Kitbase vs Google Analytics: GA4 and the Traffic It Cannot See
Kitbase vs Google Analytics 4 compared on cookieless tracking, consent banners, AI crawler visibility and AI referral attribution.
GA4 is free, ubiquitous, and structurally blind to two things that matter more each quarter: the AI crawlers reading your site, and much of the traffic AI answers send you. Kitbase costs money and covers both, alongside cookieless web analytics.
We build Kitbase. GA4 is genuinely good at several things and we’ll name them.
Where GA4 still wins
Price at scale. Free at volumes that would cost real money elsewhere. For a high-traffic site with modest analytical needs, nothing competes.
Google Ads integration. Native conversion import and audience sync. If your growth runs on Google Ads, this is difficult to replicate and often decisive on its own.
BigQuery export. Raw event data into a warehouse, which unlocks unlimited custom analysis for teams with the skills to use it.
Ecosystem. Nearly every analyst knows GA. Hiring, agency handover and documentation are all easier.
If those are load-bearing for you, GA4 stays. This isn’t a comparison where one tool wins outright.
Where GA4 is structurally blind
AI crawlers. GA4 is a JavaScript tag. GPTBot, ClaudeBot, Google-Extended, OAI-SearchBot and PerplexityBot don’t execute JavaScript, so they never fire the tag and never appear in any report. Not filtered out, not sampled away, simply absent. Since AI answers are grounded in pages the engines fetched, “which of our pages do AI crawlers read” is now a real question with no answer in GA4.
Kitbase reads requests forwarded from your server or edge, resolves each to a named crawler and vendor, verifies the claim against that crawler’s published identity, and reports per-path crawl frequency.
Much of AI referral traffic. Some AI products send a usable referrer, some don’t, and some traffic arrives with none at all because the user copied a link out of a chat. GA4 files a good deal of that under direct. Kitbase has the same underlying problem, since it’s a property of how the traffic arrives rather than a tool limitation, but it pairs referral data with crawler and citation data so you can reason about the channel rather than guess. We covered the mechanics in AI dark traffic and attribution.
Whether AI engines mention you at all. Not analytics in the traditional sense, and it’s the question determining whether that traffic exists. Kitbase tracks up to ten answer surfaces.
The privacy difference
GA4 sets cookies and processes personal data, which under GDPR and similar regimes means a consent banner, and every visitor who declines is a visitor you don’t measure. Several European data-protection authorities have scrutinised transatlantic data transfers in Google Analytics specifically.
Kitbase is cookieless by default, using a server-side anonymous identifier, so anonymous analytics needs no consent banner. Identity is an explicit opt-in rather than the baseline. We wrote about the approach in cookieless analytics.
That’s a real difference in measured traffic, not just compliance paperwork. Sites with high decline rates are analysing a biased subset without realising it.
Side by side
| GA4 | Kitbase | |
|---|---|---|
| Price | Free | From $99/mo |
| Cookies and consent banner | Cookies, banner required | Cookieless, no banner for anonymous |
| Events, funnels, retention | Yes, complex | Yes |
| Frustration signals (rage, dead clicks) | No | Yes |
| Google Ads integration | Native | No |
| BigQuery export | Yes | REST API, webhooks, exports |
| AI crawler detection | No, structurally | Yes, with identity verification |
| AI visibility in answers | No | Up to 10 surfaces |
| Sampling and thresholding | Yes | No |
| Learning curve | Steep | Lighter |
Which to run
Most sites should keep GA4 if Google Ads matters or the free price at scale is load-bearing. Adding a second tool for the AI layer is cheaper than replacing your analytics.
Replace GA4 when the consent banner is costing you meaningful measurement, when the complexity means nobody actually uses it, or when data residency is a requirement rather than a preference. Our Google Analytics alternatives guide covers the wider field including Plausible, Fathom, Matomo and Umami.
Either way, the AI crawler gap doesn’t close by changing analytics vendors. Plausible, Fathom and Umami are all JavaScript tags too, and none of them see AI crawlers either. That needs server-side collection, which is a different kind of integration.
FAQ
Can GA4 track AI crawlers like GPTBot? No. It’s a JavaScript tag and AI crawlers don’t execute JavaScript, so they never appear in any GA4 report.
Does GA4 show ChatGPT referral traffic? Partially. Some AI referrals carry a usable referrer and some arrive with none, landing in direct. This affects every analytics tool.
Is Kitbase GDPR-friendly without a banner? It’s cookieless by default with a server-side anonymous ID, so anonymous tracking doesn’t require a consent banner. Confirm against your own legal requirements.
Can Kitbase replace GA4 entirely? For web and product analytics, usually. You’d lose Google Ads integration and BigQuery export. It doesn’t do session replay.
Do I lose historical data if I switch? Your GA history stays in GA. Run both in parallel for a few weeks to build a baseline before relying on the new one.
Want cookieless analytics plus the AI crawler and citation data GA4 can’t see? Start your free trial — 7 days, no credit card required.