How this survey works.
The Abandonment Index is an opt-in reader survey, open year-round and reported once a year, every 1 January. Respondents describe the most recent book they stopped reading: its genre, publisher type, format, how far they got and why they stopped.
What stays the same
The questions and their answer options do not change between editions. That is the point of it — a number is only worth having if next year's can be set beside it. Several reasons for stopping use the exact wording of Goodreads' 2013 member poll, including phrasing nobody would write today, so that the comparison is like for like rather than approximate.
A question may be added after the thirteenth in a later edition, and labelled as that year's addition. Nothing is reworded, reordered or removed.
What this is not
Because people choose to take part, and most arrive through The Veiled Archive's YouTube channel, this website and author newsletters, the results describe the people who answered and are not a random sample of readers. Anyone telling you otherwise about a survey shaped like this one is selling something.
Every edition publishes its total responses, its collection dates, and the share of responses from each channel, so the sample can be judged rather than trusted.
What is collected
- Your thirteen answers, and the page number if you gave one.
- Which link you arrived through — the YouTube description, this site, a newsletter — recorded automatically. Question 12 asks you the same thing, and the gap between the two is itself worth knowing.
- How long the form took, to the second. It is the cheapest way to spot a submission that was not typed by a person.
No name, no email address, no account. There is nowhere on the form to put one. Nothing collected identifies you, and your free-text answer is never published unless you answered yes to the last question.
Submissions that look automated are flagged and excluded from every published figure. They stay in the raw data, marked, rather than being deleted — a number that changed because something was quietly removed is not a number anyone should trust.
One thing worth disclosing
The running totals are public from the first day of collection, on the results page. That means a respondent could have seen the current answers before giving their own, and some will have. The alternative — collecting in private and publishing at the end — would be cleaner as method and worse as an argument for taking part, and we would rather say this out loud than pretend the trade does not exist. The results page is not linked above the form for the same reason.
The comparison
Goodreads, “What Makes You Put Down a Book?”. They were read off that image. Goodreads did not publish them as text and did not publish how many members answered, so the 2013 column has no sample size and no margin of error. Goodreads, “What Makes You Put Down a Book?” · 9 July 2013 · checked 24 September 2026
That is also, more or less, why this exists. Thirteen years on, the most-quoted numbers about why readers give up on books live inside a picture with no method attached and no way to check them.
Questions
The first report publishes 1 January 2027, covering responses through 31 December 2026. Write to the Archive if you want to know anything about how it was counted; a person answers.