Audit the current setup
Inventory what's measured, what's broken, what's missing.
- Existing GA4 + GTM review
- Consent posture check
- Data-loss estimate
- Reporting needs interviews
GA4, GTM, Consent Mode v2 and Looker Studio dashboards, set up so marketing can prove ROI — not vanity metrics. Server-side tagging where it matters. GDPR-clean by default. Documentation your team can actually maintain.
Measurement looks easy until it isn't. The setups that need this work tend to share patterns:
A measurement stack your team can maintain, documented so the next person inherits a system rather than a black box.
Events, properties, audiences and data model — designed against your pipeline, not against GA4's defaults. The reference document everything else flows from.
Containers built, events fired, custom dimensions populated, audiences defined. Naming conventions documented so additions later don't break the model.
Banner integration, advanced consent signals, GDPR-clean defaults — set up so the consented and non-consented paths both work and the data tells the truth either way.
Two layers: an executive view (one screen, weekly cadence, the metrics leadership tracks) and an operator view (segmentation, channel detail, the diagnostic tools marketing actually uses).
Where the data loss from adblockers and ITP is material, server-side GTM goes in — with the right balance between cost, complexity and recovery.
The handover artefact: how the stack works, how to add events, how to debug, what to do when GA4 changes again. Plus a working session with your marketing team.
Three phases, sequenced so each one informs the next. Total elapsed time is typically three to four weeks depending on existing setup.
Inventory what's measured, what's broken, what's missing.
Build the measurement plan, then implement it.
Confirm the data tells the truth, hand over.
Measurement work touches marketing, engineering and legal — the engagement is shaped to slot in next to all three without becoming a single point of failure.
The build creates a stack they can run. Documentation is written for them. The dashboards answer the questions they actually need to answer. I leave; the system stays maintainable.
If we go server-side, your engineering team is in the loop for the deployment piece — typically a few hours of their time, with a runbook so it can be re-deployed without me.
Consent Mode v2 and GDPR are legal-adjacent decisions, not just technical ones. I'll surface the choices that need your legal or DPO sign-off rather than making them for you.
Progress updates land in your shared workspace. One scheduled mid-engagement sync, the rest async. Decisions are documented as they're made.
Tracking doesn't generate traffic, conversions or pipeline. What it does is make every other channel's contribution legible — so decisions about where to invest stop being arguments about whose dashboard is right.
For SEO and GEO specifically, the difference is sharp: measurement infrastructure is what turns "the content is working" into "the content drove €X in pipeline last quarter and these eight pages did the heavy lifting." That kind of attribution is what unlocks continued investment from leadership.
If yours isn't here, drop a line — I'll answer directly and add it.
Three to four weeks elapsed time, with the variance driven by what already exists. A clean GA4 + GTM that just needs a measurement plan and dashboards lands faster than a setup that needs to be ripped out and rebuilt.
If you serve any EEA users and run any Google ads or analytics, yes — it’s been required since March 2024 and the advanced version is what unlocks the modelled data Google now uses to fill the consent gap. Without it, your reporting is increasingly under-counting and your Ads bidder is operating with degraded signals.
Not always. The decision depends on how much data you’re losing to adblockers and ITP, how much you spend on paid acquisition (which compounds the cost of bad signals), and whether you have the budget for a small ongoing GCP / Cloudflare bill. The audit phase produces an honest estimate of the data loss before we commit either way. Server-side is a tool, not a requirement.
Yes — and that’s the default. Wholesale rip-and-replace is rarely necessary. The build phase usually extends and tightens what’s there rather than starting from scratch.
The Consent Mode v2 + modelled-data setup is most of the answer. Beyond that, the measurement plan is designed against your actual customer journey — server-side identifiers, CRM integration where useful — so the stack degrades gracefully as third-party signals continue to weaken.
Yes. GA4 + GTM is the most common stack for SaaS / tech marketing teams, but if you’re already on product-analytics tooling (Mixpanel, Amplitude, PostHog) the work integrates with it rather than displacing it — usually with GA4 owning marketing-funnel measurement and the product tool owning in-product behaviour.
The build is designed so you don’t need ongoing support — the documentation and training session are part of the deliverable for exactly that reason. If you want a quarterly check-in or ad-hoc help when GA4 ships its next breaking change, that’s available as a small retainer.
Get in touch with a short description of your current setup — GA4 in place, GTM container live, what's broken, what you're trying to prove — and I'll come back within 24 hours with whether this is the right engagement and what the next step looks like.
Choose which categories you allow. Essential cookies are always on because the site won't work without them. You can change this any time from the "Cookie preferences" link in the footer.
Required for the site to function — navigation, form submission, security.
Anonymous usage data via Google Analytics 4 — pages viewed, time on site, traffic source. Used to improve the site.
Cookies that enable ad measurement and remarketing. Off by default; only relevant if Google Ads or similar campaigns are running.
Used to remember preferences like language or display mode across visits.