Skip to main content Scroll Top

AI Visibility Tools: What Each One Actually Costs

Updated September 2026 · Written and maintained by the Progression Agency strategy team

Eight tools, checked against their own pricing pages on 15 September 2026 — because the comparison articles quote prices the vendors do not publish. Five publish full pricing. Three publish none at all.

On this page · 18 sections
  1. Which AI visibility tools actually publish their prices?
  2. What do these tools actually measure?
  3. Which answer engines does each tool cover?
  4. How much should you actually pay?
  5. What do the free options actually give you?
  6. Which tool fits which kind of business?
  7. How do you set one of these up properly?
  8. What are the common mistakes when buying one of these?
  9. Do you need a tool, or do you need the problem fixed?
  10. What should you ask a vendor before buying?
  11. How do these tools differ from ordinary rank trackers?
  12. How we checked these prices, and what we did not check
  13. Does buying a tool improve your AI visibility?
  14. How do the tools differ on the things that are hard to compare?
  15. What does the pricing tell you about where this market is going?
  16. What happens to these tools if the AI surfaces change?
  17. Vendor by vendor, what each one is actually for
  18. More from Progression Agency

The short answerAI visibility tools sample AI assistants with a fixed prompt set and record whether you are mentioned, whether you are linked, and who appears alongside you. Of the eight we checked, Otterly, Scrunch, Rankscale, Ahrefs Brand Radar and Semrush publish full prices; Profound publishes a trial and custom enterprise; Peec AI and SE Ranking publish none. Entry tiers start around $20-29 a month. None of them can tell you why you are absent — that comes from server logs and a rendering test.

Prices read from each vendor’s own pricing page on 15 September 2026 and linked below. This category changes pricing frequently; check the date before relying on any figure. Progression Agency has no affiliate or reseller relationship with any tool listed.

Entry price, where the vendor publishes oneEntry price, where the vendor publishes one
Profound, Peec AI and SE Ranking publish no monthly price for these products at all, so they cannot appear on this chart.

Which AI visibility tools actually publish their prices?

Five of the eight we checked publish full pricing: Otterly, Scrunch, Rankscale, Ahrefs Brand Radar and Semrush. Profound publishes a free trial and custom Enterprise only. Peec AI names four tiers with no numbers. SE Ranking’s product page carries no price at all. Every figure below was read from the vendor’s own pricing page on 15 September 2026.

That distinction is the first useful filter, and it is missing from almost every comparison article on this subject. A tool with custom-only pricing is not necessarily expensive — it is simply not comparable until you have had a sales conversation, which changes how you shortlist when you are evaluating six products in an afternoon.

We checked because the secondary sources disagreed with each other. Two widely-cited comparison posts quoted different prices for the same Peec AI plans, and the vendor publishes neither figure. A third quoted specific Profound tiers that do not appear on Profound’s pricing page. A fourth stated an Ahrefs minimum roughly sixteen times the published entry price. None of those articles said when they had checked.

So this page does three things differently: every price is read from the vendor, every price carries the date it was read, and where a vendor publishes nothing we say so rather than repeating a number from somewhere else.

Otterly — Publishes prices. $29 / $189 / $489, Enterprise custom.
Scrunch — Publishes prices. $300 / $500 monthly, $250 / $417 annual.
Rankscale — Publishes prices. From $20; $99 / $385 / $780.
Ahrefs Brand Radar — Publishes prices. From $50; Index $199 standalone.
Semrush — Publishes prices. $99 per domain, billed annually.
Profound — Trial only. Free trial; Enterprise custom.
Published pricing, read from each vendor’s own pricing page on 15 September 2026
ToolEntryMidTop publishedCustom tierBilling note
RankscaleEssentials from $20Pro $99Enterprise $780Custom planGrowth $385; credit-based
OtterlyLite $29Standard $189Premium $489Enterprise, from about $1,00015% off annual: $25 / $160 / $422
Ahrefs Brand RadarCustom Prompts from $50AI Visibility Index $199——Custom Prompts free in paid Ahrefs plans; Index standalone
SemrushAI Visibility Toolkit $99Semrush One from $199—Enterprise, quote$99 is per domain, billed annually
ScrunchStarter $300Growth $500—Enterprise, customAnnual: $250 / $417 (17% off)
ProfoundFree trial——Enterprise, customNo standard tier price published
Peec AINot publishedNot publishedNot publishedEnterprise, contact salesStarter / Pro / Advanced named without numbers
SE RankingNot published on the product page———5 free checks advertised
What the secondary sources claimed, and what the vendor actually publishes
Claim found in comparison articlesWhat the vendor’s own page showsVerdict
Profound is $99 and $399, billed annuallyFree trial and custom Enterprise onlyNot supported
Peec AI is $95 / $245 / $495No prices publishedNot supported
Peec AI is EUR 89-199 per monthNo prices publishedNot supported, and contradicts the claim above
Ahrefs Brand Radar needs $828/month minimumCustom Prompts from $50, free in paid plans; Index $199 standaloneSubstantially overstated
Scrunch starts near $250$300 month-to-month; $250 only on annual billingPartly right, missing the condition
Scrunch was acquired in June 2026No mention anywhere on the vendor siteUnverified; not repeated here
Which vendors publish a price at allWhich vendors publish a price at all
Checked against each vendor’s own pricing page on 15 September 2026.

