Diary Study UX Research: How to Run Longitudinal Studies

On this page

• What is a diary study in UX research?
• When to choose a diary study over other methods
• Designing your diary study: structure, prompts, and duration
• Recruiting and retaining participants without mid-study drop-off
• Running the study: keeping participants engaged day to day
• Analysing diary data: from raw logs to longitudinal insight
• Communicating diary study findings to stakeholders
• Common mistakes and how to avoid them

A diary study can show you what an interview cannot: how a user’s experience shifts across days and weeks, in their own environment, in their own words. This guide covers when to choose the method, how to design and run it, and how to analyse the data without losing the temporal signal that justifies the effort.

What is a diary study in UX research?

A diary study is a longitudinal self-reporting method. Participants log their own experiences — in real time, or close to it — across days or weeks. Instead of bringing users into a lab or scheduling a single interview, you ask them to record what happens as it happens.

Three variants are worth distinguishing:

• Structured diary studies use fixed prompts delivered at set intervals. Every participant answers the same questions on the same schedule.
• Semi-structured diary studies provide a loose framework but let participants describe events in their own words and at their own pace.
• Experience sampling method (ESM) triggers an entry based on an event (“complete this after you next use the product”) or a random time signal. ESM reduces recall lag but can feel intrusive.

The critical distinction from one-shot methods is temporal. User interviews and usability tests capture a moment — what a person thinks or does in a specific, bounded session. Diary studies capture change: how behaviour shifts, how frustration accumulates or fades, how a product is adopted across multiple real-world contexts.

Log format should match what participants will tolerate. Text works for most studies. Photos add spatial context — a cluttered desk, a confusing receipt. Short audio or video clips are richer but create transcription overhead. Don’t default to the richest format; default to the one participants will actually sustain.

When to choose a diary study over other methods

The method earns its overhead when at least one of the following is true.

The behaviour is episodic or infrequent. Expense filing, medication tracking, monthly budget reviews — these happen too rarely to observe in a lab session and too irregularly to study through a single interview.

Recall bias would distort a retrospective account. The emotion of a failed payment or a confusing onboarding step fades quickly. How people remember an experience differs meaningfully from how they felt it in the moment. A diary entry written at 9 pm captures something an interview three weeks later cannot.

You need the full journey arc. For churn research for SaaS, the signal often lives in weeks two and three of adoption — not the first session. A diary study can map where engagement drops, where habits form, and where frustration silently accumulates before a user cancels.

The real environment can’t be replicated. A commute, a clinical setting, a shared household — these contexts shape behaviour in ways a usability lab erases.

Diary studies are the wrong choice when:

• The task is short and single-session (a checkout flow, a sign-up screen)
• Your timeline is under two weeks from kick-off to debrief
• Participants have low digital literacy or low intrinsic motivation to self-report
• You need statistically generalisable data — longitudinal qualitative is built for depth, not breadth

For those scenarios, card sorting in UX research or a standard moderated test will serve you better.

Designing your diary study: structure, prompts, and duration

Start with a single, clear research question. “What do users experience during the first two weeks of onboarding?” is workable. “What do users think about the product?” is not — it produces vague logs that resist analysis.

Once your question is fixed, every design decision follows from it.

Duration. Five to fourteen days covers most behavioural cycles. Past twenty-one days, dropout rises sharply unless incentives are staged across the study. Three weeks of entries from ten participants sounds comprehensive — until you’re down to four participants by day eighteen.

Prompt design. Limit each entry to three to five questions. Mix closed questions (a 1–5 rating of how smoothly something went) with one or two open questions (“describe what happened in your own words”). The closed items reduce effort and allow basic quantification; the open items deliver the texture. If entries routinely shrink after day five, your prompt set is too long.

