At a glance
Advertising attention metrics can help teams compare exposure quality and diagnose creative or placement problems, but they do not prove incremental revenue. Use them alongside brand, commercial and causal evidence, with clear definitions and validation.
An ad can be viewable without being noticed. It can be noticed without being understood. It can be understood without changing anyone’s mind.
Advertising attention metrics try to improve the first part of that picture: whether people had a meaningful opportunity to attend to an ad, and sometimes whether they actually did. That is useful territory. It is not the same thing as proving a sale.
Find out what the metric measures
The IAB and MRC attention measurement guidelines cover a range of approaches, including visual tracking, physiological and behavioural signals, and other measures used to assess attention. A shared label does not mean every provider observes the same thing.
Some scores rely on direct observation within a research sample. Others model attention using characteristics of placements or exposure. Ask which parts are observed, which are inferred and where the model has been validated.
A score can be useful while still being an estimate. The problem arises when a precise-looking number hides a definition the buyer does not understand.
Ask five questions before buying a score
- What human behaviour does the metric claim to represent?
- How is that behaviour observed or estimated?
- Which audiences, devices and formats were used to develop the model?
- What uncertainty or coverage gaps should we expect?
- What decision will the score help us make?
The fifth question is easy to skip. If the team will not change creative, placement or investment based on the result, the measurement may become another line in the report rather than a useful tool.
Use attention to diagnose the campaign
A low-attention placement may explain why a well-made ad has little effect. A high-attention placement with weak brand recognition may expose a creative problem. A memorable ad that leaves people confused about the offer needs a different repair again.
Our view is that attention belongs between delivery and outcome in the review. It can help explain why a campaign worked or failed, but it should not replace the outcome the campaign was built to achieve.
| Evidence | What it helps you inspect |
|---|---|
| Delivery and viewability | Whether the ad had a suitable opportunity to appear |
| Attention | Whether exposure was likely to involve notice or focus |
| Brand recognition and message recall | What people took from the creative |
| Commercial outcomes | Whether the activity supported the business objective |
Do not reward attention at any cost
An intrusive placement can attract notice while damaging the experience. A startling creative device can be remembered more clearly than the advertiser. Longer exposure is not automatically better if the message could have been communicated clearly in less time.
Keep placement quality and suitability in the evaluation. Ask whether the attention came from the audience and context you wanted, not just whether a score increased.
Also consider the price. A placement with twice the attention score and three times the cost may still be worthwhile, but the score alone does not settle the investment.
Test the additional information
Choose a campaign with enough scale to compare useful alternatives. Define what you would decide using the existing measures, then examine whether attention data changes that decision in a defensible way.
Look for consistency across more than one creative or short period. If a provider claims its metric predicts a commercial outcome, ask for evidence relevant to your category and campaign, rather than accepting a general case study as proof.
Connect the test to the appropriate measurement approach. A relationship between attention and sales in observed data does not, on its own, establish that increasing the score caused the sales.
The best attention measure makes a buying or creative decision clearer. It should give the team a better question to ask, not a new number to worship.
