Google Analytics 4 (GA4) is built to answer marketing and traffic questions. Where did visitors come from, which campaigns brought them, what did they engage with, and which of them converted. It is very good at those questions, and for most websites it is the reporting tool the whole company already reads.
A different question now sits beside those. Some of the interactions with your site, API and product come from AI agents acting for real customers, and from people working with an AI assistant beside them. Product, growth and security teams want to know how much of that is happening, on which routes, and what to do about it. That is the job of Proof of Actor, the actor intelligence product from AgentCredia.
This page compares the two plainly. GA4 facts below come from Google's own documentation as of November 2026.
The short version
| Google Analytics 4 (GA4) | Proof of Actor | |
|---|---|---|
| Main question | How do people find and use my site, and what converts? | Who or what is acting on this interaction, and what is the evidence? |
| Primary data collection | The Google tag in the browser, with the Measurement Protocol to augment it | Server middleware, edge, HTTP API, MCP servers, plus an approved browser collector |
| Known bots and spiders | Automatically excluded from reports, per Google's documentation | Classified and kept, so you can see and route them |
| AI agents | Not a reporting class in Google's bot exclusion documentation | A first-class result: Agent |
| AI-assisted people | Not described as a separate class | A first-class result: Human + AI |
| Unclear traffic | Counted as a user or excluded | Reported as Unknown, with a fallback you choose |
| Evidence per result | Aggregated reports | A basis on every result: verified, declared, inferred or unknown |
| Acts on traffic | No, it reports | Routes per your policies in your own code, after observation |
The sections below explain each row.
What GA4 reports are built to show
GA4 organizes everything around users, sessions and events, and then rolls them up into reports: acquisition by channel and campaign, engagement by page and event, and conversions by source. Those reports are aggregates. They are designed to show trends across thousands of visits, and they quietly assume that every user in them is a person browsing your site or app. Known bots are removed before the numbers reach you, which keeps a marketing report clean.
That design leaves a specific blind spot. A GA4 report can tell you that sign ups rose last week. It cannot tell you how many of those sign ups were driven by an agent acting for a customer, how many by a person drafting with an AI assistant, or how many by traffic nobody can explain, and it does not hand your application an answer it can act on while the request is still open.
Proof of Actor fills that gap one interaction at a time. Each request, session or tool call gets a class: Human when the signals point to a person, Human + AI when a person acted with AI assistance that was declared or observed through a supported integration, Agent for automated or agent-origin interaction, and Unknown when the evidence does not support a reliable call. Each class carries an evidence basis, verified, declared, inferred or unknown, so the number in a report and the decision in your code rest on the same stated evidence.
Where each one runs
According to Google's documentation, GA4 collects data through tagging with the Google tag (gtag.js), Google Tag Manager or Google Analytics for Firebase. Google also offers the Measurement Protocol, which sends events directly to Google Analytics servers over HTTP for server-to-server and offline interactions. Google states that the intent of the Measurement Protocol is to augment automatic collection, not to replace it.
That design suits browsing people. It leaves gaps for agents. Many AI agents fetch pages without executing JavaScript, call your API directly, or connect to your MCP server to complete a task. In those paths there is no browser tag to fire, so GA4 typically never hears about the interaction unless you build that reporting yourself.
Proof of Actor does not depend on a tag firing. It is installed in the request path, as Node.js or Next.js middleware, in an edge layer you control, behind a generic HTTP API or inside an MCP server, and a browser collector is optional on pages you approve. The integrations page lists each path. An agent that never renders a page, and so never sends a GA4 hit, still gets a result.
How GA4 treats bots and AI agents
Google's documentation on known bot-traffic exclusion states that, in Google Analytics properties, traffic from known bots and spiders is automatically excluded. Google identifies known bots using a combination of its own research and the International Spiders and Bots List maintained by the Interactive Advertising Bureau (IAB). The same page states that you cannot currently disable known bot traffic exclusion or see how much known bot traffic was excluded.
For marketing reporting, that is a sensible default. You want campaign and conversion numbers to describe people, not crawlers.
