All insightsMeasurement & Market Mix Modelling

Establishing a Good Marketing Effectiveness Practice

Most effectiveness functions stall because they are set up to produce reports, not decisions. Here is a structure that does not.

twenty10··7 min read

Most large advertisers have a "marketing effectiveness" function somewhere. Most of them produce a steady stream of decks that nobody acts on. The cause is almost always structural, not analytical.

A good effectiveness practice has four components, deliberately separated and deliberately connected.

1. A measurement layer that is owned, not outsourced

Whoever owns the measurement methodology - MMM, incrementality, attribution - needs to be inside the business, not at an agency. Agencies can build models. They should not own the answer. The conflict of interest is too obvious: an agency cannot credibly report that the channel they sell you is underperforming.

In practice this looks like a small in-house team (one senior, one or two mid, one engineer) that commissions models, validates them, and writes the recommendation.

2. A planning layer that uses it

This is where most practices fail. The effectiveness team produces an MMM. The planning team uses last year's plan ± 5%. The two teams meet quarterly, nod politely, and nothing changes.

The fix is procedural: the planning calendar must have a fixed point at which the optimised mix becomes the *starting* point for the next plan, with explicit, written reasons for any deviation. If the planners can ignore the model with no documentation, they will.

3. An experimentation cadence

Models are hypotheses. Without experiments to validate them, an MMM drifts further from reality every quarter. A good practice runs at least four meaningful experiments a year - geo holdouts, creative tests, channel on/off - and feeds the results back into the next model build as priors.

Without this loop, you have a measurement function. With it, you have a learning function.

4. A single owner with authority

The single biggest predictor of whether an effectiveness practice produces decisions is whether there is one person with the authority to say "we are moving £5m from this channel to that one". If three people share that authority, none of them will use it.

That owner does not need to be the CMO. It often works better as a Head of Marketing Effectiveness with a direct line to the CMO and the CFO. The CFO matter is important: effectiveness work is fundamentally an investment-allocation discipline, and the CFO is the natural ally.

The maturity ladder

If you want a shorthand for where you are:

  • Level 1: You produce reports.
  • Level 2: You produce recommendations.
  • Level 3: Your recommendations move money.
  • Level 4: Your recommendations move money *and* you can show, with experiments, that the moves worked.

Most advertisers live at Level 1 or 2. The leap to Level 3 is organisational, not analytical.