Brand and performance: arbitrating horizons, not labels.
Evidence status
What was reproduced, and what it does not prove.
The sealed scenario and Python verify the budget, the signals and the illustrative gap of −€240k. They prove neither an observed brand effect, nor real incrementality, nor channel return.
- Technical specification of the brand–performance trade-off, specification in French (opens in a new tab)
v1, 31 August 2026 · CC BY 4.0 · specification in French · Method document written by marketing-mix.net. The figures are illustrative: they are not market measurements.
Key distinctions
Three conditions before calculating.
Role before channel
A video can create consideration, retarget or support an offer. Classifying budgets by platform rather than by mechanism leads to incoherent comparisons.
Compatible horizon
A seven-day conversion test undervalues an effect that changes brand search and the probability of choice over several months. Conversely, an awareness indicator is not enough to judge a promotional activation.
A portfolio of effects
The decision must combine immediate measures, intermediate signals and economic outcome. No single method observes the whole chain perfectly.
Method
Arbitrating the communication portfolio
- 01Diagnose the bottleneck
Separate lack of penetration, weak consideration, availability failure, cost of conversion and retention; the same budget ratio does not answer all of those problems.
- 02Assign a measurable role
For each budget line, write the expected behaviour, the population, the delay and the decision indicator — a vague objective produces opportunistic metrics.
- 03Triangulate the methods
Combine experimentation for the local effect, time series for aggregate effects and brand tracking for the intermediate mechanisms.
- 04Reallocate at an appropriate frequency
Revise activation quickly, but protect the investments whose signal requires several months; the decision cadence must follow the speed of the mechanism.
Two time profiles of contribution
The chart normalises total contribution to 100 to show when the effect materialises. It does not assume that a channel belongs definitively to a single role.
A short measurement window sees capture better. A portfolio decision must also cover how demand forms and persists.
Illustrative scenario: the best apparent ROAS can destroy reach
A brand moves €600,000 of reach video into retargeting, whose attributed ROAS is twice as high. Sales stay stable for six weeks, then new visitors fall by 14% and acquisition cost rises by 19%.
Retargeting captured demand that had already been created. Valuing the drop in new customers over six months, the contribution lost reaches €410,000, against €170,000 of gain attributed in the short term.
Restore an envelope for demand creation and measure separately the effect on new visitors, brand search and incremental contribution.
Acceptance conditions
What must be true to act.
- 01
Each investment has an explicit role, population and delay.
- 02
The demand-creation budget is not assessed with a short conversion window alone.
- 03
Short-term and long-term methods are reconciled on a common economic unit.
- 04
The reallocation cadence respects the delay needed for the signal to appear.
Limits
What this analysis does not prove.
- Brand effects are diffuse and can be confounded with distribution, price or experience.
- Self-reported indicators do not guarantee future behaviour.
- A universal brand-performance ratio ignores category, maturity and the growth bottleneck.
References
Works cited.
- Gordon, B. R. et al. (2019), “A Comparison of Approaches to Advertising Measurement”, Marketing Science.
- Lewis, R. A. & Rao, J. M. (2015), “The Unfavorable Economics of Measuring the Returns to Advertising”, Quarterly Journal of Economics.
- Vakratsas, D. & Ambler, T. (1999), “How Advertising Works”, Journal of Marketing.