Run diary studies where each entry builds on the last
Turn on Memory and the AI research assistant follows up on what each participant said in earlier entries, with recaps, scheduled breaks and resume built in.
Dag MalstafAt a glance
- The AI research assistant remembers what participants said in earlier entries
- Later entries open with follow-ups on the last one
- Returning participants see a recap of where they left off
- Scheduled breaks after any question, with an optional lock
- Group the interview grid by participant
Release details
- Shipped
- Area
- Interviews
- Kind
- New
- Author
- Dag Malstaf
A diary study tells one story across several days. Until now, each entry started from scratch: a participant came back on Thursday and the AI research assistant had no record of what they said on Monday. Turn on Memory and it does. It follows up on what each participant said in their earlier entries, so day four builds on day one.
Forty seconds on one week of a diary study: a short video conversation each day, follow-ups that pick up on what the participant said the day before, and a summary of what changed by Friday.
How it works
Allow repeat entries. In Advanced settings, under Security & access, allow multiple completions and set how many entries each participant can record.
Turn on Memory. The AI research assistant reads what the participant said in earlier entries and refers back to it in the next one.
Welcome them back. Returning participants see their ongoing entries with a short AI recap of where they left off, and pick up the right one.
Ask once. One-time questions, like demographics, run on the first entry and skip after that.
Schedule a break. Add a pause after any question: a statement the AI research assistant announces, a duration from one minute to an hour or your own, and an optional lock so the participant can't resume early. Use it to send people off to try the product.
Give them time. Set the abandonment timeout per study, from 3 to 14 days, so an unfinished entry stays open while the participant gets back to it.
Follow one person. Group the interview grid by participant to read each person's entries in order.
Follow-ups that pick up the thread
In the builder's test run, a participant said in their first entry that they wanted to do more sport in the evening, but tiredness got in the way. In the second entry they mentioned a run. The AI research assistant asked: "Last time you mentioned wanting to do more sport in the evening, though tiredness sometimes got in the way. How did the run last night fit into your usual evening routine?" That is the question you would ask with Monday's notes open.
Research teams asked for it in four situations
These come from teams we work with in household and personal care, beauty, food ingredients, automotive and media, and from research agencies. We kept them anonymous.
In-home usage tests. First impressions when the product arrives, then check-ins after a week of using it in place of the usual one.
Routines, day by day. A skincare routine or a household appliance, recorded across several days.
Follow-ups with context. Teams wanted later entries to build on earlier answers, so the conversation never feels repetitive.
The same people over weeks. Weekly check-ins with one group of customers to see what changes.
What this means for your research
A diary study now reads as one conversation per participant. Later entries dig into what changed since last time, and less of each entry goes to re-establishing context. You still design the study: entry routing decides which questions each entry asks, and Memory shapes the follow-ups.
Memory depends on recognizing the participant, so recruit through a route that passes the same participant ID every time. A scheduled break doesn't send reminders, so plan your own return invitations.
Diary studies with Memory in three minutes
A narrated tour of a diary study with Memory, with the four situations research teams brought to us.
Watch the diary study tools in action
Read the diary studies guide to set up your first multi-entry study.


