The data infrastructure behind every serious marketing decision.
Server-side tracking, GA4 enterprise, MCP data lake, attribution modelling, consent governance. The technical foundation on which the rest of the Fortop system rests — including BIKMA. We work with the discipline of an engineering practice, not the approximation of a reporting exercise.
Over the last seven years data collection has been rewritten from the ground up. GDPR and ePrivacy made consent a technical condition, not a disclaimer. Apple ITP, iOS 14 and the progressive deprecation of third-party cookies have halved the ability to track client-side. Ad platforms have reacted with opaque statistical models that infer what is no longer measured.
Net result: in many sectors client-side data accuracy has fallen to 60–70% (Fortop audits 2023-2025) without marketing noticing.
In parallel, marketing decisions have become increasingly automated. Bid management, audience selection, content personalisation, attribution on which the allocation of millions of euros in budget is decided — all driven by models that lean on the data we feed them. If the data is wrong, the AI takes wrong decisions very quickly.
Most agencies offer reports. We offer the data infrastructure beneath the reports — because a report on wrong data produces wrong decisions, whatever the graphics look like.
Design of the company’s data architecture: what we collect, where we store it, who uses it, how we protect it. Data models, interface contracts, lifecycle governance. It is the depth that separates a real data capability from a Google Analytics setup.
Implementation of Google Tag Manager Server-Side, GA4 enterprise with BigQuery export enabled, enterprise dataLayer, enterprise management of events and custom dimensions. Server-side first since 2019, before it became the standard.
Consent Mode v2, integration with CMPs (iubenda, OneTrust, Didomi, Cookiebot), GDPR- and ePrivacy-compliant tracking, consent decay management, operational integration with our platform GDPR Daily for continuous validation.
Data-driven attribution models, multi-touch attribution, marketing mix modelling for enterprise multi-channel clients, integration with CDP and CRM. We help you stop believing in last-click.
Company data lake consolidating marketing, sales, CRM, product, customer service data. Exposure via Model Context Protocol for integration with proprietary and market AI tooling. This is the layer where BIKMA hooks into traffic and conversion data.
Operational dashboards on Looker Studio, Power BI. Automated reporting, anomaly alerting systems, KPI scorecards for marketing leadership. The output the C-level sees, built on top of the five layers below.
We map what’s already there: tracking implemented, events that work and don’t work, gaps against declared KPIs, quality of consent, GDPR posture. It is the only phase in which we deliver “bad news”: we need a true picture to be able to design a new one.
We design the target data architecture: what needs to be collected, from where, where to store, how to transform, how to distribute. Output: a technical spec shared with the client’s internal team, which can also be implemented internally if the client prefers.
Technical execution: server-side tagging, enterprise events, CMP integration, data lake build, transformation pipelines (dbt or equivalent), operating dashboards. We work with the client’s IT teams, not in their place.
End-to-end testing of every event, accuracy validation, continuous monitoring of data quality. This is the step most implementers do not invest in — and it is the one that separates a setup that lasts three years from one that breaks after six months.
The data lake feeds BIKMA, ad platforms, personalisation systems, product analytics tools. Quarterly governance to manage evolution: new events, new integrations, new needs.
| KPI | What it measures |
|---|---|
| Data Accuracy Score | How much the data collected corresponds to what actually happened |
| Event Coverage | % of relevant interactions tracked end-to-end |
| Consent Rate | % of users giving consent to analytics tracking |
| Cross-domain Attribution Coverage | % of journeys reconstructed completely across different domains |
| Time to Insight | Average time between event and data availability in the dashboard |
| Data Lake Completeness | % of enterprise sources actually integrated into the data lake |
60-70%
Initial Data Accuracy
> 95%
Post-implementation Data Accuracy
measured on Fortop audits 2023-2025
70-80%
Optimised Consent Rate
> 90%
Cross-domain Coverage
Reduced Time to Insight from days to hours thanks to the data lake and automated pipelines.
Orders of magnitude, not guarantees. The starting point weighs.
Google Tag Manager + GTM Server-Side · GA4 360 enterprise · BigQuery, Snowflake · dbt for data transformation · Looker Studio, Power BI for reporting · iubenda, OneTrust, Didomi, Cookiebot for CMP · Segment, mParticle for CDP · Fortop Data Lake exposed via MCP · direct integrations with enterprise CRMs (Salesforce, HubSpot, Dynamics) and ad platforms (Google Ads, Meta, LinkedIn, TikTok).
We are the right partner if your company moves significant marketing budgets and those budgets are decided — even partly — by looking at dashboards.
We are the right partner if you operate in a regulated sector where consent management and GDPR compliance are critical, not a tickbox.
We are the right partner if you are in a data-layer consolidation phase (post GA4 migration, post-cookie, pre-AI marketing activation) and need a partner who works with the discipline of an engineering practice.
We are the right partner if you want a data architecture designed to last five years, not just to close the next quarter.
We are not the right partner if you are looking only for afast basic GA4 implementation: there are freelancers and specialised agencies that do that work better than us at lower prices.
We are not the right partner if the priority is the dashboard with the chart, and not the data underneath it: we work poorly in that frame.
We are not the right partner if the client’s IT is structurally closed to collaborations with external consultants: our model presupposes operational integration with internal teams.