Entry triggers. Time-based triggers (end of day) are low-interruption but introduce recall lag for events that happened at noon. Event-based triggers (immediately after using the product) minimise recall but can feel intrusive and are hard to enforce. Researcher-initiated push notifications split the difference — you control the cadence, but participants still need to remember the context. Choose based on how time-sensitive your research question is.

Piloting. Run the prompt set with one or two people before launch. Ambiguous wording discovered on day one of the full study costs you clean data from the entire cohort.

Tooling. Dscout is purpose-built for diary research and handles media uploads cleanly. Dovetail supports daily prompts with good tagging. A WhatsApp group or a shared Google Form can work for smaller studies where participants are comfortable with those tools. Choose by participant tech comfort, not researcher preference.

Onboarding. An onboarding call or short video showing participants what a good entry looks like is not optional. Skip it and you get terse, ambiguous logs from the start.

A well-structured user research plan template can help you document these decisions before fieldwork begins.

Recruiting and retaining participants without mid-study drop-off

Attrition is the most common failure mode in diary research. Most of it is preventable at the recruitment stage.

Screen for motivation explicitly. Ask directly: “This study requires you to complete one short entry per day for ten days. Do you have time for that in the next two weeks?” Anyone who hedges is a dropout risk. Screeners that only ask about product usage miss the commitment question entirely.

Sample size. Eight to fifteen participants is sufficient for longitudinal qualitative research. More entries per person consistently generates more insight than a larger sample with thin participation. A participant who completes twelve out of fourteen entries is far more valuable than three participants who each complete four.

Stage your incentive. Split the payment so a meaningful portion is held until study completion. A 60/40 split (60% at midpoint, 40% at completion) works well for ten-day studies. This single design choice reduces attrition more reliably than most other interventions.

Build in check-ins. For a ten-day study, a brief check-in on day three and day seven catches silent drop-outs while there is still time to re-engage them. Group reminders don’t do this — only a personal message to the specific participant does.

Set expectations in writing. The screener and consent form should state the time per entry (be realistic — five to ten minutes), the total commitment in days, what happens if a participant misses a day, and how data will be stored and anonymised.

Recruit across use-frequency bands. If you only recruit daily users, you’ll miss the experience arc of occasional users entirely — and those arcs often contain the sharpest insight about friction.

More on finding and qualifying the right participants: how to recruit participants for user research.

Running the study: keeping participants engaged day to day

The first message matters. Send a warm, personal kickoff note on day one — not a form email — and include a concrete example of what a good entry looks like. The example sets the quality bar without making participants feel judged.

From day two onwards, lightweight daily nudges keep momentum without tipping into nagging. If you’re running a group channel, acknowledge entries publicly — a brief reply or a reaction signal tells the participant their effort has been seen. It costs almost nothing and measurably affects continued participation.

Monitor response rates daily. If a participant goes silent for two days, send a personal message rather than a group reminder. “I noticed you haven’t logged in the last couple of days — is everything alright, or is there anything about the prompts I can clarify?” lands very differently from a generic broadcast.

Handle sensitive entries carefully. On a recent fintech engagement, several participants logged real frustration with opaque payment errors — moments where the product failed them with no clear explanation of what had gone wrong. Participants who received an empathetic acknowledgement from the research team continued logging honestly through to the end. Those who received no response, or a generic reply, began producing progressively more neutral entries. The pattern was consistent enough that we now treat researcher responsiveness as part of study design, not optional courtesy. Participants who feel observed but not heard tend to self-censor.

Keep your own field notes throughout fieldwork. Patterns visible in the live data — a cluster of frustration entries on day four, a shift in tone after a product update — are worth capturing as they emerge. Don’t wait for the analysis phase to notice them.

Analysing diary data: from raw logs to longitudinal insight

Begin with a participant-by-participant read-through, not a cross-cutting theme search. Read each participant’s entries in chronological order. The individual arc — how their experience shifted from day one to day twelve — is data. Flattening everyone into a theme map before understanding individual journeys loses the temporal signal that justifies the method.