For understanding AI agent traffic, it means the most useful data is removed before you ever see it. Excluded traffic does not appear in reports, and Google's documentation on bot exclusion does not describe a separate reporting class for AI agents acting for users, or for people working with AI assistance. Agents that are not on a known list, and agents that never execute the tag, are outside that mechanism entirely.
What Proof of Actor adds to your GA4 reports
A segment GA4 cannot build on its own. Your application receives the Proof of Actor result, so it can attach the actor class and basis to the events it already sends to GA4, for example as an event parameter registered as a custom dimension. Acquisition, engagement and conversion reports can then be split by Human, Human + AI, Agent and Unknown for the traffic the Google tag does see.
A view of what known-bot exclusion removes. Google states that known bot exclusion cannot be disabled and that you cannot see how much traffic it removed. Proof of Actor keeps agents in view as Agent results, so you can see which routes they use and decide whether to serve them. An agent comparing your plans for a real customer may be a visitor worth designing for; the guide on serving AI agents structured content shows one approach.
Coverage where the tag never fires. API calls, MCP tool calls and pages fetched without JavaScript produce no GA4 hit. Proof of Actor classifies them in the request path, which is why its actor analytics by route and workflow complement GA4 rather than duplicate it.
Evidence behind every label. GA4 matches known bots against a list and removes them; Proof of Actor states why it reached each result. A verified basis means a request signature checked out against a published key under HTTP Message Signatures (RFC 9421). A declared basis means a client or your own application said so, and an inferred basis is marked as inference. That matters because a user agent string on its own can be copied by anyone, as the user agent spoofing entry explains.
AI-assisted people inside your user counts. In GA4, a person drafting with an assistant is simply a user. Proof of Actor reports that person as Human + AI when the assistance is declared or observed, so you can compare their conversion path with unassisted users. The AI-assisted interaction entry explains the class.
Unknown as its own line. Instead of counting unclear traffic as a user or dropping it, Proof of Actor returns Unknown and tracks the Unknown rate by route.
A decision, not only a report. GA4 reports; it does not change what a visitor experiences. Proof of Actor lets your code allow, adapt, challenge, send to review or rate-limit per route, class and basis. It starts in observation mode, so installing it changes nothing until you choose to enforce a policy.
Adding actor classes to an existing GA4 setup
A practical rollout keeps every GA4 report your team already reads:
- Leave GA4 as it is. Acquisition, campaign and conversion reporting stay in place, and known-bot exclusion keeps those numbers focused on people.
- Install Proof of Actor in observation mode at the server, edge, API or MCP layer, so it classifies interactions GA4 never receives as well as those it does.
- Register an actor class dimension. Send the class and basis as event parameters on the events you already track and register them as custom dimensions, then add the dimension to the reports that matter, such as sign up and checkout funnels.
- Compare the two views. GA4 shows how classified browser traffic converts. Proof of Actor actor analytics shows the full mix by route, including agent traffic that GA4 excludes or never sees.
- Act on the routes that matter. Once the picture is clear, write policies in your code for the workflows where actor class should change the experience.
The AI agent traffic analytics page covers the measurement side in more depth, and the websites use case shows how this looks on a production site.
When GA4 is enough, and when to add Proof of Actor
GA4 is enough when your question is about marketing: which channels bring visitors, how campaigns perform, which pages engage people, and what converts. Known-bot exclusion is a feature for that job.
Add Proof of Actor when GA4's numbers raise questions it cannot answer: how much of your traffic is agents acting for customers, whether a jump in sign ups came from AI-assisted people, why API usage grew without matching sessions, or what your application should do differently for each actor. GA4 tells you how your audience behaves in aggregate. Proof of Actor tells you who is acting on each interaction, with evidence, and gives your code a way to respond.
Get started
Proof of Actor is in private beta. Through the founding developer program, selected teams can receive up to 12 months of free access within agreed usage limits, with no automatic billing. Apply for early access and tell us about your site or application and the traffic you want to understand.
Google Analytics is a trademark of Google LLC.