TruthTips: Choice Modelling - serving up trade-offs like a pro

Reduce the Say-do-gap: using trade-off

methods to understand what people

actually want.

Everyone says they want everything. But what truly makes them choose?

Trade-off methods help cut through the noise, revealing the features, benefits, and ideas that drive decisions.

Get it right, and you’ll launch something they can’t ignore.

Tip 01: Curate - Cut the noise before you Start

Don’t just throw features into the mix: curate them. Use a pre-phase (qual or short quant) to filter out fluff, duplicate ideas, and anything you wouldn’t actually act on.

Better inputs = better output.

Tip 02: Make the choice feel real

Design tasks that mirror real decisions. Include opt-outs (‘None of these’), competitors, price anchors, or feature bundles; whatever reflects how people choose.

Otherwise, you’re testing a wish-list.

Tip 03: Force Trade-offs, not fatigue

Respect the respondent. Use efficient experimental designs to keep tasks tight (8–12 per person). Add warm-ups to reduce learning curves.

More isn’t more: better is better.

Tip 04: Beyond Preference - Model Strategy

Think beyond utilities. Use simulators to test real-world scenarios: drop a feature, add a premium tier, bundle an offer. Simulate pricing, MVPs, or competitor launches to stress-test your strategy.

Tip 05: From choice data to personas

Trade-off outputs don’t just spit out preferences—they reveal patterns of choice. Use latent class models or HB scores to cluster respondents, then enrich those segments with attitudes, behaviours, and commercial value.

So what?

Choice modelling helps you cut through what people say to reveal what they’ll choose. This will help design offers, innovations, and strategies that win in the real world.

It’s the difference between guessing at value and knowing what drives demand.

Next time:

Segments that stick. The art and science of uncovering people personas: wants, needs, attitudes & more.