Description
In a mock shop, people choose repeatedly between their own first- and second-choice brands while bias cues vary in strength, measuring how much preference each cue can transfer.
This design adapts choice-based conjoint to ask how far behavioural cues can move people away from the brand they already prefer. Each respondent first names a favourite and a second-favourite brand in a category they are about to buy in. A neutral, retail-style screen then shows the two brands side by side, with information boxes expressing biases such as social proof (review ratings), authority (expert endorsement), immediacy (delivery speed), scarcity, the power of free and category heuristics at weak or strong intensities, while price is held at the expected market value. Share-of-preference models estimate how much preference each cue, or all cues together, transfers from favourite to second choice, and a fictional brand with zero prior exposure benchmarks what cues can do without brand equity. The original study ran across many categories with about 1,000 shoppers each.
Attributes
- Discipline
- Behavioral Science & Economics
- Loss type
- Siloed
- Data
- Quantitative
- Effort
- Weeks
Steps
- Select biases from the behavioural literature that act during exploration and evaluation and can be expressed on a product page.
- Select categories spanning complexity and enjoyment, recruit people who intend to buy within the category's normal purchase window, and exclude those who have already decided.
- Ask each respondent for their first- and second-choice brands in the category.
- Present paired choices on a generic retail-style page with the two logos and information boxes carrying bias expressions at graded intensities, holding price constant.
- Collect about ten choices per respondent and estimate the share-of-preference change for each cue and combination with a conjoint-style model.
- Run scenarios: the second brand simply appearing, single-cue boosts, all cues on the second brand, and all cues on a fictional new brand.
- Validate the strongest effects with controlled in-market tests before acting on them.
Sources
- Decoding Decisions: Making Sense of the Messy Middle (Rennie et al., 2020)
Also known as
- Messy middle purchase simulation
- Bias supercharging test
- Second-choice brand switching experiment
- Simulated purchase bias experiment
Categories
- Analytics & Behaviour
- Behavioral Science & Economics