Business intelligence your decisions can rely on
The Single Source of Truth across marketing, CRM, and backend. No more CPA fairy tales from the ad networks. True customer acquisition cost, real LTV, AI-ready for 2026.
- BigQuery
- dbt
- Fivetran
- Airbyte
- Data (Looker) Studio
- Power BI
✓ EU region (BigQuery Frankfurt) · SCCs & DPA with Google Cloud
In short
We consolidate marketing spend, CRM stages, and real revenue into one source, in BigQuery, modelled in dbt, served to Data (Looker) Studio or Power BI. One CAC, one LTV, one calculation logic. No more three systems reporting three numbers for the same bank account.
- Who it's for
- E-commerce & B2B whose numbers get questioned internally
- What you get
- Marketing Data Hub, dbt models & executive dashboards
- Entry
- Audit from €2,900, Build in 3 to 8 weeks
01Quick self-check
You're in the right place if:
Tick what applies.
02How it works
Four phases, each with one clear job.
No one-tool Swiss-army-knife. Each phase uses the tool that fits its job.
- Fivetran · Airbyte
Ingestion
Pre-built connectors pull Stripe, HubSpot, Meta, Google Ads, and Shopify incrementally into the EU region.
- BigQuery EU
Storage
BigQuery in EU multi-region: storage and compute separated, pay-per-query, no clusters to size.
- dbt · marts
Transformation
dbt turns raw tables into tested models. When sales changes the CAC definition, it changes in one place.
- Looker Studio · Power BI
Output
Executive view on one page, operational views for marketing and sales, anomaly alerts in Slack.
03What we build
Was wir bauen
Meta reports €3.50 CAC. HubSpot says €47. Stripe shows €89 in actual acquisition cost. Three numbers, three systems, one bank account. This is not a reporting problem. It is an architecture problem.
Revenue Intelligence & Executive BI is the data architecture that consolidates those three numbers into one. Marketing spend from the ad platforms, lead stages from the CRM, revenue from the payment backend, all loaded into BigQuery, transformed in dbt, served to Data (Looker) Studio or Power BI. One source, one calculation logic, one dashboard per audience.
We don't build isolated dashboards, we build resilient ELT pipelines on which dashboards are the end product. The prerequisite is clean server-side data collection: loading marketing data into a warehouse fed by broken tracking only automates the error. See Measurement & Privacy Engineering.
04The difference
Excel silo vs. consolidated data hub.
Architecture: EU data residency in BigQuery Frankfurt, SCCs and DPA with Google Cloud. PII excludable or pseudonymised per pipeline.
05The building blocks
Two maturity stages. One foundation.
Tier 1 is the foundation Tier 2 doesn't work without. Predictive on inconsistent data automates hallucinations.
Tier 1 · Foundation
Clean marketing-to-CRM mapping, BigQuery as the central data foundation, dbt models for the core metrics. The attribution scaffold everything else builds on.
Concretely: Fivetran setup (Meta, Google Ads, HubSpot, GA4), BigQuery project in EU region with IAM and cost controls, dbt with staging/intermediate/mart, dashboards for marketing and leadership.
Tools: Fivetran · BigQuery · dbt · Looker Studio
Tier 2 · Executive BI & Predictive
Stripe and backend integration, LTV cohort analysis, anomaly detection, and forecasting at channel and cohort level.
Concretely: Stripe API in BigQuery, LTV models per acquisition cohort, anomaly alerts on ROAS drift, forecasting layer (Prophet or ML depending on data volume).
Tools: Stripe · BigQuery ML · Prophet · Slack
06In plain terms
The key terms, briefly explained.
CAC
Customer Acquisition Cost. The real, Stripe-reconcilable acquisition price, not the ad network's CPA claim.
LTV
Customer Lifetime Value per acquisition cohort, calculated from historical transaction data instead of estimated.
ELT over ETL
Load first, transform later: raw data lands in BigQuery, dbt models it there, versioned and tested.
dbt
data build tool. Enforces Git versioning, per-model tests, and a clean split between source data and business logic.
07Who it is for
When it pays off.
CMO / VP Marketing
Defend the marketing budget in front of the CFO. Need a report that shows CAC and LTV from Stripe data, not from Meta attribution. On one page.
Founder / Leadership
Want to know which channel brings actual customers, not which channel claims the conversion. The bridge between marketing spend and the bank account.
Head of Data / BI
Have BigQuery, need the marketing and CRM sources cleanly connected and modelled in dbt. Without custom Python that breaks on every Meta API update.
Performance teams (E-commerce)
Want ROAS, POAS, and real margin in one view. Not across three tools running three different attribution models.
08Deliverables
What you end up with.
