Top drink
Espresso
25% of answered quizzes
First-party research · United States
What US coffee finder users choose
These are anonymous preferences from people who completed a WhichCapsule finder, not a survey of all coffee drinkers or a measure of sales. The rolling window and sample size appear with the results.
WhichCapsule Coffee Insights
What coffee drinkers are actually looking for
A rolling, anonymous view of completed finder answers, top quiz matches and Amazon buy-button clicks.
Last 90 days12 completed quizzes Anonymous
Original vs Vertuo
100% / 0%
Capsule-system preference
Preferred strength
Medium
60% of answered quizzes
Milk drinks
42%
Selected milk-first options
Small kitchens
20%
Machine seekers choosing small space
Most viewed capsule match
Capriccio
9 signals
Most viewed machine match
Vertuo Pop+
12 signals
Most clicked capsule
Firenze Arpeggio
4 signals
Most clicked machine
CitiZ Platinum
4 signals
Buy clicks are outbound Amazon clicks—not confirmed orders. Internal tests are excluded.
How the data is calculatedHow we calculate these insights
Sample and period
The US preference sample counts completed finder events from the last 90 days. Internal tests are excluded. Each completed event is a response, not necessarily a unique person. Preferences are shown only after at least five answered events in the relevant group.
What percentages mean
Drink, strength and system percentages use answered responses for that question. Milk and small-space percentages use eligible answered responses for those questions. Missing answers are not treated as a “no.” Values are rounded, so a split may not add to exactly 100%.
Product signals
The product cards count result views and outbound Amazon clicks, with a minimum of three events per displayed product. Views are not independent product tests; clicks are not confirmed orders, ratings or market share. Merchant relationships do not change the quiz score.
Limits and privacy
This is a self-selected group of site visitors and may change as new responses arrive. We publish aggregates, not individual answers or sessions. If the threshold has not been met or data is unavailable, we show no percentages rather than estimates.