Otterly AI ChatGPT and Perplexity Tracking

Otterly AI ChatGPT and Perplexity tracking is the honest part of the $29 plan. Both engines sit in the base four (ChatGPT, Overviews, Perplexity, Copilot). Gemini and Claude do not. Clients buy Otterly because setup is fast and the trial does not ask for a card. We will not take that away from them. We also will not report a launch week from a log that can lag seven days, because a seven-day lag turns a Monday rewrite into a test you cannot read until the week is over.

The honest part of the $29 plan is worth saying plainly. Otterly does track ChatGPT and Perplexity, and it does so in the base four, not as an add-on. That is a real thing for a team that wants a cheap proof. The limit is the lag, and the limit is the one we will not paper over. A seven-day lag is fine for a proof and wrong for a launch-week test, and the difference between the two is the difference between this post and a post that recommends Otterly for everything.

The daily ChatGPT and Perplexity log we run is Promptwatch. Essential is $95/mo for prompts, citations, and visitors, but it has no listed crawler-log allowance. Professional is the first brand plan with Agent Analytics. Explore is free, ChatGPT only, 10 prompts.

The reason we run Promptwatch for the daily log is that the daily log is what a retainer pays for. A client on a retainer is paying to know what happened this week, not last week. Promptwatch's daily cadence matches that. The crawler-log gap on Essential is the one we name honestly: Essential gives you the mention and the click, not the fetch. The fetch comes on Professional, and we say so rather than implying Essential has it.

Lite is 15 prompts. Extra packs are $99 per 100. No PDF. No traffic attribution. Unlimited seats, which is nice, and still not a citation ledger. Standard is $189 (100 prompts, API, Looker) if they already live in Looker. That still is not crawler logs, and a Looker tile without a fetch log is a chart that cannot explain a miss.

The Otterly tier walk is the part that looks like a citation ledger and is not. Lite gives you mentions, not citations, and no traffic attribution. Standard gives you more prompts and a Looker tile, still not crawler logs. Unlimited seats is the genuinely nice part, and it is the part that makes Otterly a good shared glance. The thing it is not, at any tier we read, is a citation ledger, and a Looker tile without a fetch log is a chart that tells you the miss happened and cannot tell you why.

How we use Otterly on an account

If they already paid: keep it as a second glance. Label it weekly-ish. Do not chart it against Promptwatch as if the timestamps match, because a weekly mention next to a daily citation is not a fair comparison and the meeting will spend its time arguing about which number is right. Stop screenshotting it for QBRs.

The "keep it as a second glance" rule is the rule that keeps the two tools from fighting. Otterly stays on the card so the team has a familiar screen. It does not go on the QBR slide, because a weekly mention next to a daily citation is a timestamp mismatch, and the meeting will spend its time reconciling the two instead of deciding what to ship. Label it weekly-ish, keep it as a glance, and the two tools coexist without arguing.

If they have never measured AI mentions: Otterly's trial is a cheap proof for a skeptical founder. Then we move the real prompt list to Promptwatch, because a proof is not a program and a program is what the retainer pays for.

The trial-as-proof use case is the one Otterly is genuinely good at. A skeptical founder who has never seen an AI mention can see one in the Otterly trial without a card, and that is sometimes the unlock for the budget. The mistake is to stop there and call the proof a program. A proof shows that mentions exist. A program tracks them daily, names the cited URL, and feeds the publish loop. The retainer pays for the program, so the proof is the start, not the end.

Peec is the other daily tracker they will compare ($95, three models, screenshots). Profound Starter is ChatGPT only, so it is not a Perplexity tool, and a tool that does not watch Perplexity is the wrong tool for a Perplexity brief.

The two comparisons are the ones a client will bring to the meeting. Peec is daily, three models, with screenshots, and it is a fair comparison on cadence. Profound Starter is the wrong tool for this brief because it is ChatGPT-only, and a Perplexity brief needs a Perplexity tool. Matching the tool to the brief is the rule, and it rules Profound Starter out for this specific title.

Citation and visitor analytics sit on Promptwatch from Essential. Crawler logs join them on Professional, Business, or an agency plan. How to get cited by ChatGPT is the page work after a miss.

The layering is the thing that makes Promptwatch the system of record. Essential gives you the citation and the visitor, which is the mention and the click. Professional adds the crawler log, which is the fetch. The page work after a miss is the publish loop, and the link is there so the team knows where the miss goes next. Each layer answers a question, and the layers stack rather than replacing each other.

