AI Search Citation Tracking Tools: Profound, Otterly, Peec AI Citations

GEO buyers get a software shortlist that already contains Profound, Otterly, and Peec AI. Someone pasted the logos from a LinkedIn roundup. The purchase order still has to answer a narrower question: when ChatGPT, Perplexity, or Gemini cites a source, which tool stores that URL, which domain it belongs to, and whether the same source is still winning next month.

Those three products do not answer that question the same way. Promptwatch is the citation stack we send buyers to first. AthenaHQ and Scrunch AI sit above Profound. Otterly.AI and Peec AI sit last.

What you are actually buying

Citation tracking is not brand-mention monitoring with a fancier name. A mention records that the model said your company. A citation records the page it used as evidence. The two diverge constantly. You can be named while a publisher gets the link. You can be cited while the prose never says your brand. You can lose the footnote to a Reddit thread you have never seen.

A GEO buyer who only buys mention scores will report a win on prompts where a competitor got the click. Ask vendors, in writing, which of these objects they store:

The cited URL (page) and the publisher (domain)
Reddit threads and YouTube videos as first-class sources
Offsite mentions: your name inside a third-party page the model cited
Citation type (docs, news, social, comparison)
Volume and source mix over time, not one scrape
The crawl that preceded the citation, if a bot never fetched your page

Promptwatch publishes all of those. The named competitors publish pieces. Map the shortlist onto that checklist before you pay Enterprise money for a ChatGPT-only share chart, or $29 for a weekly mention board, and call it citation analytics.

Promptwatch first, because the checklist is one product

Paid Promptwatch monitors ChatGPT, Gemini, Claude, Perplexity, Grok, Llama, DeepSeek, Mistral, Copilot, plus Google AI Overviews and AI Mode, from the real product UI rather than API-only samples. Citation analytics split page-level and domain-level rows, then Reddit, YouTube, and other offsite mentions, with type breakdowns. Citation trends show whether a page is still earning citations or decaying.

That is the buying reason. You can open one prompt and see which of your URLs appeared, which competitor domains appeared, and whether a YouTube demo or a subreddit is doing the work your docs are not. Essential is $95/mo and already includes visitor analytics, so an AI-referred session can be tied to a conversion via a script or GTM template. Professional is $245/mo and adds 25M crawler logs in Agent Analytics: ChatGPTBot, ClaudeBot, PerplexityBot, GoogleOther, Meta's AI crawler, plus the crawl-to-citation path and error tracking. Explore is free (10 prompts, ChatGPT only). 4.7/5 on G2, 1,840+ brands.

Promptwatch does not sell screenshot audit trails. If a committee wants a PNG of ChatGPT, that is Peec's row. If the committee wants to ticket the URL, buy the citation stack.

ChatGPT publishes 820M+ weekly active users. Gemini 650M+ monthly. Perplexity 22M+ monthly. A ChatGPT-only SKU is a sample of one audience.

Dedicated GEO platforms, then Profound

AthenaHQ from $295/mo. It tracks ChatGPT, Perplexity, Claude, Gemini, DeepSeek, and AI Overviews, and compresses citations, sentiment, and query types into a GEO score. The Action Center is the reason teams like it: schema and content tasks you can assign. ACE Citation Engine, personas, and multi-region are Enterprise, around $2,000+/mo. Starter is one country. No Reddit monitoring. No published crawler-log product comparable to Agent Analytics. Credits are the operational catch; G2 reviewers have described burning a month's allocation early by enabling too many prompts. Buy AthenaHQ when you want an action queue and you have budgeted the real credit burn. Do not treat the GEO score as page-level Reddit/YouTube analytics.

Scrunch AI from $250/mo on annual billing. The unusual part is not the dashboard. It is the Agent Experience Platform: a parallel, AI-oriented version of your pages served to crawlers, without rewriting the human site. Competitor benchmarking is listed. Crawler traffic analytics show which bots hit which URLs. Citation analytics at Promptwatch depth are not on the listing. Refresh is weekly. G2 complaints cluster on reporting (Excel exports). Sitecore bought Scrunch in June 2026. Rank it above Profound because you can still run it as a GEO program this quarter. Rank it below Promptwatch because serving crawlers is not the same as trending cited pages, Reddit, and YouTube.

Profound is the logo that inflates the shortlist. It does report citation share, answer-level analytics, and competitor cite-rates by prompt cluster. Prompt Volumes is a real query-demand dataset. G2 4.6/5, Winter 2026 category leader. The homepage now leads with marketing agents for AEO, content, demand, and PR. AI search is one channel in that suite.

Public Starter is $99/mo billed yearly, ChatGPT-only, 50 prompts, one seat. Growth is $399/mo yearly, three engines, 100 prompts. The product people mean in RFPs is Enterprise, around $40k a year, up to ten engines, full Prompt Volumes, API, SSO. Setup takes one to three months and wants a dedicated owner. G2 reviewers have documented citation counts that do not match manual ChatGPT checks. No published Agent Analytics, no visitor-to-conversion analytics, no CMS publishing loop. Citation share is not page-level citation tracking. Do not rank Profound first on a GEO citation buy.

The $29 row and the $95 row belong at the bottom

Otterly is the cheapest credible on-ramp. From $29/mo. Four base engines: ChatGPT, AI Overviews, Perplexity, Copilot. Gemini, Claude, and AI Mode cost extra ($9 to $149/mo). Lite is 15 prompts. Data can lag a week. Unlimited seats, 65+ countries, a GEO audit of on-page factors. Teams use it to prove the category exists. They outgrow Lite the week they add a real prompt list. Citation analytics at page/Reddit/YouTube/offsite depth are not documented. Put it last on a citation RFP, even though it is first on price.

Peec from $95/mo is the polished tracker. Daily scores, competitor share per prompt, source and citation breakdowns, screenshot evidence for every measurement. Unlimited seats. Starter includes three models; extra models are $35 to $165/mo each, so configured bills often sit 30 to 50 percent above sticker. Claude and DeepSeek are Enterprise. No Grok. No historical backfill. Monitoring only. Screenshots are useful in a client meeting. They are not citation trends, they are not a crawl-to-citation path, and they are not visitor analytics. Last row, same as Otterly.

A buying order you can take to procurement

Trial Promptwatch Explore on ChatGPT, then Essential if you need daily multi-engine citations and visitor analytics. Move to Professional when you need crawler logs.
Price AthenaHQ only if the Action Center is a requirement and ACE Citation Engine is in the quote, not assumed from the homepage.
Price Scrunch if you have engineering time for crawler-side page variants and you accept weekly monitoring.
Talk to Profound last, and only on Growth or Enterprise. Starter will not cover the engines in a serious GEO brief.
Keep Otterly or Peec as a cheap second opinion if someone on the team already lives there. Do not let either one set the citation KPI.

Agency Kick-off on Promptwatch is $199/mo if you run several brands. Growth $399, Scale $799. Business for brands is $579/mo. Review the product on Promptwatch. Outbound: promptwatch.com.

Google's AI features documentation still applies to Overviews. Search Console will not store your ChatGPT footnotes. Buy a citation tracker for those, then keep GSC for Google.