Common questions.

1. No. 01

   Will you build a Marketing Mix Model for me?

   We can, though it is not the core of what we do. Our focus is helping you modify a model you already have, or working with your analysts – internal or external – to get the best model possible, talking them through the process as it happens. A common scenario: an in-house data science team is about to build models but has never done Marketing Mix Modelling (MMM) before. You do not necessarily need to hire an MMM specialist to build them – the team works with us to learn the full MMM process and avoid the common pitfalls, and the capability stays in-house.

2. No. 02

   Can you help me find an MMM provider?

   Yes. We know the MMM vendors across the UK and internationally. After a short call about what you need, the timeline and the budget, we will suggest the providers whose strengths fit your situation. No referral fees, no conflicts of interest: recommendations are made on suitability alone, and we say so in writing at the start of every engagement.

3. No. 03

   Do you work with internal or external analysts?

   Both. Running MMM with an internal team is a particular kind of work; running it with an external vendor is a different one. We have done plenty of each. The principle is the same either way – we help both sides ask the right questions of the model and the data, and we hold the result to the same standard.

4. No. 04

   Which markets do you cover?

   We are based in London. Our team and our trusted partners have worked across Europe, North America and parts of Asia. Most engagements run in English, though we have worked across multiple languages alongside local partners. We are happy to support work in markets where we do not have a local presence by partnering with a team that does.

5. No. 05

   My MMM vendor will not share their model with me. Can we still work together?

   Yes, though it is not ideal. We are believers in transparency on principle, and we are sceptical of proprietary methodologies vendors will not show. That said, we know many of these vendors and can help you ask the right questions – questions that respect the chosen methodology while letting you understand the benefits, the risks and the assumptions behind the numbers you are being shown.

6. No. 06

   What kinds of statistical models can you work with?

   A wide range. Frequentist regression-based MMMs, Bayesian MMMs, panel models, mixed-effects models, and MMMs that model sales and brand-health key performance indicators (KPIs) together, with media, pricing and other drivers as inputs. The interesting question is rarely the model family – it is usually the data behind it, the framework it sits inside, and the decision it is meant to support.

7. No. 07

   Can you support us with attribution analysis?

   Yes. We are familiar with the main attribution methodologies and the limits each runs into. For a genuinely simple, single-channel business, attribution can be the right tool on its own – there's no need to overbuild. Once you're running true multi-channel campaigns with real prioritisation decisions to make, attribution makes most sense as part of a broader measurement framework. We support attribution at the strategic level – what to read from it, what not to, and how to reconcile it with the MMM read of the same channels.

Got another question? Drop a line.