Marketing mix modelling is back at the centre of the CFO conversation- twenty10Multi-touch attribution alone can't survive cookie deprecation: triangulation is the new default- twenty10Geo lift tests and incrementality studies are calibrating the next generation of MMM- twenty10Clean rooms + MMM: how retailers and CPGs are closing the measurement gap- twenty10The UK MMM landscape: who's serious, who's selling a dashboard- twenty10Comparing marketing mix modelling providers in the UK- twenty10Brand vs performance: measuring long-term equity effects in a P&L your CFO can read- twenty10Marketing mix modelling is back at the centre of the CFO conversation- twenty10Multi-touch attribution alone can't survive cookie deprecation: triangulation is the new default- twenty10Geo lift tests and incrementality studies are calibrating the next generation of MMM- twenty10Clean rooms + MMM: how retailers and CPGs are closing the measurement gap- twenty10The UK MMM landscape: who's serious, who's selling a dashboard- twenty10Comparing marketing mix modelling providers in the UK- twenty10Brand vs performance: measuring long-term equity effects in a P&L your CFO can read- twenty10
All insightsAttribution

Is Multi-Touch Attribution Done?

MTA is dying quietly in most marketing stacks. Here is why: what killed it, what marketers are replacing it with, and where a residual use case still exists.

twenty10··5 min read

A pattern has emerged in conversations with advertisers this year: the MTA licence is up for renewal, and no one is quite sure why they are renewing it. Most of the numbers are already ignored. The media team optimises to platform ROAS. The finance team looks at MMM. The board looks at revenue. MTA sits between all of them and moves nothing.

That is a symptom, not a strategy. Multi-Touch Attribution: the discipline of tracing every touchpoint in a user's journey and assigning a fractional credit to each: is dying quietly across most marketing stacks. The question is whether to prolong the decline or make the break.

What killed MTA

Three things, in sequence.

The identity graph collapsed. iOS ATT stripped mobile app identifiers. Safari and Firefox blocked third-party cookies years ago. Consent-mode opt-outs remove 30-60% of web sessions in Europe. The user-level journey that MTA was designed to reconstruct no longer exists in most of the data.

The platforms went private. Google, Meta, TikTok and Amazon each optimise inside a walled garden that no MTA vendor can see into. The MTA journey ends at the platform boundary and starts again on the other side, with the platform's own attribution telling a different story.

Modelled conversions took over. Platforms now backfill missing signal with their own probabilistic models. An MTA vendor stitching those together is not measuring reality: it is averaging vendors' guesses.

Where MTA still has a residual use

There is a narrow, honest use case left. For a purely digital, direct-response business with a single primary channel and clean first-party data, a lightweight MTA can still add value as a within-channel optimisation signal: which creatives, keywords or audiences to bid on. That is a real, if diminished, job.

What MTA cannot do any more is answer the incremental question: what happened because of marketing that would not have happened otherwise. That question belongs to experiments and to MMM.

The replacement stack

For most advertisers, the replacement is a triangulation model:

  • Server-side conversion tracking for platform optimisation. Owns first-party identifiers, feeds them back into the platforms via APIs.
  • Incrementality testing for causal reads. Geo holdouts, ghost bids, lift studies. Run continuously against the largest channels.
  • Marketing Mix Modelling for allocation. Refreshed monthly, calibrated against the experiments, delivered as a decision system.

None of those three is MTA. Together they do the job MTA promised and never delivered.

The renewal decision

The question to ask before renewing an MTA contract: what decision did the MTA output actually change last year? For most advertisers the answer is none, or the same decision the MMM would have produced. That is the case for redirecting the budget: to the experimentation programme that is not yet funded, or to the MMM refresh that is once a year when it should be monthly.

MTA is not being killed. It is being outgrown.