What do these tools actually measure?

Whether you were mentioned for a given prompt, whether you were linked as well as named, which competitors appeared alongside you, which sources the answer drew on, and how all of that changes over time against a fixed set of prompts.

The important word is measure. These are instruments, not interventions. They tell you the current reading and they tell you whether it moved. They do not tell you why it moved, and they cannot tell you what to change on the page — which is a different kind of work, done with server logs, rendering tests and editorial judgement.

This matters when budgeting. A visibility subscription and an optimisation engagement are complementary rather than alternative, and a business that buys only the first ends up with a well-measured problem it has not addressed.

Mentions — Metric. How often you are named at all.
Citations — Metric. How often you are named with a link.
Share of voice — Metric. Your mentions against competitors'.
Sources — Metric. Which domains the answer drew on.
Sentiment — Metric. How you are described, not just whether.
Position — Metric. Where in the answer you appear.
What each metric actually tells you
MetricWhat it countsWhat it does not tell you
MentionsHow often the brand is named in an answerWhether anyone could click through
CitationsHow often the brand is named with a linkWhether the link was in a useful position
Share of voiceYour mentions relative to competitors’Whether the prompt set is representative
SourcesWhich domains the answer drew onWhy those domains were selected
SentimentHow the brand is characterisedWhether the characterisation is accurate
PositionWhere in the answer you appearHow much of the answer a reader actually read
Proprietary visibility scoreA vendor-defined compositeAnything comparable outside that vendor
What a visibility tracker cannot see, and what does see it
QuestionTrackerWhat answers it instead
Did an AI crawler reach my site?No90 days of server logs
Does my page render without JavaScript?NoA fetch with JavaScript disabled
Am I blocked in robots.txt or the WAF?NoThe live file plus a bot-agent request
Does my schema contradict my page?NoA schema validator plus a read of the copy
Is my answer buried in its section?NoAn editorial audit
Am I being mentioned this week?YesThe tracker
Who else is being mentioned?YesThe tracker
What these tools can and cannot tell youWhat these tools can and cannot tell you
The bad rows are not faults. They are simply outside what a visibility tracker observes — they come from server logs and rendering tests.
ChatGPT — Engine. Largest audience; covered by every tool here.
Google AI Overviews — Engine. Index-led; covered by most.
Perplexity — Engine. Most citation-dense; covered by most.
Gemini — Engine. Covered by most.
Copilot — Engine. Bing-fed; coverage varies.
Claude, Grok, DeepSeek, Mistral — Engine. Only the broadest tools cover these.

Which answer engines does each tool cover?

Every tool here covers ChatGPT. Most add Google AI Overviews, Perplexity and Gemini. Coverage diverges at Copilot, Claude, Grok, DeepSeek and Mistral, which only the broadest platforms track.

Engine count is the specification most often quoted and least often useful. If your buyers use ChatGPT and Google, a tool covering ten engines gives you eight columns you will never read, and you will have paid for them. The question to ask is narrower: does it cover the engines your customers actually use, in the country and language they use them in?

Region and language are the specifications that genuinely get overlooked. Answers differ by both, and a tool that only samples United States English will not tell you what a buyer in another market sees.

Answer engines named publicly by each vendor
ToolEngines it namesCount
RankscaleChatGPT, Perplexity, Google AI Mode, AI Overviews, Gemini, DeepSeek, Mistral, Claude, Grok, Copilot10
Profound (Enterprise)ChatGPT, Perplexity, Google AI Mode, Gemini, Copilot, Grok, DeepSeek, Claude, AI Overviews9
Ahrefs Brand RadarAI Overviews, ChatGPT, Gemini, Perplexity, Copilot, AI Mode, plus Claude as an option7
ScrunchChatGPT, Claude, Gemini, Perplexity, Google AI Mode and AI Overviews, Meta6
SE RankingGoogle AI Overviews, AI Mode, ChatGPT, Gemini, Perplexity5
OtterlyChatGPT, Google AI Overviews, Perplexity, Microsoft Copilot4
SemrushChatGPT, Google AI, Gemini, Perplexity4
Profound (Trial)ChatGPT only1
Answer engines covered, by vendorAnswer engines covered, by vendor
Counts are of distinct engines each vendor names publicly. A larger number is not automatically better — coverage of the engines your buyers use matters more than the total.

How much should you actually pay?

For one brand and a modest prompt set, the published entry tiers run from about $20 to $99 a month. Agencies and multi-brand teams land in the $300 to $800 range. Enterprise is a conversation rather than a price, at every vendor that offers it.

