Matching people to places they actually want
Built preference-elicitation into the browsing experience
Stated preferences are unreliable; revealed preferences from behavior are more predictive
Used temporal discounting principles to rank urgency
People discount future options; surfacing high-match locations reduces present bias
Toyota's online configurator was losing customers mid-funnel. Analytics showed high drop-off between trim selection and lead submission — a 'paradox of choice' problem with 200+ configuration combinations.
Preference instability under complexity: when people can't hold all options in working memory simultaneously, they satisfice on salient features (color, price) and defer on others (tech packages, drivetrain). This creates regret risk and abandonment.
A 3-week diary study tracking preference stability across repeat visits, followed by a conjoint analysis to identify the minimum-viable configuration set. A progressive-disclosure prototype reduced options to a 4-question guided flow.
Figures accurate as of patent grant dates. Mockups illustrative — no real Toyota UI reproduced. No logos used.