When a Week of Cravings Became a Year of Missed Targets
A real-world example of Projection bias in action
Context
A mid-stage food-tech startup built a meal-planning subscription that lets users pick weekly menus and receive pre-portioned ingredients. Early testers loved the convenience during a seven-day pilot, so the product team used that enthusiasm to set aggressive long-term retention and supply commitments.
Situation
After a successful seven-day trial with 3,200 users, the product and growth teams extrapolated weekly usage into an annual retention forecast and negotiated six‑month supply contracts with ingredient vendors. Pricing, marketing messaging, and hiring plans were all aligned to the projected subscription base they assumed would keep choosing the same meals month after month.
The bias in action
Team members assumed that the preferences and routines demonstrated during the short pilot would persist, failing to account for fluctuating factors like weekends, holidays, social plans, seasonal tastes, and menu fatigue. Product managers used the pilot's 42% weekly reorder rate as a proxy for long-term customer lifetime value, and finance baked that into revenue and hiring models. Because the pilot period captured a temporary state (users motivated to try something new and keen to reduce cooking effort), planners overprojected steady demand and ignored evidence that appetite and schedules change. Decisions were made under the implicit belief that users' current tastes and circumstances would remain stable over months.
Outcome
Within three months of launch active subscribers dropped to 18% of the original cohort, far below the forecasted 42% retention. The company was left with excess inventory orders and monthly vendor minimums they could not meet, leading to $120,000 in perishable waste in the first quarter and renegotiation penalties. Revenue projections missed targets by $1.2 million for the first year and hiring plans had to be frozen, slowing product development and eroding investor confidence.