The variable that moves your bill is almost never the engine count — it is the prompt allowance, or the credit consumption that stands in for it. Otterly’s tiers are explicitly priced at 15, 100 and 400 prompts. Scrunch’s are 350 and 700 custom prompts. Semrush’s toolkit includes 25 custom prompts per domain. Rankscale and Ahrefs price by credits and checks instead, which is cheaper at volume and considerably harder to forecast before you start.

Work out how many prompts you need before you look at any pricing page, because that number determines your tier at almost every vendor, and it is a number you can only get by writing the prompt set. Most single-brand businesses need fewer than they expect: sixty to two hundred prompts covering real commercial questions is a serious programme, and fifteen is enough to establish whether you have a problem at all.

Prompt and credit allowances at each published tier
Tool and tierPriceAllowanceUnit
Otterly Lite$2915search prompts
Otterly Standard$189100search prompts, +100 for $99
Otterly Premium$489400search prompts, +100 for $99
Semrush AI Visibility Toolkit$99 per domain, annual25custom prompts, daily
Scrunch Starter$300350 custom + 1,000 industryprompts, 3 users, 5 page audits
Scrunch Growth$500700 custom + 2,500 industryprompts, 5 users, 10 page audits
Rankscale Pro$991,200credits, up to 4,800 answers
Rankscale Growth$3855,500credits, up to 22,000 answers
Rankscale Enterprise$78012,000credits, up to 48,000 answers
Ahrefs AI Visibility Index$19983 per dayprompts, 2,500+ monthly checks
Profound TrialFree10prompts, one-time, ChatGPT only
Prompt allowance at the published entry tierPrompt allowance at the published entry tier
Rankscale and Ahrefs price by credits and checks rather than by prompt count, so they are not directly comparable here.
How the three pricing models behave
ModelUsed byPredictable?Cheapest whenWatch out for
Per promptOtterly, Scrunch, SemrushYesPrompt set is small and stableAdd-on prompt packs at list price
Per creditRankscale, AhrefsNoVolume is highConsumption is hard to forecast before you start
Per domainSemrushYesOne brandMulti-brand cost multiplies
Custom onlyProfound, Peec, ConductorNoEnterprise scaleA sales cycle before you can compare
How the pricing models differHow the pricing models differ
1 = weak, 3 = strong. Credit pricing is the hardest to forecast and the cheapest at scale.
Peec AI — No published price. Four tiers named, no numbers.
SE Ranking — No published price. Product page shows none; 5 free checks.
Conductor — No published price. Enterprise sales motion.
Custom-only pricing — What it costs you. A sales cycle before you can compare.
Annual-only billing — What it costs you. A year of commitment to find out.
Credit pricing — What it costs you. Hard to forecast before you start.

Want this done for your site?We build and maintain the search, content and paid programmes described on this page.

Get a free proposal

What do the free options actually give you?

Enough to find out whether you have a problem. Profound’s trial tracks ten prompts on ChatGPT once. SE Ranking advertises five free checks. Ahrefs includes Custom Prompts in its paid plans at no extra charge. And asking the assistants yourself costs nothing at all.

Start with the free version of the method before buying the automated version of it. Take the ten questions a buyer would actually ask, put them to ChatGPT, Perplexity and Google’s AI surfaces, and write down who appears. That is precisely what these tools automate — they run it more often, across more engines, and they keep the history. Those are real benefits, and they are worth paying for once you know the exercise is telling you something.

What the manual version gives you that a dashboard does not is the experience of reading whole answers. A share-of-voice number tells you that you appeared. Reading the answer tells you that you appeared as a footnote after three competitors were described in detail, which is a different situation entirely.

What you can do before paying anythingWhat you can do before paying anything
Do this first. It tells you whether you have a visibility problem worth paying to measure.
Free and low-cost ways to get a baseline
MethodCostEffortWhat it gives you
Asking the assistants manuallyNothingAn hourFull answers, and who else appears
Profound trialFreeMinutes10 prompts, ChatGPT only, one-time
SE Ranking free checksFreeMinutes5 checks
Ahrefs Custom PromptsIncluded in paid Ahrefs plansMinutesTracking of prompts you define
Rankscale EssentialsFrom $20MinutesEntry-level automated tracking
Otterly Lite$29Minutes15 prompts across 4 engines, daily

Which tool fits which kind of business?

Enterprises and agencies get the most from paid platforms, because the cost is spread across many brands and the reporting is the deliverable. A single small business usually gets most of the value from a disciplined manual prompt set and no subscription at all.

This is the recommendation most comparison articles avoid making, because a page that concludes ‘you may not need any of these’ is not commercially useful to whoever is paying for the article. But it is the honest answer for a twelve-page site with one location: the money is better spent fixing crawler access and page structure than on measuring the absence those problems cause.

