Tools
What does AI visibility cost for a company your size?
By Arnav Mukherjee, founder of TofuBofu · September 30, 2026
Updated 13 October 2026. Profound's pricing page no longer lists the $99 Starter and $399 Growth plans this post originally priced. We last saw them on 7 September 2026 and they were gone by 10 October 2026. Read on 13 October 2026, its pricing page shows a free 7-day trial and a custom-priced Enterprise plan, and the table and answers below are corrected to match.
TL;DR
- Budget $95 to $499 a month under 50 people, paying monthly. The spread is engine count.
- Read the tier, not the feature page. Profound's $99 Starter said "ChatGPT tracking only" on 7 September 2026 and was gone by 10 October 2026.
- Don't pay for Claude twice. Peec puts it on Enterprise only; Otterly sells it as an add-on.
- Get your baseline free before you buy anything. Track runs 5 queries on 5 engines monthly.
- Price the work, not the dashboard. Most of this category measures and hands the fixing back.
Somebody typed this into Google last month and landed on our site: "most affordable geo platforms for b2b saas companies under 50 employees". They didn't find what they wanted.
We hadn't written it, so Google matched them to a page about our B2B SaaS vertical instead, which answers a different question entirely and left a buyer with a budget and a headcount constraint reading generic positioning copy. Our average position on that query is 26.8, which is page three, and we deserve it.
So here's the page. Every price below was read at the vendor's own pricing page and carries the date it was read, because this category re-prices fast, because engine bundling moves inside a single quarter, and because comparison articles about it have been wrong before, including one of ours that quoted a competitor's prompt-database figure seven times at a number we'd misread. Sorted by budget.
How we priced this
- Every figure is from the vendor's own pricing page, never a comparison article. Read dates: Profound, Peec and Otterly on 28 August 2026, Profound re-read and screenshot-verified 7 September, Peec re-read 8 September in both billing modes, Scrunch 8 September, Suede and TofuBofu 30 September, Peec and Otterly again on 15 October 2026. Profound was read again on 10 October and 13 October 2026, by which point its Starter and Growth tiers were gone. Profound, Peec and Otterly were re-read on 31 October 2026 with the same prices and plans, and Scrunch on 1 November 2026.
- We list the tier's own words where a tier is narrower than its headline price suggests, because that's where buyers get caught.
- Company size is a sorting device, not a rule. It correlates with budget and it does not determine what you need. The last section says what does.
- We sell one of these. TofuBofu's own numbers are in the table on the same terms as everyone else's, and the concession section says where we're the wrong answer.
Solo founder or under ten people: what can you get for nothing?
More than you'd expect, and you should exhaust all of it before you spend anything, because the three options below will between them tell you whether you have a visibility problem worth paying to fix.
- TofuBofu Track, free forever. 5 queries per scan across all five engines, one scan a month, one brand, plus who AI names instead of you. Read from our own pricing page, 30 September 2026.
- Otterly Lite, $29 a month. The cheapest paid entry we found. Base tiers cover four engines, and Claude, Google AI Mode and Gemini are paid add-ons on top.
- Asking the engines yourself, free. Type your category's buying question into each assistant and read the answer. It doesn't scale and it isn't a trend line, but at this size it will tell you whether you have a problem at all.
At this size the honest advice is simple. You probably don't need a subscription yet. What you need is to know whether you're absent at all, and that answer is available free from more than one of the sources above, so spending $29 a month to learn it is a decision you can defer until you've seen the free version and found it wanting.
Under fifty people: why is this band so confusing?
Because vendors land within a few dollars of each other month to month and sell noticeably different things: Peec Starter at $95 and our Fix at $99, with Profound's $99 Starter alongside them when we read it on 7 September 2026 and gone by 10 October 2026. This is the band our searcher was in. It's also the band where reading the tier column rather than the headline price changes the answer most, because nearly the same headline price buys three chosen models at one vendor and every engine at another.
| Tool | Price | What that tier covers, in their words |
|---|---|---|
| Peec Starter | $95/mo $80 annual | Their tier says "Choose 3 models", and Claude appears only in the Enterprise column. |
| Profound Trial | Free 7 days | "50 prompts run daily for 7 days" on "ChatGPT, Gemini, Google AI Overviews", read 13 October 2026. The only paid plan is Enterprise, priced "Custom". The $99 Starter read on 7 September 2026, "ChatGPT tracking only", was gone by 10 October 2026. |
| TofuBofu Fix | $99/mo | 10 queries per scan, all five engines, weekly, plus about 10 drafted articles in a typical month. |
| Otterly Standard | $189/mo | Four engines at the base price, with Claude, Google AI Mode and Gemini sold as paid add-ons. |
| Peec Pro | $245/mo $205 annual | Still three chosen models included, with more as paid add-ons, and Claude is still in the Enterprise column. |
| Scrunch Starter | $300/mo $250 annual | 350 prompts. Its page lists seven platforms, Claude and Gemini among them, read 1 November 2026. |
Stated as a sentence, because a table isn't quotable: when we read Profound's page on 7 September 2026, $99 a month bought a tier whose own description said "ChatGPT tracking only", a tier gone by 10 October 2026, and at $95 a month you can buy one that includes three models, with more as paid add-ons and Claude locked behind Enterprise. The headline prices are close. What they query is not.
