Brand Measurement Isn't Broken. The Brief Is.

Brand StrategyCustomer Experience ExcellenceCustomer Value PropositionData Analytics
Written by Jaslyin Qiyu

I came away from a recent marketing roundtable disturbed by how disconnected many brand marketers felt from measurable attribution and ROI for their campaigns. While the disconnect is understandable, the underlying rationale is also not good enough.

The research says marketers are being held to an ROI standard using instruments they themselves barely trust. The usual response, another dashboard, makes that worse.

My argument is this. The technical problem is real but narrower than we claim. The bigger failure is that brand is often measured as a media outcome when it is delivered as an experience, and nobody has agreed with finance and business what proof looks like before the money moves.

The problem is real, and it is a trust problem

Much of what circulates on this topic is vendor content. These are the figures I would stand behind:

  • CMO Alliance’s 2026 report (50 senior marketing leaders) found CMOs rate confidence in their own strategy at 7.5 out of 10 but trust in attribution data at 2.8 out of 5. 19% run no formal attribution model at all.
  • IAB’s State of Data 2026 (400+ buy-side decision-makers) found 60% to 75% say their advanced measurement falls short on rigor, timeliness, trust or efficiency. Only 39% use attribution, incrementality and MMM together.
  • Perion and Advertiser Perceptions found only 22% of CMOs strongly agree they have the measurement insights to justify marketing’s value to their CFO.
  • NIQ’s CMO Outlook found 84% of CMOs treat ROI as the primary metric for budget allocation, and 74% say scrutiny to prove it has increased.

Read together: we are accountable to a number we do not believe. Headlines claiming “75% of CMOs can’t prove ROI” overstate it. The IAB finding is about tools falling short, not marketers failing, and that distinction points to the fix.

The technical problem is narrower than we pretend

What is gone is deterministic, user-level mapping: this person saw this brand ad, then bought. Privacy rules, signal loss and platform silos ended that, and each platform now reports conversions through its own lens.

What is not gone is aggregate causal measurement. Marketing mix modelling, geo experiments and long-window holdouts can all estimate how brand investment moves sales, price tolerance and acquisition cost. They are slower and costlier than a click report. That is an effort problem, not a physical limit. An academic study of 15 advertising experiments run inside Facebook is the usual reference for treating attribution output as directional reporting rather than proof, and it is why 52% of US brand and agency marketers now use incrementality testing.

So “brand can’t be measured” is sometimes a genuine technical statement and sometimes cover for not having done the work. A CFO can usually tell which one they are hearing.

The real gap is expectation

Finance applies the standard of performance marketing, fast, deterministic and attributable, to something that works on a different timeline. Brand effects build over quarters and years. They show up as lower cost to acquire, stronger conversion on everything else and less price sensitivity. Demand click-level proof and brand will always lose. NIQ found the share of CMOs who say their CEO and CFO believe in long-term brand building fell from 80% to 69% in a year.

Marketing owns part of this:

  • Brand is rarely written as a financial hypothesis. “Awareness” is a metric, not an outcome.
  • There is no agreed standard of proof before the money moves, so the method gets argued after the result.
  • The spend is often not legible to finance. Regions file the same cost under different categories, and the CFO gets three answers to one question.

Perion’s research points the same way. Marketers using unified measurement closed trust gaps with their CFOs. The shared framework mattered, not the dashboard.

Brand has a different job at different stages

A single brand ROI standard applied to a challenger and an incumbent will misjudge both.

Brand has a different job at different stages
Dimension Newly established Mature
Starting point Unknown to most target customers Known, even by people who never buy
Job of brand Awareness, then consideration, then conversion though low value goods can jump ahead to conversion with a highly attractive promotion but recurring purchase at original price becomes a challenge Be first choice, especially when wallets tighten and choices broaden
Credible signals Reach, aided awareness, branded search growth, brand lift, geo tests, repeat purchases Win rate, price realisation, retention, share of wallet, share of consideration, total customer lifetime value
Measurement difficulty Easier: a large effect against a near-zero baseline Harder: a small effect against a large baseline

The mature case is the one that trips teams up. Awareness metrics are meaningless there, and performance channels easily take the credit, because existing demand can look like marketing-created demand.

One caution here - known is not the same as known for the right thing. A mature brand’s hypothesis cannot be “we are known.” It has to be “we are chosen, for a reason that justifies our price.” That is testable.

Brand and experience are one thing, measured twice

Experience is the lived outcome of the brand: what customers and employees go through across products, services, touchpoints and people. Brand is the promise. Experience is the delivery. Splitting them is why brand gets judged on media metrics while much of its real effect is booked elsewhere.

The most useful thing to measure is the promise-delivery gap. Brand spend that raises expectations the experience cannot meet is worse than wasted. It pulls customers in and speeds up churn, complaints and negative word of mouth. Strong experience under a vague brand leaves value on the table, because customers cannot say why they stayed or recommend you.

For mature brands especially, CX data can be the best metric a CFO will accept: retention, referral, price realisation, win rate against cheaper alternatives. Marketing that calls brand unmeasurable while CX holds the retention data in another dashboard is ignoring its own best evidence.

Two cautions:

  • Synergy is not merging. Name who owns the promise and who owns the delivery, and hold both to one shared outcome. Otherwise “everything is brand” becomes “nobody owns anything.” and accountability remains in silo.
  • Employees are a delivery channel. In service-heavy categories the people are the product at the moments that matter, so understanding of the promise and the ability to deliver it belong in the same case.

One promise-to-outcome case, built with finance

Stop defending brand with a media report. Build one case, jointly owned by marketing, CX and finance, adjusted for the brand’s stage. Three questions test whether you have one:

  1. Can marketing state, in one sentence, the business outcome brand spend is meant to move and the time horizon?
  2. Has finance agreed in advance what evidence would count, and what would prove the hypothesis wrong?
  3. Is the budget coded consistently enough that spend can be tied to that hypothesis?

If any answer is no, the problem is organisational before it is technical. If all three are yes and the measurement still cannot confirm it, you have a genuine methods problem. The route there is to use attribution for tactical optimisation, incrementality tests to calibrate, and marketing mix modelling for allocation.

The measurement industry will keep selling better dashboards. The CFO is not asking for a dashboard. The CFO is asking whether we know what we are doing, and whether we can say so before the money is spent. Brand measurement isn’t broken. The interpretation and alignment are.

Sources

Figures are as reported in the pages below.

Jaslyin Qiyu
Next

Organic is dead, or is it? What to pay for and what to earn in 2026