Enterprise, many brands — Buyer. Paid tooling earns its cost quickly.
Agency, many clients — Buyer. Per-client dashboards are the value.
Mid-market, one brand — Buyer. Entry tier, small prompt set.
Small business — Buyer. Manual checks first; tool later if at all.
One-off diagnostic — Buyer. An audit, not a subscription.
In-house SEO team — Buyer. Export and API matter most.
Tool fit by buyer type
BuyerStart withWhyRoughly
Enterprise, multiple brandsProfound or Scrunch EnterpriseEngine breadth, SSO, API, supportCustom
Agency, many clientsRankscale or ScrunchPer-client dashboards and seat counts$385-500
Mid-market, one brandOtterly Standard or Semrush100 or 25 prompts is usually enough$99-189
Small business, one brandManual checks, then Otterly LiteDiagnose before subscribing$0-29
Already paying for AhrefsAhrefs Custom PromptsIncluded in the plan you have$0 extra
One-off diagnosticAn audit, not a subscriptionYou need the why, not the readingProject fee
Effort against value, by buyer typeEffort against value, by buyer type
For a single small brand, a disciplined manual prompt set returns much of the value at none of the cost.

How do you set one of these up properly?

Write the prompt set first, add competitors, fix region and language, take a baseline, and then leave the wording alone. The prompt set is the asset; the tool is the instrument that reads it.

Almost every disappointing tool deployment we see has the same cause: it was bought before anyone decided what to measure. The dashboard fills with numbers about prompts nobody chose carefully, the numbers move for reasons nobody can interpret, and within two quarters the subscription is renewed out of habit rather than use.

What setting one up actually involvesWhat setting one up actually involves
The prompt set is the asset. The tool is the instrument that reads it.
  1. Write sixty to two hundred prompts covering the questions buyers actually ask.
  2. Phrase them the way a buyer would, not the way your marketing does.
  3. Include comparison prompts, not only category prompts.
  4. Add every competitor you want measured against, including the unexpected ones.
  5. Set the regions and languages your customers actually search in.
  6. Take a baseline and record the date it was taken.
  7. Do not change a single prompt’s wording afterwards.
  8. Re-run on a fixed cadence — monthly is right for most businesses.
  9. Read whole answers periodically, not only the dashboard.
  10. Export the data somewhere you control, so your history is not rented.
Setup checklist, and how long each step really takes
StepTimeWho does itCommonly skipped?
Writing the prompt setHalf a dayMarketing plus salesYes — and it is the whole asset
Adding competitors20 minutesMarketingSometimes
Region and language setup10 minutesMarketingOften
Baseline and dating it10 minutesMarketingOften
Connecting exports or APIAn hourTechnicalUsually
Scheduling the monthly re-run5 minutesMarketingNo, but it lapses
Buying coverage you ignore — Mistake. Engine count is not a feature if your buyers use two.
Buying before a prompt set exists — Mistake. The tool cannot write it for you.
Changing prompts between runs — Mistake. Destroys the only comparison you had.
Treating a score as an outcome — Mistake. Proprietary indices compare to nothing outside the tool.
Reading one run as a trend — Mistake. Answers vary between runs by design.
Assuming it tells you why — Mistake. Trackers observe; they do not diagnose.

What are the common mistakes when buying one of these?

Buying engine coverage you will never read, buying before a prompt set exists, changing prompts between runs, treating a proprietary score as an outcome, and reading a single run as a trend.

The score problem deserves particular attention. Several vendors produce a composite visibility index. Inside that tool, tracked consistently, it is a perfectly reasonable relative measure. Outside it, it means nothing — it cannot be compared to another vendor’s score, to an industry benchmark, or to anything a board would recognise. Reporting it upward as though it were an external metric is how a measurement programme loses credibility the first time somebody asks what good looks like.

Mistakes, what they cost, and the fix
MistakeWhat it costsFix
Buying coverage you ignorePaying for eight unread columnsCount only the engines your buyers use
Buying before a prompt set existsA dashboard of meaningless numbersWrite the prompts first
Changing prompt wordingThe comparison you were paying forFreeze the wording
Reporting a proprietary score upwardCredibility, the first time it is questionedReport mentions and citations instead
Reading one run as a trendChasing noiseRequire three consistent measurements
Never reading whole answersMissing how you are characterisedRead a sample every month
No exportRenting your own historyCheck export and API before buying

Do you need a tool, or do you need the problem fixed?

If you already know you are invisible, a tracker will confirm it every month at a monthly cost. The diagnosis — why, and what to change — comes from server logs, rendering tests and an editorial audit, none of which a visibility tool performs.

The sequence that wastes least money is: establish manually whether you appear, fix the technical and structural reasons you do not, and then subscribe to something that measures whether the fixes worked. Subscribing first inverts that, and produces a year of well-charted absence.

A visibility tool is a thermometer. It is genuinely useful, and it has never once cured a fever.