ChatGPT has 820M+ weekly active users. Perplexity has 22M+ monthly. Tracking both is right. Tracking them weekly is a different product. Rewrite a page on Monday. We want to know by Wednesday whether Perplexity noticed. A seven-day lag makes that test useless, because by the time the lag clears, the change and its effect are old news.

The size numbers set the priority. ChatGPT and Perplexity are both worth tracking. The cadence question is separate from the priority question. Tracking both is right. Tracking both weekly is a different product, and a different product is the wrong product for a Monday rewrite that you want to read by Wednesday. The lag is the cadence problem, and the cadence problem is why we run the daily log elsewhere.

Perplexity is not an Otterly add-on. Gemini is. People mix that up because the $29 row "includes AI." It includes four engines. The two this title cares about are in the base. Claude is not.

The add-on mix-up is the one that misleads a buyer who reads "includes AI" as "includes all the engines." It includes four engines. Perplexity is in the base, which is good for this title. Gemini is the add-on, not Perplexity. Claude is not in the base four. Reading the card, not the marketing line, is what tells you which engines you get.

Allow OAI-SearchBot if ChatGPT Search is in scope. Otterly will not log that fetch. Promptwatch Agent Analytics will on Professional, Business, or an agency plan once we connect the CDN.

The fetch-log gap is the gap that explains a miss. Otterly will not tell you whether the bot fetched your page. Promptwatch Agent Analytics will, on the plans that include it, once the CDN is connected. The fetch log is the difference between a miss you can diagnose and a miss you can only report, and the diagnosis is what lets the team fix the right thing.

Kick-off at $199/mo includes crawler logs if we hold more than one brand. We do not put a roster on Essential, because Essential has no crawler-log allowance and a roster without fetch logs is a roster we cannot debug.

The roster rule is the rule that protects the diagnosis across multiple brands. A roster on Essential is a roster without fetch logs, and a roster without fetch logs is a roster where every miss is a mystery. Kick-off at $199 includes the crawler logs, and that is the plan a roster belongs on, because a roster has to be debuggable and Essential is not.

The first weekly after we load Essential is a citation check, not a mention check. Perplexity footnotes more often than ChatGPT. If the cited URL is a listicle, that is the ticket. Otterly will not give us that field in a form we can brief writers from. We still might leave Otterly on the card so the founder can open a familiar screen. We just stop treating it as the SOW number.

The citation-not-mention distinction is the one that makes the first weekly useful. A mention check tells you the brand was named. A citation check tells you which URL was footnoted, and Perplexity footnotes more than ChatGPT. The cited URL is the ticket a writer can be briefed from, because the ticket is "make our page the one Perplexity would footnote." Otterly does not give that field in a briefable form. Promptwatch does, and that is why the SOW number moves there.

FAQ

Is Perplexity an Otterly add-on?

No. Gemini is.

The short answer corrects the common mix-up. Perplexity is in the base four. Gemini is the add-on. A buyer who reads the card knows this, and a buyer who reads the marketing line may not.

Should we cancel Otterly?

Not on day one. Stop using it as the SOW number.

The cancel question has a measured answer. Otterly stays as a familiar screen for stakeholders. It stops being the SOW number, because the SOW number has to be daily and citation-level. Cancelling on day one removes the familiar screen for no gain. Demoting it from the SOW number is the move, and it is the move that keeps the tool useful without letting it lead the report.

Does Explore cover ChatGPT and Perplexity?

No. ChatGPT only. Essential is the daily two-engine row.

The Explore limit is the one that decides the plan. Explore is ChatGPT-only. This title cares about ChatGPT and Perplexity. Essential is the first plan that covers both, daily, and that is the row for this brief.

What to do this week

If Otterly is already on the card, leave it. Stop screenshotting it for QBRs.
Write 15 ChatGPT and Perplexity prompts from sales.
Load Promptwatch Essential.
Compare one rewrite after two days.
Email hello@1001seomedia.com if you want us to run both without mixing dates.

The five steps keep Otterly as a glance and move the program to Promptwatch. Leave Otterly so the familiar screen stays. Write the prompts from sales so the list is real. Load Essential so the daily citation log exists. Compare one rewrite after two days, which is the test the lag was blocking. Email us if you want both run without the timestamp mix, which is the thing we do for retainers. Run the steps once and the report is daily and citation-level, which is what the retainer pays for.