Marketing Measurement: Predictions for the Year Ahead
Five trends shaping measurement and effectiveness practice - and what each one means for advertisers.
Every January the industry publishes a stack of predictions. Most are wrong, vague, or both. Here are five that we expect to actually shape measurement and effectiveness work over the next twelve months.
1. The retreat from last-click is finally happening
Five years of arguing that last-click attribution undercounts upper-funnel work is finally landing. Expect more advertisers to formally retire MTA-only reporting in favour of an MMM-plus-experimentation stack. The shift is being forced by privacy changes as much as conviction, but it is overdue either way.
2. In-housing peaks and partly reverses
The wave of in-housing MMM and analytics teams that began in 2020 will plateau. Some functions stay in-house - the strategic, interpretive work. Some quietly move back to specialist partners - the heavy lifting on model builds and data engineering - because the talent market makes it uneconomic to do everything internally. Expect a hybrid model to become the norm.
3. Retail media gets its first real measurement scrutiny
Retail media networks have grown faster than any other channel, largely on the strength of self-reported ROAS numbers that the buyers cannot independently verify. Expect at least one major advertiser to publish a serious incrementality study on retail media, and for the results to recalibrate the market.
4. AI gets folded into MMM workflows, not into the maths
The useful AI in MMM is the workflow layer: automated data cleaning, variable selection, model diagnostics, plain-English summaries for stakeholders. The underlying maths is still regression and Bayesian inference. Vendors that lead with "our AI replaces MMM" will lose ground to vendors that lead with "our AI makes MMM 5x faster".
5. The CFO becomes the most important stakeholder
As marketing budgets come under harder scrutiny, the audience for effectiveness work is shifting from the CMO to the CFO. Practices that can produce defensible, investment-grade reporting - with clear assumptions, confidence intervals, and validation - will get protected. Practices that produce glossy decks of channel ROIs will not.