Server logs — Diagnostic. Whether crawlers arrived at all.
JS-off render — Diagnostic. What the crawler actually reads.
robots and WAF check — Diagnostic. Whether you are blocked by accident.
Schema reconciliation — Diagnostic. Whether markup agrees with the page.
Answer-position audit — Diagnostic. Whether your answer leads its section.
Entity consistency — Diagnostic. Whether you describe yourself one way.
Tool or engagement — what each one answers
Your questionToolEngagement
Am I mentioned?YesYes
Who else is mentioned?YesYes
Did it change this month?YesYes
Why am I not mentioned?NoYes
Can crawlers reach my site?NoYes
Does my page render without JavaScript?NoYes
What should I change?NoYes
Did the change work?YesYes

Want the diagnosis rather than the dashboard?

We audit crawler access from your server logs, test every commercial template without JavaScript, restructure the pages that matter, and hand you the prompt set and baseline. You keep all of it, including the prompt set.

/contact-us

Does it publish a price? — Question. Custom-only means a sales cycle.
Which engines, by name? — Question. Not 'all major AI platforms'.
How many prompts at this tier? — Question. The real price driver.
Can I export the data? — Question. Or you are renting your own history.
How often does it run? — Question. Daily and weekly are different products.
What does the score mean? — Question. If it is proprietary, it compares to nothing.

What should you ask a vendor before buying?

Six questions, and the first one removes about a third of any shortlist: does it publish a price? After that, which engines by name, how many prompts at this tier, can I export, how often does it run, and what does the score actually mean.

Does it publish a price?

Custom-only pricing is not a red flag in itself — enterprise products are frequently priced that way for sound reasons. It is a scheduling fact: you cannot compare six products in an afternoon if three of them require a call first.

Which engines, by name?

‘All major AI platforms’ is not an answer. The list is short, public and easily stated, and a vendor who will not state it is hoping you will not check against the engines your buyers actually use.

How many prompts at this tier?

This is what you are really buying. Compare allowances rather than headline prices, because a cheaper tier with a quarter of the prompts is not cheaper.

Can I export the data?

Without export, your measurement history lives inside a subscription. Cancel it and the baseline you spent a year building goes with it.

How often does it run?

Daily and weekly are different products at similar prices. For most businesses weekly is ample; for fast-moving categories it is not.

What does the score mean?

If the answer is a proprietary composite, treat it as an internal trend line and never as an external benchmark.

Vendor questions and what a good answer sounds like
QuestionGood answerWarning sign
Do you publish pricing?Yes, here is the pageWe’ll scope it on a call, always
Which engines?A named list‘All the major AI platforms’
Prompts at this tier?A number‘Generous limits’
Can I export?Yes, CSV and APIOn the enterprise tier only
Run frequency?Daily or weekly, stated‘Regularly’
What is the score?Here is the formula’s inputsIt’s proprietary
Regions and languages?Named and configurableNot mentioned
How to choose, in orderHow to choose, in order
Do these in order. Most shortlists shorten dramatically at question two.
What the secondary listicles got wrongWhat the secondary listicles got wrong
Each bar is one discrepancy found. This is why the prices on this page carry a date and a link.

Not sure which of these applies to you?Tell us the situation and we will say plainly what we would do first, and what we would not.

Talk it through

How do these tools differ from ordinary rank trackers?

A rank tracker records a position in a list. There is no list inside an AI answer, so these tools record something different: whether you were mentioned, whether you were linked, and who appeared with you — across repeated samples, because the answer changes between runs.

That last point is the structural difference and it explains the pricing. A rank check is one cheap deterministic request. An AI visibility check is a sampled observation of a non-deterministic system, which has to be repeated to mean anything. You are paying for repetition, which is why prompt allowance rather than keyword count is the unit of pricing.

Rank tracking against AI visibility tracking
Rank trackerAI visibility tracker
What it recordsA position in a listWhether you are mentioned or cited
DeterminismBroadly repeatableVaries between runs by design
Unit of pricingKeywordsPrompts or credits
Frequency neededDaily is plausibleRepetition is required for meaning
Competitive viewWho ranks above youWho appears with you
ActionabilityModerateLow on its own
Glossary of terms these tools use differently from each other
TermUsual meaningWhere vendors differ
PromptOne tracked questionWhether variations count separately
MentionBrand named in an answerWhether partial matches count
CitationNamed with a linkWhether any link counts or only attributed ones
Share of voiceYour share of mentionsWhich competitor set is the denominator
CreditOne unit of consumptionWhat consumes one varies widely
Answer engineAn AI surface trackedWhether AI Overviews and AI Mode count as one or two
AuditA page-level checkDepth varies from a title check to a render test

How we checked these prices, and what we did not check

Every price on this page was read from the vendor’s own pricing page on 15 September 2026 and is linked below. We did not take a briefing from any vendor, we have no affiliate relationship with any of them, and we have not run a controlled comparison of their accuracy.