Timeline visualisation. Plot each participant’s entries on a time axis and colour-code by sentiment or theme. Even a simple spreadsheet version reveals patterns that affinity mapping misses: does frustration cluster in the first three days and then recede? Does it spike again after a specific product event?

Frequency analysis. A pain point mentioned by one participant once is noise. The same pain point appearing across multiple participants across multiple days is signal. Recurring friction carries more weight than one-off complaints.

Affinity mapping. This works well at the theme level, but apply it after individual timelines are understood. Affinity mapping in qualitative research is a powerful synthesis tool — just not the first tool to reach for with longitudinal data.

Quote tagging. Tag each entry excerpt with participant ID, day number, and theme. That lets you reassemble clips chronologically in reports, so you can show a participant’s journey through a single issue rather than a decontextualised quote.

Triangulation. Where you have quantitative data — usage logs, rating-scale items from the diary itself — use it to validate or complicate self-reported patterns. If participants report increasing confidence but their usage logs show declining engagement, that tension is worth investigating.

Behaviour change over time. This is the unique return on the method. Did frustration with a specific feature reduce as participants developed workarounds? Did early adopters’ language shift from uncertainty to fluency? Identifying these temporal arcs is what separates diary findings from interview findings.

Communicating diary study findings to stakeholders

Stakeholders who commission diary research often expect a theme list. Deliver a journey instead.

Lead with the longitudinal arc in any presentation. Frame the opening not as “here are five themes we identified” but as “here is what the first two weeks looked like for users who stayed engaged, and here is where the ones who churned started to diverge.” That framing makes the temporal value of the method immediately legible.

Use participant timelines as a visual anchor. A single slide showing one participant’s journey from day one to day fourteen — colour-coded by sentiment, annotated with two or three direct quotes — communicates more than a dense findings deck. It makes the data feel real.

Anonymised diary quotes carry particular weight in stakeholder presentations because they were written in the moment, not reconstructed in an interview. A quote typed at 11 pm after a frustrating experience reads differently from one recalled a week later.

Frame the findings in terms of what changed, not just what users think. “By day eight, participants had developed workarounds for the notification failures they described in days two through four — but those workarounds added an average of three extra steps to a core workflow” is more actionable than “users find notifications unreliable.”

Map every finding to a product moment or feature touchpoint. Without that mapping, design and product teams have no clear entry point for action.

Guidance on presenting these outputs effectively: communicating research findings to stakeholders.

Common mistakes and how to avoid them

Running too long without staged incentives. Studies that run three weeks or more on a single end-of-study payment routinely lose a large proportion of participants before the final week. Stage the incentive, or shorten the study.

Prompt overload. Ten questions per entry turns a diary study into a chore. Entries get shorter, blander, and less honest as the study progresses. If you find yourself wanting to ask eight questions, ask which three matter most.

Treating all entries as equivalent. A day-one entry reflects novelty and effort to impress. A day-twelve entry reflects familiarity and, often, honest frustration or comfort. Interpret entries in their temporal context — early entries and late entries are different kinds of evidence.

Skipping the pilot. Ambiguous prompts discovered on day one of the live study contaminate all the data that follows. Thirty minutes piloting with a colleague or a friendly user is cheap insurance.

Recruiting on availability rather than motivation. Participants who said yes because they had a gap in their calendar are not the same as participants who said yes because the topic matters to them. The latter complete studies; the former often don’t.

Analysing too early. Starting affinity mapping before you’ve read every participant’s entries in full risks building a theme structure that misrepresents the individual arcs. Finish the chronological read-through first.

---

A clear next step: if you’re considering a diary study for a current project, write your research question in one sentence, then ask whether it requires understanding change over time. If the answer is yes, the method is justified. If the answer is no, a shorter method will give you the same insight with less overhead.

---

About Glasgow Research — Glasgow Research helps B2B SaaS teams turn customer and market research into product decisions. Work with us.