Webflow's AEO Analytics API: What Should GTM Teams Report?

What did Webflow just ship for AEO reporting?

On October 13, 2026, Webflow added an AEO Analytics API to its Data API. According to the Webflow developer changelog, it lets you read how AI bots visit your site and how answer engines mention and cite you. For a revenue team, that turns AI visibility from a screenshot habit into data you can pipe into a report.

My view is simple: AI search is a pipeline question, not a vanity question. The hard part was never the opinion. It was the evidence. Most teams I talk to check ChatGPT or Perplexity by hand, paste a few answers into a slide, and call it tracking.

An API changes the shape of that work. Once numbers come out of a system on a schedule, they can sit next to pipeline, demo requests, and closed revenue in the same report. That is the part I care about, so this article skips the hype and looks at what the release actually contains and what a go-to-market team should do with it.

What exactly is in the AEO Analytics API?

The changelog lists thirteen endpoints in four groups: bot traffic, prompts, visibility, and citations. Bot traffic shows which AI crawlers request your pages. Prompts return tracked questions and the answers engines gave. Visibility measures how often responses mention you. Citations show which domains and pages engines link to when they answer.

The bot traffic group has two endpoints. One returns total requests from tracked AI bots, split by provider or category. The other ranks the pages and files those bots requested most. For many marketing teams, this will be the closest thing to a crawl log they ever look at, because few of them ever open raw server logs.

The prompts group lists the prompts you track, their executions, and a single execution with its full response text. The visibility group returns a visibility score, described as the share of responses that mention your organization, plus a breakdown by topic or prompt. The citations group is the richest. It covers citation rate, top cited domains, cited paths for a domain, one page's share of citations, competitor citations, and a full list of citations with mentions.

I like that the competitor endpoint is framed narrowly. It returns responses that cited a competitor but not you. That is a ready-made gap list, and gap lists are what sales and content teams can actually work from.

Who can use it, and what do you need to switch on?

Access is gated. The changelog says the API only works for sites with the AEO analytics entitlement, every endpoint needs the sites:read scope, and Webflow AI must be enabled for the workspace for everything except the two bot traffic endpoints. Check Webflow's official developer docs for current plan details before you promise anyone a dashboard.

That last detail matters for planning. If your workspace has not enabled Webflow AI, you can still read bot traffic, but the prompt, visibility, and citation data will not be available to you. Before scoping any reporting work around this, I would confirm both conditions with whoever administers the workspace. It is a short check that saves an awkward conversation later.

The scope requirement is also worth noting for anyone wiring this into a warehouse or a tool like Make or n8n. A read-only token is enough. You do not need write access to the site to report on it, and I would not grant more than the job needs.

Why should a GTM team care about AI citations at all?

Buyers now ask AI assistants for shortlists before they ever visit a vendor site. If an engine names three tools and you are not one of them, you lost a deal you never saw. Citation data is the earliest signal you can measure for that invisible part of the funnel, which makes it a GTM metric, not an SEO curiosity.

Traditional attribution starts at the first tracked touch. A form fill, a demo request, a UTM tagged click. Everything that happened before that touch is guesswork. AI assistants have added a large new chunk of buyer research to the guesswork zone, because the conversation happens inside ChatGPT, Gemini, Claude, or Perplexity and leaves no referrer behind.

Citation share does not solve attribution. What it does is give you a leading indicator. If your share of citations on category prompts goes up, you should expect more branded search and more inbound that says "I found you through ChatGPT" in the "how did you hear about us" field. If it drops, you have a few weeks of warning before pipeline does.

That is why I keep pushing this toward the revenue team rather than the marketing team alone. The person who owns pipeline coverage should see this number, because it moves before their number moves.

Which AEO numbers should go into a weekly GTM report?

Pick three, not thirteen. I would report citation rate on your core category prompts, the list of competitor-only citations, and the top pages AI bots request. Those three answer the questions a revenue leader actually asks: is the company showing up, who is winning instead, and are engines even reading the pages that matter.