That last limitation is worth stating plainly, because it is the one a buyer most wants resolved and the one no published comparison currently answers honestly. Testing whether two trackers agree about the same brand on the same prompts on the same day would be genuinely useful work. It is also expensive, and anyone claiming to have done it should be asked for the method before the conclusion is believed.

What this page does assert is narrow and checkable: these are the prices the vendors publish, this is the date we read them, these are the engines each vendor names, and these are the specific places where widely-circulated comparison articles disagree with the vendor’s own page. Prices change; the date is on every figure so you can tell how stale this is.

What we read

Each vendor’s own pricing page, plus the product page where pricing was absent.

What we recorded

Plan names, published monthly prices, billing terms, prompt or credit allowances, and the answer engines each vendor names publicly.

What we did not do

Trial every product, compare their accuracy against each other, or accept a vendor briefing. None of those inform anything stated here.

Our interest

Progression Agency sells answer engine optimisation services. We do not sell, resell or receive commission on any tool listed. A page that recommended buying nothing would suit our commercial interest as well as this one does, which is the reason to state the interest rather than hide it.

A page that recommended buying nothing would suit our commercial interest as well as this one does, which is the reason to state the interest rather than hide it.

Does buying a tool improve your AI visibility?

No. Measuring something does not change it. A tracker tells you where you stand and whether your work moved it; the work itself is crawler access, rendering, structure, corroboration and entity consistency.

This is obvious stated plainly and routinely obscured in how these products are marketed, because ‘optimisation platform’ sounds more valuable than ‘measurement instrument’. Several of these tools do include page audit features, and those are genuinely useful — but a page audit inside a visibility tracker is a lightweight check, not a technical audit, and none of them read your server logs.

If your budget only stretches to one thing this quarter, fix the site. You can measure it manually in the meantime for nothing.

What actually moves AI visibility, and whether a tool does it
LeverImpactDoes a visibility tool do this?
Unblocking AI crawlersDecisive — binaryNo
Server-side renderingVery highNo
Answer-first structureHighNo, though some audit features touch it
Third-party corroborationHigh, slowNo
Entity consistencyModerateNo
Schema reconciliationModeratePartially, via page audits
Knowing whether any of it workedEssentialYes — this is the job
If you only do one thing this quarter
Your situationDo thisNot this
You have never checkedAsk the assistants manuallyBuy a subscription
You checked and you are absentAudit crawler access and renderingBuy a subscription
You fixed the technical issuesTake a baseline, then subscribeAssume it worked
You are already measured and visibleCorroboration and content depthAdd a second tool
You manage many brandsBuy the platformTrack manually at that scale

How do the tools differ on the things that are hard to compare?

Four specifications decide day-to-day usefulness and none of them appear in a pricing table: run frequency, region and language coverage, how competitors are configured, and whether the page-audit feature is a real check or a title inspection.

Run frequency

Daily tracking sounds better than weekly and is only worth paying for in categories where the answer genuinely changes that fast. For most businesses a weekly sample with a stable prompt set produces a cleaner series, because daily data on a noisy signal invites over-reading.

Region and language

Answers differ by both, substantially. A tool sampling only United States English will not tell you what a buyer in another market sees, and this is the specification most often discovered after purchase rather than before.

Competitor configuration

The competitor set is the denominator of every share-of-voice number you will look at. Some tools let you define it precisely; some infer it. An inferred competitor set produces a number that moves when the inference changes, which is indistinguishable from your performance moving.

Page audit depth

Several tools include page audits. These range from a genuine structural check to a title and heading inspection. None of them fetch your pages with JavaScript disabled, which is the check that matters most, so treat the audit as a bonus rather than a reason to choose.

Seats and sharing

Otterly includes unlimited team members at every tier; Scrunch prices seats explicitly at 3 and 5. For an agency this is a larger cost difference than the headline price.

Historical data on signup

Some platforms show historical visibility from before you subscribed. That is genuinely useful for establishing a trend quickly, and it is a real differentiator worth asking about.

That is genuinely useful for establishing a trend quickly, and it is a real differentiator worth asking about.

What does the pricing tell you about where this market is going?

Two things. Prices are clustering into a sub-$100 self-serve tier and a $300-plus team tier, and the vendors with the broadest engine coverage are the ones least likely to publish a price at all.

The second pattern is the more interesting one. Engine breadth is expensive to maintain — each surface changes independently — and the vendors carrying the most of it are pricing through sales conversations rather than pricing pages. That is a reasonable commercial choice and it has a consequence for buyers: the products hardest to compare are the ones doing the most.

The self-serve tier

Roughly $20 to $99 a month, fifteen to a hundred prompts, four to six engines. Enough for one brand to establish whether it appears and whether that is changing.

The team tier

Roughly $189 to $500, hundreds of prompts, multiple seats and dashboards. This is where agencies and multi-brand marketing teams sit.

The enterprise tier

Custom at every vendor offering it, with SSO, API, compliance and support attached. The price is a conversation and the differentiators are procurement features rather than measurement.

