SaaS Development Company in Amsterdam

Adtech / Digital Media

An Amsterdam-based adtech company (Series A, €8 million raised) was building an analytics platform for European digital publishers — an alternative to Google Analytics that provides audience insights without relying on third-party cookies or transferring data to US servers (a concern amplified by Schrems II and the AP's enforcement of data transfer rules). The prototype worked but faced: accuracy limitations (cookie-less analytics using IP-based and fingerprint-based tracking was only 72% accurate for user identification — publishers needed higher accuracy for audience measurement that advertisers would trust), no real-time capabilities (the prototype processed data in daily batches — publishers needed real-time dashboards for breaking news, campaign monitoring, and editorial decisions), limited EU deployment (data was processed on a single AWS region — enterprise publishers required data residency guarantees in specific EU countries), and no consent integration (the prototype ignored consent signals — in the post-GDPR EU market, analytics must respect TCF consent and process only consented data). The company needed 40 publisher clients within 12 months to demonstrate product-market fit for a Series B.

We rebuilt the platform as a privacy-first, real-time analytics SaaS with EU data residency and consent-native architecture. (1) Privacy-preserving identification: replacing cookies and fingerprinting with a privacy-preserving approach. First-party data: publishers embed a JavaScript snippet that generates a first-party identifier (stored in the publisher's domain, not a third-party cookie). This identifier persists across sessions on the same publisher site without cross-site tracking. Probabilistic modelling: for users who reject all tracking (approximately 30% in the EU), the system uses statistical modelling based on aggregated, anonymised traffic patterns — providing directionally accurate audience estimates without identifying individuals. Consent-native: the system checks TCF consent signals before processing any data. For users who consent to analytics (TCF Purpose 8 — Measure content performance): full analytics with first-party identifier. For users who consent to limited purposes: aggregated-only analytics. For users who reject all: no data processing, with probabilistic modelling filling the gap. Result: 91% user identification accuracy (from 72%) while fully respecting consent choices. (2) Real-time streaming architecture: replacing daily batch with real-time. Event ingestion via Kafka (processing 2.8 billion events daily across all publishers). Real-time aggregation using Apache Flink — computing audience metrics (unique visitors, page views, engagement time, scroll depth) with sub-second latency. Dashboard delivery via WebSocket — publisher editors see real-time audience metrics for every article, enabling editorial decisions based on live reader behaviour. Historical analytics: daily aggregation into ClickHouse for historical queries spanning months of data at sub-second query speed. (3) EU multi-region deployment: data residency in multiple EU countries. Primary: AWS eu-west-1 (Ireland) — the default for most publishers. Options: eu-central-1 (Frankfurt) for German publishers with strict data residency requirements, eu-south-1 (Milan) for Italian publishers. Per-publisher data isolation: each publisher's data is stored and processed entirely within their chosen region — no cross-border data transfer. EU-only processing: all staff access from EU locations, all data sub-processors EU-based — creating a compliance posture that withstands AP scrutiny. (4) Publisher intelligence: beyond basic analytics. Content performance: which articles drive subscriptions, which drive ad revenue, and which drive engagement — helping publishers optimise their editorial strategy. Audience segmentation: first-party audience segments based on reading behaviour — "tech enthusiasts who read 5+ articles/week" — that publishers can use for advertising targeting without third-party data. Subscription propensity: predicting which readers are most likely to subscribe, enabling targeted paywall and subscription offers. Competitor benchmarking: anonymised, aggregated benchmarks — "your engagement is 15% above the median for news publishers your size" — enabling publishers to understand their performance in context.

Publisher adoption: 52 publisher clients in 12 months (30% above target), including 3 of the top 10 European news publishers. Combined monthly audience: 680 million unique users across all publishers. Total ARR: €2.1M. Accuracy: 91% user identification (from 72% — 26% improvement) while fully respecting consent. Publishers report that advertisers accept the audience data for campaign planning — the accuracy is sufficient for commercial use. Real-time: sub-second dashboard updates during breaking news events. One major publisher reported that real-time analytics changed their editorial workflow — editors now monitor article performance in real-time and adjust homepage placement based on reader engagement, increasing overall engagement 18%. Privacy compliance: zero AP complaints across all publishers. The consent-native architecture means the platform only processes data it has consent for — eliminating the compliance risk that drove publishers away from Google Analytics. Two publishers specifically cited "AP-proof analytics" as the reason for choosing the platform. Data residency: 34 publishers chose Ireland, 14 chose Frankfurt, 4 chose Milan. German publishers (Axel Springer, FUNKE Mediengruppe) required Frankfurt data residency — the multi-region architecture made this possible without a separate deployment. Subscription propensity: publishers using the subscription prediction model report 22% improvement in subscription conversion rates — the model identifies readers at the optimal moment for a paywall trigger. Series B: the 52 publishers, €2.1M ARR, and privacy-first positioning contributed to a €28 million Series B. The lead investor's thesis centred on "the only analytics platform purpose-built for EU privacy regulation." Development investment: €390,000 over 16 weeks.