Endowed Progress Effect
The endowed progress effect is Joseph Nunes and Xavier Drèze's finding that people work harder to finish something they believe is already under way. In their 2006 car-wash study, customers given a ten-stamp loyalty card with two stamps already filled in completed it at 34 per cent, against 19 per cent for customers given a blank eight-stamp card. Both groups had to buy eight washes. What changed was whether the task read as continuing or as starting.
What it is
Nothing is added except a belief. The two free stamps do not reduce the number of car washes anybody has to buy, and the customers can count. What the endowment does is reclassify the situation: a person with a blank card is deciding whether to begin, and a person with two stamps is deciding whether to abandon something they have started. Those are different decisions, and people make them differently.
The study is "The Endowed Progress Effect: How Artificial Advancement Increases Effort", in the Journal of Consumer Research in 2006. Three hundred loyalty cards were distributed at a car wash. One group received a card requiring eight stamps, blank. The other received a card requiring ten stamps, with two already applied. The real requirement was identical. Completion ran at 34 per cent for the endowed group and 19 per cent for the blank-card group. Those figures were verified against the primary source on 1 September 2026.
Eyal renders the same result in Hooked as "82 per cent higher completion", which is the relative gain between 19 and 34 and is arithmetically correct. Stated bare, it reads to a skimming eye as a percentage-point improvement, and it is doing so in a book aimed at people who will use it to justify a design decision. Both rates belong in any quotation of this finding.
There is a family resemblance to other effects on this site that runs through effort rather than framing. The IKEA effect says that work you have done raises what the output is worth to you. The endowed progress effect says that work you have not done, credited to you by someone else, raises the odds you will do the rest. One trades on labour performed, the other on the appearance of labour already banked, and the second is considerably cheaper to supply.
The evidence status here is primary-cited. The citation is correct, the figures come from the paper, and nobody on this site has audited the loyalty-programme and goal-pursuit literature that grew around it. Treat the 34 against 19 as a well-attributed single field study, in one product category, with 300 cards, rather than as a summary of what twenty years of research has established.