Why You Need a Model, Not a Hunch

Conversations about banner design tend to stall on opinion. One person says a friendlier banner will lift opt-ins; another worries about the design cost. A small revenue model ends the debate by attaching a number to the question: what is one percentage point of consent rate worth to us?

The model below needs only data you already have in your ad platform and analytics. It uses your numbers, not industry averages, because eCPMs vary enormously by format, geography and audience.

The Inputs

• Eligible monthly impressions from regions where consent is requested (for example EU, EEA and UK traffic).
• Consent rate in those regions today.
• eCPM for consented traffic, taken from your ad platform split by consent state.
• eCPM for non-consented traffic, where ads run without personalisation or with limited signals.

If you cannot split eCPM by consent state yet, that is your first task. Without that split, any estimate is guesswork.

The Formula

Revenue from a region is the sum of two parts:

1. Impressions × consent rate × consented eCPM ÷ 1,000
2. Impressions × (1 − consent rate) × non-consented eCPM ÷ 1,000

The value of one point of consent rate is therefore:

Impressions × 1% × (consented eCPM − non-consented eCPM) ÷ 1,000

The key variable is the gap between the two eCPMs. A wide gap makes consent rate very valuable; a narrow gap means other levers matter more.

A Worked Example With Made-Up Numbers

These figures are purely illustrative. Replace them with your own. Suppose a site serves 2,000,000 eligible impressions a month in consent regions, with a consented eCPM of $4.00 and a non-consented eCPM of $2.50. The gap is $1.50. One point of consent rate moves 20,000 impressions from one bucket to the other: 20,000 × $1.50 ÷ 1,000 = $30 per month. A ten-point improvement would be worth $300 per month at that traffic level.

That might be small or large depending on your scale, which is exactly the point: now you can compare it with the effort required.

Turn It Into Decisions

• Prioritise work. If a point is worth little, spend your time on fill rate or page speed instead.
• Set a break-even. If a banner redesign costs a known amount, divide by the monthly value of a point to see how many points it must win.
• Model by segment. Run the formula separately by country, platform and language. Segments with a big eCPM gap and large volume deserve attention first.
• Sanity-check seasonality. Q4 eCPMs usually differ from Q1, so refresh the inputs each quarter.

Limits of the Model

Keep the simplifications in view. Consent rate may change the mix of visitors you measure. Demand for non-personalised inventory can shift over time. Correlation is not causation: a revenue change after a banner update may have other causes. Treat the output as a range for planning, not a forecast you promise to anyone.

Where FlexyConsent Fits

FlexyConsent is a Google-certified CMP (CMP ID 317) with a banner in 57 languages and support for IAB TCF 2.4 and Google Consent Mode v2. All plans include the same features, and there is a free plan, so cost of tooling is rarely the variable that decides the outcome. Your own eCPM gap is.

Key Takeaways

• Put a number on consent rate using your own impressions and eCPM data.
• The value of a point equals impressions × 1% × the eCPM gap ÷ 1,000.
• Split eCPM by consent state first; without it, the model cannot work.
• Run the model by segment and refresh it every quarter.
• Use the result to prioritise work and set break-even targets, not as a guaranteed forecast.