What is missing from the market

Nobody is publishing an independent accuracy comparison — whether two trackers agree about the same brand, on the same prompts, on the same day. Until someone does, all of these numbers are internally consistent and externally unvalidated.

What happens to these tools if the AI surfaces change?

Their coverage changes with them, which is a maintenance cost you are renting rather than carrying. That is a genuine argument for buying rather than building, and it is the strongest case these products have.

Every one of these surfaces alters its behaviour without notice. User agents change, answer formats change, citation behaviour changes, and new engines appear. A tool absorbs that work. A spreadsheet of manual checks does not, which is the point at which most manual programmes quietly stop.

It also means the specification you bought is not the specification you will have in a year. Check how a vendor has handled past changes before assuming they will handle the next one.

Why coverage changes

Each answer engine is a separate product with its own roadmap. A tracker must maintain a distinct integration per surface, and those break.

Why your baseline can break

If a vendor changes how it counts a mention, your series changes with it. Ask whether methodology changes are announced and whether history is restated.

Why exports matter more than they look

An export is insurance against both a vendor change and a vendor exit. In a market this young, some of these products will not exist in three years.

What to keep yourself

The prompt set, the raw answers where you can get them, and the dates. Those are yours and they survive any vendor decision.

Specifications that are hard to compare, and how to ask about them
SpecificationWhy it mattersHow to ask
Run frequencyDaily and weekly are different productsHow often, exactly?
Regions and languagesAnswers differ by bothWhich, by name?
Competitor configurationIt is the denominator of share of voiceDo I define the set or do you infer it?
Page audit depthRanges from structural to title-onlyDo you fetch with JavaScript disabled?
SeatsLarger cost difference than headline price for agenciesHow many seats at this tier?
Historical dataEstablishes a trend fasterDo I see history from before I signed up?
Methodology changesCan break your baselineAre changes announced and is history restated?
Risk of building this yourself instead
FactorBuild it yourselfBuy a tool
CostStaff time, ongoingSubscription, predictable
Engine coverageWhatever you maintainMaintained for you
Surface changesYou absorb themVendor absorbs them
HistoryYours permanentlyYours only if you export
MethodologyYou control itVendor controls it
Realistic outcomeLapses within two quartersRenews and is used

Prefer to see the numbers on your own site?We will run the audit described above and walk you through what it finds.

Request an audit

Vendor by vendor, what each one is actually for

Read this as positioning rather than ranking. Each of these is a reasonable choice for a particular buyer and a poor one for others, and the differences that matter are prompt allowance, engine coverage and whether you can compare the price at all.

Otterly

The clearest published pricing in the category: $29, $189 and $489 tied explicitly to 15, 100 and 400 prompts, with unlimited team members at every tier and 15% off annual. Four engines — ChatGPT, Google AI Overviews, Perplexity and Copilot. Best for a single brand that wants to know its tier before talking to anyone.

Rankscale

The widest published engine list we found — ten named surfaces including Claude, Grok, DeepSeek and Mistral — priced by credits from $20 Essentials through $99, $385 and $780. Credit pricing is the hardest to forecast and the cheapest at volume, which suits agencies more than single brands.

Scrunch

Published tiers at $300 and $500 month-to-month, $250 and $417 annually, with seats, custom and industry prompts, personas and page audits counted explicitly at each tier. Six platforms including Meta. The explicit seat counts make agency arithmetic straightforward.

Ahrefs Brand Radar

Custom Prompts from $50 and free in every paid Ahrefs plan, with the AI Visibility Index at $199 standalone. If you already pay for Ahrefs, this is the cheapest serious starting point in the category — which is the opposite of what the comparison articles say about it.

Semrush AI Visibility Toolkit

$99 per domain billed annually, 25 custom prompts, four engines; Semrush One combines it with the SEO toolkit from $199. Per-domain pricing is predictable for one brand and multiplies quickly across several.

Profound

A free trial covering ten prompts on ChatGPT, then custom enterprise. The published enterprise feature list is the broadest here — nine answer engines, SSO/SAML, SOC2, API, custom regions and languages — and there is no published price for any of it.

Peec AI

Four tiers named Starter, Pro, Advanced and Enterprise, with no numbers published against any of them. Widely quoted in comparison articles at two mutually contradictory price points, neither of which appears on the vendor’s own page.

SE Ranking AI Visibility Tracker

Five surfaces including Google AI Mode, with five free checks advertised and no pricing on the product page. Sits inside a broader SEO platform, which suits teams already using it.

The numbers on this pageThe numbers on this page
All figures read from vendor pricing pages on 15 September 2026.

More from Progression Agency

The rest of the library, by what you are trying to do

AI, AEO and what is changing

Websites and design

Choosing and working with an agency

Social, content and brand

By industry and by situation

Want this done for your site?We build and maintain the search, content and paid programmes described on this page.

