I'm Gabriel Landaeta Kranwinkel, a computer engineer who spent a decade in digital marketing before coming back to the technical side. For the last 5+ years I've worked with measurement setups end-to-end: from the data layer in the browser, through GA4 and server-side GTM, to the BigQuery pipelines analysts can actually trust.
Your GA4 numbers don't match Google Ads, Meta, or your CRM. You've stopped trusting any of them.
Your BigQuery bill keeps climbing, and nobody wants to touch the queries anymore.
Your server-side or consent setup is a black box. Something in it broke a few weeks ago, and nobody can figure out what..
I come from computer science, so the technical layer is home turf: JavaScript, data layers, network inspection, server-side logic. But a decade running marketing and analytics teams taught me to translate it. To turn "what does the business really want to measure?" into a tracking setup that survives contact with reality.
Today I work inside the measurement team of a heavily regulated bank. Every tag, every export has to be correct and defensible. Same standard for smaller projects: clean setups that hold long-term.
Happy taking a project from scratch, or dropping in to audit and fix what's already there. Based in Berlin, working remote with clients across Germany and abroad.
I take care of the GA4 infrastructure for a heavily regulated banking environment. That runs from data layer design through GTM implementation to the Dataform pipeline I architected: the one that turns nested BigQuery exports into flat, sessionized tables the whole team can query. And I run the server-side GTM setup that lets marketing teams deploy Google Ads, Meta, and LinkedIn tags inside the bank's privacy constraints.
Owned the tracking tech and infrastructure for the whole group's sites and apps on GA360. Worked on the group-wide GA3 to GA4 migration: 3,000+ unique events across 30+ websites and 12 mobile apps. Also owned server-side tagging, Meta CAPI and TikTok Events API integrations, and the JavaScript templates the client- and server-side GTM setups ran on.
Ran the corporate Data-Driven Marketing project across six markets: Austria, Switzerland, Poland, Czech Republic, Hungary and Slovakia. Mostly programmatic, search, and social spend, plus measurement frameworks built with the local teams and agencies. Before that, launched the digital projects behind renault.es and dacia.es.
When the team shifted from the GA interface and Data Studio to BigQuery, the data showed up but the strategy didn't. Analysts wrote complex queries straight off the raw event-level exports, and costs grew every month. The same metric meant different things on different dashboards, depending on who wrote the query. I came in to fix that: first the data strategy and shared definitions, then the technical layer to back them up. The Dataform pipeline now sits between the raw exports and everyone who uses them.
The team had moved from the GA interface to raw BigQuery exports without a data strategy. Analysts hit the firehose directly with complex, expensive queries. Costs grew. The same metric meant different things to different people.
I defined the data model and shared metric definitions first, then architected the Dataform pipeline to back them. It takes the daily BigQuery exports and produces documented, analysis-ready tables. Nobody has to unnest anything by hand anymore. Currently migrating it to dbt.
Monthly BigQuery costs dropped by roughly 80%. Queries got faster and simpler. And the less measurable win: analysts spend their time on analysis instead of wrestling with raw GA4 schemas.
for a website or app: data layer, GA4, and client- and server-side GTM. Built around what the business needs to measure.
A health check of your GA4, GTM, and data layer: what's broken, what's double-counting, what's leaking. With a prioritized fix list at the end.
for Google Ads, Meta CAPI, TikTok Events API, LinkedIn CAPI, Pinterest, and others. Consent-aware and deduplicated across channels, feeding ad platforms signals they can bid on.
Raw GA4 exports become clean, modeled tables that analysts and dashboards can rely on. Cost-efficient because nobody has to re-run expensive unnesting logic on every query.
CMP setup (Usercentrics, OneTrust, CookieBot) and Consent Mode configuration. GDPR-compliant measurement without the guesswork.
Two flavors: replatforming a website without losing tracking continuity, or moving from one analytics or tag-management setup to another. Planned so the data history stays intact.
Have a setup that needs building or auditing? Tell me a bit about it and I'll get back to you.