Citation rate is the headline. Track it on a fixed set of prompts that mirror how your best buyers describe their problem, not on every prompt you can think of. A stable prompt set is what makes week-over-week comparison honest.

The competitor-only list is the action item. Each row is a question where an engine trusted someone else's page over yours. That is a content brief, a comparison page, or a missing proof point waiting to happen. I would hand that list to whoever owns content every Monday.

Bot page requests are the sanity check. If the pages AI crawlers request most are old blog posts and your pricing and product pages barely appear, you have a structure or linking problem to fix before any content push will pay off. For a deeper look at the manual version of this work, my walkthrough on tracking AI search visibility without enterprise tools covers the spreadsheet approach many teams still need.

How should you connect AEO data to pipeline and CRM data?

Join on time and topic, not on individual leads. AI citation data will not tell you which contact read which answer. What you can do is line up weekly citation share by topic against weekly inbound volume and self-reported source by the same topic, then look for movement together over a quarter.

In practice I would set this up in three layers. First, a scheduled pull from the API into a simple table in Airtable, BigQuery, or a Google Sheet, one row per prompt per week. Second, a topic tag on each tracked prompt that matches the topic tags you already use for inbound leads in HubSpot or Salesforce. Third, a weekly view that shows citation share and lead count side by side for each topic.

Self-reported attribution is the bridge. If your demo form asks how someone found you, add an option for AI assistants. I wrote about how to qualify a lead that came from an AI assistant, and the short version is that these leads often arrive with a shortlist already formed, so sales needs to know which engine and which question got them there.

Be careful with correlation. A rising citation rate and a rising lead count in the same month can both be caused by a launch or a conference. One quarter of data is a hypothesis, two quarters is a pattern worth acting on.

What are the limits of this kind of data?

AEO analytics measures a sample of answers to prompts you chose. It does not see every conversation a buyer has with an AI assistant, and engine answers change from run to run. Treat the numbers as a directional panel, like a brand tracking survey, not as a census of AI search behavior.

Prompt selection is the biggest source of bias. If you only track prompts where you already rank well, your visibility score will look great and mean nothing. I would include at least a few prompts where you know a competitor dominates, so the report can show progress or the lack of it.

Engine variance is the second limit. Ask the same question twice and you can get different sources. That is why weekly or monthly trends are more useful than any single day. It is also why I would never tie a bonus or a content writer's pay to a single week's citation rate.

The third limit is entitlement. Not every Webflow site will have this data, and teams on other platforms will need a separate tool. If you work across several sites, plan for the report to have gaps rather than pretending every property is measured the same way.

Why does an API matter more than a dashboard?

A dashboard is useful for a marketer looking at one site. An API is different because it lets the data leave Webflow and land wherever revenue reporting already lives. That shift is what makes this release useful for GTM engineering, since pipeline reviews rarely happen inside a website builder.

I wrote earlier about Webflow's AEO analytics and what it means for site owners. A dashboard view answers "how is the site doing" for the person who logs into Webflow. The API answers a different question: can this data sit in the same place as pipeline, so the people who own revenue can see it without logging into another tool.

For agencies and freelancers, the API also means client reporting can be automated. I would rather build one scheduled pull and a clean weekly summary than export screenshots by hand for every client. That is the kind of system I build for clients, and it fits my brand line: I build websites and the automations behind them for lead generation.

What should you do next?

Confirm access first, then build the smallest useful report. Check that your site has the AEO analytics entitlement and that Webflow AI is enabled, pick ten to twenty prompts that mirror how buyers describe the problem, and set up one weekly pull of citation rate, competitor-only citations, and top bot-requested pages.

Run it for a full quarter before you change strategy based on it. Share it with whoever owns pipeline, not only with whoever owns the website. And keep the prompt set stable, because a moving prompt set makes every trend line a lie.

If you want help wiring the AEO Analytics API into your CRM reporting, or deciding which prompts are worth tracking for your category, reach out. Let's chat about what your buyers are asking and whether the engines are pointing them to you.