Why SealMetrics exists
Most enterprise eCommerce companies in Europe optimize advertising investment on top of 13% of their traffic. This is what happens when that gets normalized.
Most enterprise eCommerce companies in Europe optimize their advertising investment on top of 13% of their traffic. This chapter is about what happens when that number stops being a measurement problem and becomes a decision problem.
It's written for CMOs, growth leads, and measurement owners — the people who sign off on six or seven-figure media budgets and have to defend the result.
The problem we see
Not 13% of some edge case. Thirteen percent of total — after GA4 consent banners are rejected, ad blockers strip events, and Intelligent Tracking Prevention purges cross-site cookies on Safari and Firefox. Across our tests on European enterprise sites, that number holds with small variance.
Thirteen percent is what reaches the dashboard. Decisions are made anyway — because the dashboard is the only thing on the table.
"The dashboard looks complete. That's the part that should worry you."
What breaks in decision-making
Bidding algorithms are statistical models. When you feed them sampled, modeled, or partially observed conversion data, they don't refuse — they extrapolate. The output looks like a campaign performance report. It isn't. It's a confident extrapolation from a biased sample.
A 10M€ eCommerce running on 13% data is not allocating 87% of its budget against ghosts. It is allocating 100% of its budget on a phantom of the truth. Every bid, every audience, every retargeting list is a guess dressed as a measurement.
The cost compounds. Budgets shift toward channels that appear to convert — because those channels are measured better, not because they perform better. The campaigns that actually drove revenue go unmeasured and unfunded. Over twelve months, the marketing plan rewrites itself around what GA4 can see, not around what your customers actually do.
Our position
SealMetrics is built on three premises. They are constraints, not features — they define what we will and will not build.
Data should be complete
Captured server-side, before ad blockers see the request and before consent banners are decided. Aggregate, anonymous, never personal. The full population, not a sample of the consenting majority.
Attribution should be honest
Last-click on 100% of events. Not because last-click is sophisticated — it isn't. Because complete last-click outperforms modeled multi-touch every time the budget moves. We're explicit about this in chapter 09.
Compliance should be architectural
No cookies, no localStorage, no fingerprinting. Hosted in Dublin. GDPR and ePrivacy compliant by design — not by legal interpretation. The legal grounding lives at /security while we finish the dedicated Open chapter on compliance.
Who we don't serve
We are not a GA4 replacement for a 200K€ DTC startup. The product fit starts roughly at 10M€ in annual eCommerce revenue, where the cost of bad investment decisions exceeds the cost of measuring properly. Below that, GA4 plus a content management system usually works.
We are also not a privacy-first lightweight tool. Plausible, Fathom, and Simple Analytics are excellent at what they do — they are not what we compete with. We compete with GA360, Adobe Analytics, and Piwik PRO — at a fraction of the cost, with complete data, and without leaving the EU.
If your measurement problem is "I need a simple analytics dashboard", this is the wrong product. If your problem is "I'm spending real money on incomplete signal", the rest of this document explains why we exist and what we do about it.
Continue with chapter 02 — Who we build for.
Was this chapter useful? Tell us →