Two traps live in this band.
- Counting engines on the feature page instead of the tier column. This is where Claude keeps disappearing: a vendor can list six engines across its marketing and still sell you three at your price point, and nothing about that is dishonest because the tier column said so.
- The add-on ladder. Otterly's extra engines are priced per tier, so a full roster gets more expensive as you move up, which is the opposite of what most buyers assume bundling does.
Fifty to two hundred and fifty: what changes?
Two things change at this size, and neither of them is depth of measurement, which is the axis most vendors use to justify the step up in price.
- Brands, not headcount. A group at this size usually tracks more than one product or market, and multi-brand is where per-seat and per-brand pricing starts to bite. TofuBofu's self-serve plans cover one brand each; Rule, from $1,999 a month, covers 5 brands, each with its own scans and trend.
- Someone to do the work. At ten people the founder does it. At 150 there's a marketing team, and the question flips from "what does it cost" to "who implements what it finds". That is the line between our Elevate at $499, where we write and stage the blogs and you publish, and Dominate at $999, where we publish them after you approve.
- Published ceilings appear. Peec Advanced is $495 a month, or $420 if you commit annually. Otterly Premium is $489. Profound Growth was $399 for "3 Answer Engines tracked" when we read it on 7 September 2026, and by 10 October 2026 Profound's page carried no paid price at all. Above these, most vendors stop publishing.
Get the baseline before you price anything. A free TofuBofu scan puts your real buying questions to all five engines and shows you which ones already name you. Then you know what you're shopping for.
Run a free scanEnterprise: why do some vendors publish no price at all?
Because their buyer is a procurement process rather than a founder with a card, and a published price would anchor a negotiation that they would rather open with a scoping conversation about your particular risk exposure. That isn't cynical. It's how enterprise software has always been sold.
Suede Web Systems is the clean example. We went through their whole site on 30 September 2026: no pricing page, no pricing link, no dollar figure anywhere, only "Book a Demo". That's consistent rather than evasive, because their four solution pages are Public Affairs and Policy, Campaigns and Advocacy, Corporations and Regulated Industries. A public affairs team buying narrative intelligence is not comparison-shopping a $99 card payment.
Otterly's Enterprise is custom too, read 31 October 2026, with a floor of 1,000 prompts rather than a dollar figure. Profound, Peec and Scrunch all list Enterprise as custom. Treat unpublished pricing as information about the sales motion, not about quality: expect a scoping call and a number that depends on what you ask for.
When is headcount the wrong question entirely?
Everything above sorts by company size, because that's how people search and because headcount is a decent proxy for budget. It's a poor proxy for need. Three cases where sorting on it will actively mislead you:
- A narrow category needs fewer queries than a broad one, whatever the headcount. Twelve people selling one thing in one vertical may need ten well-chosen buying questions and nothing more.
- Several brands beats several people. Two products under one roof is a bigger pricing input than doubling the team.
- Regulated or reputation-led work is a different product, not a bigger tier. If your risk is what a regulator or a journalist hears, a vendor-recommendation tracker is the wrong shape and Suede's category is the right one.
And the concession we owe you on our own pricing: TofuBofu is the wrong buy if all you want is a deeper measurement. Profound samples harder and Peec's interface is cleaner for a pure tracker. We price around writing and publishing the fix and then re-scanning to show it moved, so if your team already has the writing covered you're paying for something you won't use.
Frequently asked questions
What is the cheapest way to track AI visibility?
Free, if you only need a monthly baseline. TofuBofu's Track tier is free forever and runs 5 queries per scan across all five engines, once a month, on one brand. The cheapest paid entry we found is Otterly at $29 a month for its Lite tier, read at their pricing page on 28 August 2026, and that tier covers four engines with Claude, Gemini and Google AI Mode sold as paid add-ons. So the honest comparison isn't $0 against $29, it's what each one queries. A tier that excludes the engine your buyers use is not cheaper, it's narrower.
What should a company under 50 people expect to pay?
