Communication synergies: measuring the interaction without double counting.
Key distinctions
Three conditions before calculating.
Comparable exposures
The four groups must be constituted before the outcome is observed, with random assignment or a credible identification strategy. Naturally multi-exposed users do not form a control.
Distinct interaction
The synergy is computed as a difference of differences: joint result minus activity A, minus activity B, plus control. It must not be credited in full to both activities.
Value after cost
A statistically positive interaction can remain too small to fund the creative, the targeting or the sequencing complexity. The decision applies unit contribution and specific costs.
Method
Designing and interpreting a synergy measurement
- 01State the interaction
Describe why A should change the response to B — priming, evidence, repetition, coverage or recall; define the order, the delay and the population before the test.
- 02Create four cells
Size control, A, B and A+B over the same period with verifiable exposure rules; avoid contamination and measure treatment adherence.
- 03Compute the effects
Estimate effect A, effect B, total effect and the interaction; report levels, differences, uncertainty and any negative effects on the guardrails.
- 04Arbitrate the sequence
Value the additional conversions, subtract the costs of both activities and of their coordination, then check whether a simpler version keeps most of the effect.
The 2×2 factorial design and the interaction term
The four cells make visible the effect of each activity and the additional conversions that exist only in the combination.
conversions
conversions
conversions
conversions
(1,260 − 1,000) − (1,120 − 1,000) − (1,070 − 1,000)70 conversionsThe combination produces 70 synergy conversions
Four comparable groups produce 1,000 conversions in control, 1,120 with paid alone, 1,070 with owned alone and 1,260 with both. Contribution per conversion is €45; paid costs €4,000 and owned €1,000.
Paid effect: 120. Owned effect: 70. Joint effect: 260. Interaction: 260 − 120 − 70 = 70 conversions, that is €3,150 of contribution. The joint plan creates 260 × 45 − 5,000 = €6,700 net.
The combination is retained if the interval of the interaction excludes a material loss. The 70 conversions are kept in an interaction line separate from the reporting of the two activities.
Acceptance conditions
What must be true to act.
- 01
The mechanism, the order and the window of the interaction are pre-specified.
- 02
The four cells have a counterfactual and a controlled exposure.
- 03
The interaction term is computed separately from the own effects.
- 04
The contribution of the interaction covers the cost of coordination and creative.
Limits
What this analysis does not prove.
- An interaction depends on the level of spend, on the creative and on the sequence.
- Contamination between groups reduces the ability to identify the effect.
- A two-activity test does not describe every interaction of a complete portfolio.
References
Works cited.
- Naik, P. A. & Raman, K. (2003), “Understanding the Impact of Synergy in Multimedia Communications”, Journal of Marketing Research, 40(4), 375–388.
- Montgomery, D. C. (2017), Design and Analysis of Experiments.
- Lewis, R. A. & Rao, J. M. (2015), “The Unfavorable Economics of Measuring the Returns to Advertising”, Quarterly Journal of Economics, 130(4), 1941–1973.