Diary study: see how readers actually use the e-books and academy over two to three weeks
Use it when: the question is about behaviour over time and in real life (when and where people read, what makes them come back, what makes a member drift away), which a single usability session or interview cannot show. Remote and asynchronous, so cheap and suited to a small, spread-out audience.
The method (NN/g, Kim Flaherty, 2024)
- Pick the entry trigger (combinations are common): - Event-based (in-situ logging): log an entry each time the thing happens (opened an e-book, visited the members' area). Best when events are not so frequent that logging becomes a chore. - Interval-based: one entry at a fixed time (e.g. each evening). Fine when you do not need every interaction; beware time-of-day skew. - Signal-based (experience sampling): you send a prompt (a text message) and they answer. Best when people may forget to report on their own.
- Set the length from how often the behaviour happens. NN/g's examples: a few times a day -> 1 week; once or twice a week -> 3 weeks; a long purchase journey -> as long as the journey takes. Choose a period that gives enough entries across the whole sample.
- Recruit for saturation, then over-recruit for dropouts. NN/g's guide: small discovery project, fairly similar users, small problem space ~5-12 people; large discovery, varied users ~12-30; academic or generalisable work ~30-50. Diary studies lose people to life events, so recruit a few extra.
- State exactly what an entry needs (how many, what to include, by when) and tie any thank-you to completion, explained up front.
- Choose the simplest tool that captures what you need. NN/g: plain text answers can go through a messaging app (WhatsApp, Telegram, Messenger); richer answers (screenshots, set questions per entry) suit a form (Google Forms) or a doc.
- Pilot it with 1-2 people for a few days to fix instructions and tools; pilot data is not analysed.
- Pre-study brief (short call or written note): what to report, how often, the thank-you, questions answered. Post-study interview: walk back through their entries, ask about gaps and changes.
- Analyse over time: go back to the research questions, then look at how behaviour evolved, what triggered it, and the whole journey around the product.
Entry template (keep it under 2 minutes)
- What did you do just now? (one line)
- Where were you and on what device?
- What prompted it?
- Anything that got in the way or annoyed you?
- Optional screenshot.
How we apply it to Cedwyn (PROPOSAL; any contact with readers or members is a proposal for the owner's approval)
- Research question: what brings an academy member or e-book buyer back in weeks 1-3, and where do they stall?
- Who and how many: 6-8 volunteers from existing members/buyers (small, fairly similar group: NN/g's 5-12 band), recruiting 8-10 to allow for dropouts.
- Trigger and length: event-based ("each time you open an e-book or the members' area") plus one interval prompt on day 3, 7 and 14 for people who logged nothing. Expected use is once or twice a week, so 3 weeks.
- Tool: a 5-question Google Form linked from the prompt; no new software, nothing installed.
- Thank-you: non-monetary (a credit in the acknowledgements or an early look at the next title); no cash and no price changes.
- Pilot: one volunteer for 4 days first.
- Output: a week-by-week journey note (triggers, devices, blockers, drop-off moments) feeding the onboarding file (academy-member-onboarding-first-week-activation.md) and the Kano feature list (kano-feature-survey-better-worse.md).
- Privacy: entries stay internal; publish only anonymised themes, never names or contact details.
Sources
- NN/g, "Diary Studies: Understanding Long-Term User Behavior and Experiences" (Kim Flaherty, 29 Mar 2024): triggers, length table, participant bands, tools, pilot, brief, analysis: nngroup.com
- NN/g video, "Field Studies vs. Diary Studies": nngroup.com
Raw Markdown: diary-study-longitudinal-reader-behaviour.md · version and changes: changelog · machine index: agents.json