Case Study · Drivers and MaxDiff, then rotated monadic message testing with forced choice
Why the most-liked message finished last: a two-phase message test
In a two-phase message test for a consumer wellness brand, the message that tied first on top-box likeability (42%) finished last of 6 on forced choice (13.3%). The message that ranked last on likeability won its concept on forced choice. A ratings-only test would have picked the wrong message.
The Challenge
A message test can only rank the messages it is given. The client wanted messages written from evidence, which meant measuring drivers and barriers first, then testing what the marketing team wrote against them.
Our Approach
Phase 1 measured drivers and barriers with three MaxDiff exercises on 13 product features (fair-share threshold 7.7%), 12 benefits and 12 barriers (both 8.3%). The client team and its agency then spent about four weeks writing messages against that output. Phase 2 tested 5 message concepts of 6 messages each in rotated monadic cells, with every respondent seeing 2 of the 5 concepts, and added a forced-choice ranking to the usual diagnostics.
What the Research Found
- ›Within one concept, the message tied first on top-box likeability at 42% finished 6th of 6 on forced choice, at 13.3%.
- ›The message 6th of 6 on likeability, at 35%, finished first in that concept on forced choice at 20.7%, within 0.6 points of the highest-scoring message anywhere in the test (21.3%).
Outcome
The client received a message hierarchy with the forced-choice result alongside each score. A third wave repeated the exercise for a second product line.