Ingestion & storage
- Fivetran or Airbyte connectors for all relevant sources (ads, CRM, payment, shop)
- BigQuery project in EU region with IAM and cost controls
- Incremental loads, schema-drift handling, backfill mechanics
Transformation & models
- dbt project with staging, intermediate, and mart layers
- Core models: dim_customers, fct_orders, mart_cac_by_channel, mart_ltv_cohorts
- Data-quality tests per model (dbt tests, freshness checks)
Output & executive BI
- Executive dashboard (Data Studio, Power BI, or Tableau), one page, PDF-ready
- Operational views for marketing and sales
- LTV cohorts, anomaly alerts in Slack, forecasting layer (Tier 2)
Governance & handover
- KPI definitions document for internal alignment
- Documentation via dbt docs (what is calculated how, who owns it)
- Handover session plus 30-day post-launch support
What we do NOT do
Engagement depths
Three depths. Clear scopes.
No retainer trap.
Audit Sprint
We audit what is wrong. Prioritised report + action plan.
plus statutory VAT
Included
- Full analysis of the existing setup
- Prioritised report with concrete actions
- 90-minute walkthrough with your team
Not included
- Implementation (follows in the Build Sprint)
- Code in your app or website
When it fits
When the setup runs but the numbers are questioned internally.
Fixed price by scope (Audit Sprint / Audit Sprint Plus). 50% credited toward a Build Sprint if commissioned within 30 days.
Request an Audit Sprint →Build Sprint
Fresh build or restructure, built to spec.
plus statutory VAT
Included
- Marketing data hub (funnel.io or BigQuery)
- Attribution model + executive dashboard
- KPI definition document + handover
Not included
- Campaign execution, media buying, creative
- Tool licences (billed directly, no markup)
When it fits
When a clean rebuild beats patching in production.
Final fixed price after scope definition.
Discuss a Build Sprint →Managed Evolution
Ongoing partnership. Analytics as a product.
plus statutory VAT
Included
- Monthly development + roadmap
- QA on every release deploy
- Slack support, < 4 h response (Mon–Fri)
- Monthly report + executive summary
Not included
- 24/7 on-call rotation
- Campaign operations
When it fits
When analytics has to keep growing with you.
Monthly cancellation after the minimum term.
Request Managed Evolution →All prices net, plus statutory VAT. For companies in Germany, Austria and Switzerland.
Why dbt instead of plain SQL views in BigQuery?
dbt enforces three things raw views do not: Git versioning, automated tests per model, and a clean layer separation between source data and business logic. With seven sources feeding ten downstream metrics, the view-only approach breaks on the first schema change. A view that nobody tests is a bomb on a fuse. dbt fixes that in days, not in weeks of incident response.
How do you connect HubSpot data in a privacy-first way?
Through the official HubSpot API via Fivetran. Fivetran supports [EU multi-region](https://fivetran.com/docs/using-fivetran/fivetran-dashboard/account-settings/troubleshooting/use-fivetran-multiple-regions); the sync runs in EU-Frankfurt and the data stays in the EU region. PII fields (email, phone) can be excluded or pseudonymised per pipeline. With SCCs and a DPA in place, the connection stays audit-clean. An on-prem alternative via Airbyte Self-Hosted is available when the cloud vendor is contractually ruled out.
What does the architecture cost monthly to operate?
A typical mid-market configuration (5 sources, ~100M rows/month, 4 dbt models daily, 2 production dashboards) lands at €300–800 BigQuery, €200–500 Fivetran depending on connector mix, plus €50 dbt Cloud if used. Data (Looker) Studio is free; Power BI Pro is €10 per user. Operating cost scales with data volume, not with report count. Custom-ETL maintenance falls away.
Why BigQuery and not Snowflake or Redshift?
For DACH mid-market setups, BigQuery is the more pragmatic tool in most cases: no cluster sizing, EU multi-region out of the box, native GA4 integration via the export, transparent pay-per-query pricing. Snowflake wins for multi-cloud scenarios and very high concurrency. Redshift sits more naturally in AWS-centric stacks. We recommend the warehouse to fit the situation, not the preference.
Integrations
From the integrations catalog
These are the tools we wire up for this service. Each catalog page rates the tool honestly: setup effort, GDPR classification and when it fits.
Next step
Which CAC is right: Meta, HubSpot, or Stripe?
An Audit Sprint clarifies in 10 business days which number is wrong in which system and what a consolidated data foundation looks like. Prioritised report. 60-minute walkthrough call. No follow-on contract, no forced retainer.
- Entry
- €2,900–4,500 net
- Delivery
- 3–8 weeks
- Scope
- 5 modules