Diary Study

A diary study asks participants to record their own experience over days or weeks, usually through short prompted entries. It exists to capture what a single session cannot: anything infrequent, anything that unfolds slowly, and anything whose context you would never see.

The contrast with an interview is about distance from the event. An interview collects a reconstruction, shaped by memory and smoothed in the telling. A diary entry is written near the moment, while the irritation is still specific.

What it is good for

Infrequent work. A month-end close, a quarterly review, an annual renewal. Nothing that happens twelve times a year can be observed on demand.

Slow-forming impressions. Whether a product is still useful in week three, when novelty has gone and the data has piled up. This is the question most closely tied to retention, and it is invisible in any single session.

Context you cannot enter. Work that happens across locations, out of hours, or in places an observer would change simply by being there.

Emotional arcs. Where confidence builds or erodes across a first month, which nobody can report accurately afterwards because the ending colours the memory.

Comparisons over time. The same task in week one and week four, by the same person.

Designing one that survives

Attrition is the method's characteristic failure, and it is almost always self-inflicted.

Keep entries under two minutes. Three questions, not ten. A form that takes fifteen minutes gets filled in for three days.

Prompt at the moment, not at the end of the day. A notification tied to the event, or to a plausible time for it, collects something specific. "Tell us about your week" collects a summary.

Allow voice and photos. Speaking for forty seconds is easier than typing a paragraph, and a photograph of a desk carries more than a description of it.

Run the cycle, not a fixed fortnight. Two weeks of a monthly workflow is half a story.

Pay properly and say what it involves up front. Underpaying for a three-week commitment produces attrition disguised as bad luck.

Recruit more than you need. Eight to twelve per group, expecting to lose some.

Reading what comes back

Entries are short, uneven and written in a hurry, so the analysis is different from interview transcripts.

Read chronologically, per participant, first. The arc through one person's three weeks is usually the finding. Sorting everything by theme on the first pass destroys exactly that.

Watch for what stops appearing. A participant who mentions a feature daily in week one and never again in week three has told you something no survey would.

Treat gaps as data. Missing entries on the busiest days are a signal about when the product is least usable, not just missing responses.

Follow up with an interview. A diary study pairs naturally with a closing conversation that asks about the entries. The entry says what happened; the interview explains it.

Choosing it over the alternatives

Diary studies get reached for when something cheaper would answer the question, and skipped when nothing else can.

The distinguishing condition is elapsed time. If the answer depends on what happens between sessions — habit forming or not forming, enthusiasm decaying, a workaround being invented in week two — nothing else reaches it. If it does not, something cheaper will do.

In practice

A team wants to know why trials go quiet after the first week, and interviews with churned trials keep returning the same unhelpful answer: "we got busy".

A three-week diary study with ten new trials replaces it with something specific. Entries in week one are enthusiastic and detailed. In week two they thin out, and the recurring note is that the person has to re-import data every time because the product does not keep it between sessions. By week three, half the participants have stopped opening it.

The re-import problem never appeared in an interview, because after a month it had compressed into "we got busy". In the diary it is written down four separate times, in four participants' own words, on the day it happened.

Where teams get it wrong

• Asking for too much per entry. The most common cause of a study collapsing in week two.
• Prompting at the end of the day. A summary instead of an account.
• Running shorter than the cycle. A fragment that reads like a pattern.
• Sorting by theme first. The per-person arc is the thing diaries uniquely capture.
• No closing interview. You have what happened and not why.