Between $95 and $499 a month, and the spread inside that band is mostly engine count rather than depth. Peec starts at $95 a month paid monthly, or $80 if you commit annually, and each tier below Enterprise says 'Choose 3 models', with Claude available only on Enterprise. Profound listed a $99 Starter described on their own page as 'ChatGPT tracking only' with 50 prompts when we read it on 7 September 2026, and had removed it by 10 October 2026, leaving a free 7-day trial and a custom Enterprise plan on 13 October 2026. TofuBofu's Fix is $99 a month for 10 queries per scan across all five engines, weekly. All figures read from each vendor's own pricing page, Peec on 28 August 2026 and re-read in September, Profound on the dates given, TofuBofu from its live pricing page on 30 September 2026.
Why do some vendors not publish a price at all?
Because they sell to procurement rather than to a founder with a card. Suede Web Systems is the clean example: we checked their whole site on 30 September 2026 and there's no pricing page, no pricing link and no dollar figure anywhere, only 'Book a Demo'. That's a deliberate choice and it matches their buyer, who is a public affairs or corporate communications team rather than a founder. Unpublished pricing isn't a red flag by itself. It does tell you the sales motion involves a call and a scoping exercise, which is a different purchase from a card on a checkout page.
Does paying more get you more engines?
Usually yes, and that's the part worth reading closely, because engine coverage is the main thing the tiers actually differ on. Profound's own tier descriptions, read on 7 September 2026, ran from 'ChatGPT tracking only' at Starter to '3 Answer Engines tracked' at Growth to 'Up to 9 Answer Engines tracked' at Enterprise. Starter and Growth were gone by 10 October 2026, and on 13 October 2026 the page lists a free 7-day trial on three engines and an Enterprise plan with 'Up to 9 Answer Engines tracked'. Scrunch moved the other way: read on 1 November 2026, its Starter tier is $250 a month billed annually, or $300 month to month, and its page lists ChatGPT, Claude, Gemini, Perplexity, Google AI Mode and AI Overviews, and Meta. On 8 September 2026 its $250 Core tier listed four engines, with Claude and Gemini on Enterprise. Otterly's base tiers cover four engines with Claude, Google AI Mode and Gemini as paid add-ons that scale with the tier. TofuBofu is the exception by design: every plan including the free one queries the same roster, and what changes is how many queries you run and whether we write the fixes.
Is a bigger company automatically on a bigger tier?
No, and treating headcount as the input is the mistake this page is trying to correct. What actually sets the tier is how many buying questions matter in your category and how many brands you're tracking. A 12-person cybersecurity firm in one niche might need 10 well-chosen queries. A 400-person group with five product lines needs 25 per brand and five brands, which is a different plan for reasons that have nothing to do with headcount. Headcount is a useful first sort because it correlates with budget, and it's a bad final answer.
What does the monthly price not include?
On most of the category, the work. Most tools on this page price monitoring first, and the briefs or drafts some of them add still need someone to approve, publish and check them. That's the split worth pricing against, because a $99 dashboard plus a person to act on it is not a $99 line item. Check three things on any quote: which engines are included at that exact tier rather than on the feature page, how many queries the number buys, and whether anything gets written and published or only measured. The third one is where the real cost difference sits.
How often do these prices change?
Often enough that every figure on this page carries a read date, and you should treat any undated price in this category as unreliable. We've watched entry tiers move and engine bundling change inside a single quarter. Every number here was read at the vendor's own pricing page rather than taken from a comparison article, because secondary summaries in this category have been wrong before. If you're making a decision on one of these figures, open the vendor's page and check it yourself; it takes a minute and the page you're reading may be older than the price.
Sources and further reading
- Profound pricing. $99 Starter, $399 Growth, Enterprise custom, read 28 August 2026, re-read and screenshot-verified 7 September 2026. Starter and Growth were gone by 10 October 2026. Read 13 October 2026, the page lists a free 7-day Trial and a custom Enterprise plan.
- Peec AI pricing. Billed annually, $80 Starter, $205 Pro and $420 Advanced; month to month the same tiers read $95, $245 and $495. Read 28 August 2026, re-read 8 September in both billing modes, and re-read 15 October 2026, when the page opened on the monthly view at the same prices.
- Otterly pricing. $29 Lite, $189 Standard, $489 Premium, Enterprise from $1,000. Read 28 August 2026.
- Scrunch pricing. $250 Core, Enterprise custom, read 8 September 2026; Starter $250 billed annually or $300 monthly, Growth $417 or $500, Enterprise custom, read 1 November 2026.
- Suede Web Systems. No published pricing, verified across the site 30 September 2026.
- TofuBofu pricing. Free, $99, $499, $999 and from $1,999. Read 6 November 2026.
