Diary studies are the only instrument here that captures a decision rather than an interaction. They are also the easiest to ruin, because a participant who works out what you want to hear will produce it every day for a week.
What to put in the public brief
- The exact number of days and entries expected
- Roughly how long each entry takes
- What happens if they miss a day
- That entries are written, with photos where relevant
- The full reward and when it is paid
Keep the study prompts and any product identity private.
What to put in the private instructions
Give one prompt per day, and vary it
Identical prompts produce identical entries by day three. Rotate between what happened, what they decided, and what annoyed them.
Ask about the day, not the product
The product should appear in the diary only when it genuinely came up. If it never does, that is the most important finding the study can produce.
Say what to do about a missed day
Write it the next morning and mark it as late rather than skipping or backfilling silently. An honest gap is data; a reconstructed entry is noise.
Ask for one photograph a day where it makes sense
A screen, a shelf, a receipt. Photographs anchor entries to real moments and make retrospective invention obvious.
Proof worth requiring
Ask for
Dated entries with timestamps, any photographs, and a short closing summary written after the final day rather than instead of it.
Do not ask for
A completed diary submitted all at once at the end. It defeats the entire method, and it is the single most common way a diary study becomes a long survey.
When this task type is the wrong tool
If your question is about a single interaction, this is an expensive way to answer it. A usability or onboarding test does it in a session.
If you need numbers rather than narrative, a survey will give them to you, with the caveats that page sets out.