Get a free proposal

AI, AEO and what is changing

Frequently asked questions

What are AI visibility tools?
Software that samples AI assistants with a fixed set of prompts and records whether your brand is mentioned, whether it is linked, which competitors appear alongside it, and how that changes over time.
Which AI visibility tools publish their prices?
Of the eight checked on 15 September 2026: Otterly, Scrunch, Rankscale, Ahrefs Brand Radar and Semrush publish prices. Profound publishes a free trial and custom Enterprise only. Peec AI names tiers without numbers. SE Ranking’s product page shows none.
What is the cheapest AI visibility tool?
Of published prices, Rankscale Essentials from $20/month, then Otterly Lite at $29/month for 15 prompts across four engines. Ahrefs includes Custom Prompts in its existing paid plans at no extra cost.
How much do AI visibility tools cost?
Published entry tiers run roughly $20 to $99 a month for one brand; agency and multi-brand tiers run $300 to $800; enterprise is custom at every vendor that offers it.
Why do comparison articles disagree about the prices?
Because several of them publish figures the vendors do not. We found quoted Profound tiers that do not exist on Profound’s pricing page, two different Peec prices from two sources where Peec publishes none, and an Ahrefs minimum overstated many times over.
What is the best AI visibility tool?
It depends on engine coverage you will actually use and prompt allowance you actually need. For one brand, Otterly or Semrush at the entry tier. For agencies, Rankscale or Scrunch. For enterprise, Profound or Scrunch Enterprise.
Do I need an AI visibility tool at all?
Not to start. Asking the assistants your buyers’ questions manually costs nothing and tells you whether you have a problem. Buy the automation once you know the exercise is telling you something.
What is the difference between an AI visibility tool and a rank tracker?
A rank tracker records a position in a list. AI answers have no list, so these tools record mention, citation and competitive presence across repeated samples, because the answer varies between runs.
Which AI engines do these tools track?
All of them track ChatGPT. Most add Google AI Overviews, Perplexity and Gemini. Copilot, Claude, Grok, DeepSeek and Mistral are covered only by the broadest platforms.
How many prompts do I need to track?
Sixty to two hundred is a serious programme for one brand. Fifteen is enough to establish whether you have a problem. Prompt allowance is the main price driver at most vendors.
What is an AI visibility score?
A vendor-defined composite. It is a reasonable internal trend line and meaningless outside that tool, so it should not be reported upward as an external benchmark.
Can these tools tell me why I am not mentioned?
No. They observe. The why comes from server logs, a rendering test without JavaScript, and an editorial audit — none of which a visibility tracker performs.
Do AI visibility tools check whether crawlers can reach my site?
No. That requires server logs, and no tool in this category reads them.
Is there a free AI visibility checker?
Several offer limited free checks — Profound’s trial covers 10 prompts on ChatGPT once, SE Ranking advertises five free checks. Asking the assistants yourself is free and unlimited.
What is an LLM SEO tool?
The same category under a different name. LLM SEO tool, AI visibility tool, AEO tool and GEO tool describe substantially the same products.
Are ChatGPT SEO tools the same thing?
Usually not. Most products sold as ChatGPT SEO tools help you use ChatGPT to produce SEO work — prompts, writing, audits. Visibility tools measure whether ChatGPT mentions you. Different jobs.
How often should the tracking run?
Weekly is ample for most businesses; daily suits fast-moving categories. What matters more is that the prompt wording never changes between runs.
Can I export my data?
Check before buying. Several vendors gate export and API behind higher tiers, which means your measurement history lives inside a subscription until you upgrade.
Do these tools work outside the United States?
Most support multiple regions and languages, and answers genuinely differ by both. Confirm your markets are covered rather than assuming.
Should an agency buy one tool or one per client?
One platform with per-client dashboards and enough seats. Rankscale and Scrunch publish seat and dashboard counts at each tier, which makes that arithmetic straightforward.
What does a page audit inside these tools do?
A lightweight on-page check. It is useful and it is not a technical audit — none of them fetch your pages with JavaScript disabled or read your logs.
Will buying a tool improve my AI visibility?
No. Measuring does not change the reading. The work is crawler access, rendering, structure, corroboration and entity consistency.
How current are the prices on this page?
Every figure was read from the vendor’s own pricing page on 15 September 2026, and each vendor is linked so you can check it yourself. Prices in this category change frequently.
Do you have an affiliate relationship with any of these tools?
No. Progression Agency sells answer engine optimisation services and does not sell, resell or receive commission on any tool listed here.

Want this done for your site?We build and maintain the search, content and paid programmes described on this page.

Get a free proposal

Get a free marketing proposal

Tell us what you are trying to grow and we will come back with a plan, not a pitch deck. Same-day reply on weekdays.

Privacy Preferences
When you visit our website, it may store information through your browser from specific services, usually in form of cookies. Here you can change your privacy preferences. Please note that blocking some types of cookies may impact your experience on our website and the services we offer.
Contact Us