Framework · Marketing Data Strategy

Most marketing teams aren't short on data. They're short on the right level of capability.

A five-rung maturity ladder for marketing data — and why most teams are trying to climb rung 4 before they've stabilised rung 2.

Role Framework developed from 12 years working across analytics, product, and strategy roles

5 maturity levels 4 jump types: tech, collaboration, trust, infrastructure 12 years across analytics, product & strategy
01 — The Ladder

Five levels, each with its own job to finish

Each level solves a specific problem. Skipping one doesn't make a team faster — it makes every level above it unreliable.

1
Get the basics right Foundation

Fix the foundation before anything else — if it's lying to you, everything built on top of it is too.

  • Server-side tracking, so signals survive cookie blocking
  • Consent management that actually works — not a banner nobody reads
  • Clean event data landing in the warehouse with proper attribution hygiene
2
Know your audiences Solid ground

This is where most teams live. It's solid ground — it's just not the finish line.

  • First-party audience segments flowing into Braze, Meta, and Google
  • A/B testing with real holdout groups — not just creative variants
  • A next-best-action model that matches the right product to the right user
3
Know your customer's future Mindset shift

You stop optimising for the cheapest click and start optimising for the most valuable customer. Harder to build — worth every bit of the effort.

  • LTV models embedded directly into campaign bidding — not sitting in a notebook
  • Lookalike modelling built on your best customers, not your most recent ones
  • Probabilistic identity resolution stitching together web, app, and CRM signals
4
Prove marketing is working Evidence over projections

Marketing stops being a cost centre and starts walking into budget conversations with evidence. Most CMOs want to be here — fewer actually are.

  • Marketing Mix Modelling that decomposes revenue across channels, seasonality, and external factors
  • Incrementality tests running continuously, not just once a year
  • Multi-touch attribution that stops crediting the last click for everything
5
Let the machines run (some of) it Not fully here, industry-wide

The infrastructure decisions made at Levels 1–4 determine whether this level is ever realistic.

  • A semantic data layer so AI agents have something reliable to reason over
  • Agents that monitor performance, flag upstream data issues, and rebalance budgets within guardrails
  • Reinforcement learning optimising bid decisions on outcome signals, not just click-through rates
02 — The Framework

From data hygiene to AI autonomy

The same five levels, mapped against business impact and core capabilities.

The Marketing Maturity Framework: From Data Hygiene to AI Autonomy — five stages from real-time signal activation to AI-driven autonomous agents
03 — What Each Jump Actually Requires

The bottleneck changes at every level

The teams that move fastest aren't the ones with the biggest budgets — they're the ones who resist the temptation to skip levels.

1 → 2

A tech problem

Tracking, consent, and clean pipelines.

2 → 3

A collaboration problem

Data science and marketing actually working together — not just handing files back and forth.

3 → 4

A trust problem

Getting finance and marketing to agree on what proof of growth actually looks like.

4 → 5

An infrastructure & patience problem

The groundwork from Levels 1–4 determines whether autonomy is ever realistic.

Let's talk

Tackling something like this?

Tell me your biggest data or AI bottleneck and I'll send you how I'd approach it — the same way I worked this problem. No résumé dump, no pitch. Just a useful conversation.

Start the conversation →