Skip to content
home

Marketing measurement consultancy

Measurement your board will trust

We build the data foundations, measure what your marketing really delivers, and test what we can't model, so budget decisions rest on evidence rather than attribution.

What we do

From messy data to measured decisions

Four practice areas, one aim: marketing decisions you can defend.

  • Marketing mix modelling

    Bayesian MMM that separates what marketing caused from what it coincided with, with budget recommendations your finance team can audit.

    Explore MMM
  • Data and analytics engineering

    The weekly pipeline, dbt models, control variables and quality checks that measurement depends on.

    Explore data engineering
  • Analytics implementation

    GA4, server-side tagging and consent mode set up so the numbers going in are right.

    Explore implementation
  • Conversion optimisation

    Research-led experiments on your site and funnels, designed to be read out honestly.

    Explore conversion optimisation

How we work

A consultative engagement, not a black box

  1. Discovery call

    Thirty minutes to understand the decision you need to make.

  2. Diagnostic

    We assess your data and measurement against what the decision needs.

  3. Delivery

    We build, model, test or review, working in your tools.

  4. Handover

    You own the pipeline, the method and the documentation.

Our platform

We're building CausalPanda for our own work

CausalPanda is our measurement platform, and it is in the final stages of building. We're making it so the models, tests and reports we deliver can stay with you after we leave. We're inviting a small number of teams to try it early and help shape it.

  • Early access for a small number of teams, in return for feedback
  • Built so your team keeps access after an engagement
  • Partner access you grant and can revoke
Illustrative: the Causal Readiness Score: whether this data and design can support a causal claim, before any model runs.68/100Causal readinessModel with caveatsHistory length92156 weeks, above the 104-week minimum.Spend variation48Radio and print barely vary.Control coverage55No price or competitor variable supplied.Experimental evidence30No completed test on this snapshot.

Agency? Deliver measurement under your own brand

White-label MMM, data engineering, implementation and testing.

FAQ

Before we talk

No. We work in your tools. CausalPanda, our platform, is in the final stages of building, and it will never be a condition of working with us.

Newsletter

Get new articles by email

Let's talk about the decision you need to make

Thirty minutes, no slides.