When this is the right call.
- $5M+ ad-spend brands where spreadsheet attribution has hit its ceiling.
- Performance agencies needing white-label warehouse-grade attribution for client books.
- Series A+ B2B SaaS blending first-touch, multi-touch, and self-attribution survey data.
- Teams with a data engineer (or willingness to retain one) post-handover.
When you're better off elsewhere.
- Brands wanting a pre-packaged dashboard product - buy Triple Whale or Northbeam instead.
- Companies under $1M/yr ad spend - incremental insight doesn't justify warehouse build.
- Teams without engineering bandwidth to own the system after the 60-day support window.
A working system in weeks - not a slide deck in months.
- 01
Audit · Week 1
Source-of-truth mapping, channel taxonomy, success metric definitions.
- 02
Build · Weeks 2–4
BigQuery schema, dbt models, ingestion pipelines, dashboard layer.
- 03
Validate · Week 5
Reconcile against platform numbers, MMM model fit, dashboard QA.
- 04
Train · Week 6
Workshops, documentation, runbook handover.
Every line. On the box.
Productised means published. No hidden scope, no surprise invoices, no scope creep dressed as a change order.
- BigQuery dataset architecture
- dbt models for events + orders + ad spend
- Ingestion: Meta, Google, LinkedIn, TikTok, Klaviyo, Shopify
- Last-touch + first-touch + linear + position-based
- MMM-lite (regression-based) for spend optimisation
- Incrementality test framework
- Looker Studio dashboard (or Looker / Metabase)
- Daily refresh + monitoring
- Self-serve channel filters
- Documentation + SQL examples
- Stakeholder training (3 hours)
- 60-day post-launch support
Three tiers. Transparent.
Ideal: Foundational attribution stack
- Last-touch + GA4
- 3 channels
- 1 dashboard
- Documentation
Ideal: Most $5M+ ad-spend brands
- MTA + BQ + Looker
- All channels
- Incrementality framework
- Training
Ideal: Multi-brand groups, mature data orgs
- MMM + incrementality + LTV
- Multi-brand
- Custom integrations
- Quarterly optimisation
"We finally trust our ROAS numbers. The MMM-lite layer told us we'd been over-investing in last-touch Google by 28%. [PLACEHOLDER]"
Common questions.
Often paired with this.
Server-Side GTM + CAPI Sprint
Cookieless-resilient measurement on Google Cloud, with Meta + TikTok + LinkedIn CAPI, dedupe, and Consent Mode v2.
MarTech Audit & Migration
Full audit of your 20+ tools, TCO model, migration plan, and executed cutover with zero data loss.
Receipts.
Ready to scope Attribution System Build?
Book a 20-minute call. Bring one bottleneck. We'll diagnose live. Starts at $7,500 fixed.