbing contractor.
Read the detail
Software and SaaS buyers are on G2 and Capterra. Buyers of agencies, MSPs and professional services are on Clutch. A local services buyer is on Google Business Profile, which is Google's own surface and therefore structurally different from the rest. Consumer-facing brands are on Trustpilot.
You do not have to guess which applies to you, because the scan records which of these the engines actually cited when answering your category's buying questions. That is the whole argument for measuring before acting.
What we will not do, stated plainly
We do not write reviews, solicit them on your behalf, or place them. That is your work with your customers, and on the self-serve plans what we provide is guidance and measurement.
Read the detail
Fabricated reviews are off the table entirely and always will be. They violate every platform's terms, they are illegal under FTC rules in the United States, and they get removed, which means the money buys you a penalty and a gap.
This matters more in this channel than anywhere else we work, because the whole reason an engine trusts a review is that you did not write it. A vendor offering to manufacture your corroboration is offering to destroy the thing that makes it valuable.
What it costs
The citation gaps are on every plan, including the free one. Paid plans add the loop that verifies a fix.
| Plan | Price | Questions | Scans |
|---|---|---|---|
| Track | Free | 5 | 1/mo |
| Fix | $99/mo | 10 | 4/mo |
| Dominate | $499/mo | 25 | 4/mo |
No plan includes review generation, because we do not do it. What scales with the plan is how many questions get asked, how often, and whether a person reads the list with you.
See where you stand first
Run a free scan and find out which of these gaps you actually have.
Get your free auditFrequently asked questions
Can TofuBofu get reviews for my company?
No, and you should be wary of anyone who says they can. Getting reviews is your work with your customers. What we do is measurement and guidance: which review and directory sources the engines cited for your category, which of them cited a competitor while leaving you out, and which platform matters for your industry. On the self-serve plans that is guidance, not a done-for-you service. Fabricated reviews are off the table entirely, because they breach platform terms, fall foul of FTC rules in the United States, and get removed.
Which review platform matters most for AI visibility?
It depends on your industry, which is why most answers to this question are wrong. Software and SaaS buyers are on G2 and Capterra. Agencies, MSPs and professional services buyers are on Clutch. Local buyers are on Google Business Profile, which is Google's own surface and therefore read natively by Gemini and AI Mode. Consumer-facing brands are on Trustpilot. Rather than guessing, your scan records which of these the engines actually cited when answering the buying questions in your category.
Does being listed on Clutch or G2 mean AI will recommend me?
No, and this is the single most useful thing on this page. In our 15-region study of managed IT providers, 139 of the 228 firms listed on Clutch were named by no engine at all. The listing existed and the citation did not. A directory profile makes you eligible to be cited; it does not make you cited. What closes the gap is corroboration an engine can actually read: specific, recent reviews that describe the work in the language a buyer uses.
How many reviews do I need?
The honest answer is that volume is the wrong target. Kevin Indig's analysis of G2 data associates a 10 percent increase in reviews with roughly 2 percent more AI citations, which is real but modest, and it is an association rather than a mechanism. A review that names the problem you solved in a buyer's own words is a sentence an engine can lift into an answer. Fifty ratings with no text are not. Optimise for reviews that say something specific.
Do all six engines weigh reviews the same way?
Not remotely, and the spread is larger than most people expect. In our managed IT study Perplexity corroborated 39 percent of the firms it named against an independent check, and Claude corroborated 2 percent. Perplexity and ChatGPT lean on review and directory sources heavily; Claude answers most stable vendor questions without searching at all, so no amount of review work reaches it quickly. Google Business Profile is a special case, because it is Google's own surface and Gemini and AI Mode read it natively.
What about Google Business Profile for a local firm?
It is the highest-leverage review surface for anything with a service area, and it is structurally different from the rest. Google owns it, so the reviews are not a third-party source Google has to go and find, they are already inside Google's own data. If your buyers search with a place name attached, that profile is doing more work for you on Gemini and Google AI Mode than any software review site will.
How do I know whether a new review actually changed anything?
Re-scan and compare. That is the only honest test, and it is why the measurement matters more than the advice. The scan stores every answer word for word, so when a question you were losing starts naming you, you can read the answer that changed and see which source it leaned on. On the free plan you can re-run the same questions monthly; the paid plans do it weekly and alert you when an answer moves.
Sources and further reading
- Our 15-region managed services study: Where the 139-of-228 Clutch figure and the per-engine corroboration rates come from. One profession, four engines, July 2026.
- Dan Petrovic, citation mining of OpenAI grounding metadata: Measured which candidate sources OpenAI retrieves and which it actually cites, across 27,351 probes.
- Kevin Indig, G2 review analysis: G2 at roughly 22.4 percent share of voice for software, and 10 percent more reviews associating with about 2 percent more AI citations.
- G2 2026 AI Search Insight Report: 51 percent of B2B software buyers begin vendor research on an AI chatbot more often than Google, and 69 percent chose a different vendor than planned based on an AI answer.
- Which engine leans on which channel: Our channel by engine cross-tab, including the reviews row and where each figure came from.
Related solutions