Blog / Research

User Research Plan Template: Structure Any Study

By Vadim Glazkov June 9, 2026 14 min read

A practical user research plan template covering every section you need — research questions, methods, participant criteria, timeline and success metrics — so

On this page

Why a Research Plan Is Worth Writing Before You Recruit Anyone

Studies without a written plan drift. Scope expands mid-recruitment, sessions attract the wrong participants, and the data that comes back answers questions nobody actually asked. By the time you notice, two weeks and a recruitment budget have gone on material you cannot use.

A research plan solves this before it starts. It forces three decisions teams habitually defer: what question the study must answer, who can actually answer it, and what counts as a useful result. Thirty minutes of planning costs far less than re-recruiting a fresh cohort because the first screener was too broad.

The plan doubles as a stakeholder alignment document. When a product director, a designer, and a researcher all read the same page and sign off, late objections — “but did you ask about X?” — largely disappear. Disagreements surface early, where they are cheap to resolve.

Treat the plan as a living artefact, not a gate. Update it when something changes: a research question sharpens, a session has to move remote, a segment turns out to be unreachable. The revision history becomes a decision log for the whole study.

For teams building scalable research operations, the plan is foundational. A library of well-structured plans makes it possible to onboard new researchers, reuse screeners, and spot gaps in what the organisation has and has not studied.

The 8-Section User Research Plan Template (with Guidance Notes)

The template below covers every field a study needs, organised into eight sections. Each section maps directly to a field in the working document — nothing is decorative. The structure works for qualitative studies (interviews, usability testing), quantitative studies (surveys, analytics review), and mixed-method designs where one approach feeds the next.

Scan the sections now to orient yourself, then use the guidance notes to fill each one in.

Section 1 — Research Background and Business Context

This section answers one question: why does this study need to happen now?

Describe the product, feature, or decision the research is informing. Note what stakeholders already believe or assume — capturing existing hypotheses prevents the study from simply confirming what the team wanted to hear. Connect the work to a specific business goal or OKR so it has a clear sponsor and a clear beneficiary. Name who commissioned the research and who will act on the findings.

Three to five sentences is usually enough. If you need more, the scope may be too broad. Avoid internal jargon; a new team member should be able to read this section and understand why the study matters without any background.

A clear context section makes turning research findings into decisions considerably easier, because the link between evidence and action is established before a single session takes place.

Section 2 — Research Questions (Not Interview Questions)

Research questions (RQs) are strategic. Discussion guide questions are tactical. Conflating the two is one of the most common planning errors we see.

A research question defines what the study must learn. An interview question is one prompt you might use in a session to gather evidence relevant to that learning goal. The distinction matters because RQs gate method choice: you should not select a method until you know what you are trying to understand.

Good RQs tend to open with How…, Why…, or To what extent…. Aim for two to four core RQs. More than four usually signals that the study is trying to do too much.

Weak: “What do users think of the new dashboard?” Stronger: “To what extent does the revised dashboard help operations managers identify anomalies without additional support?”

The weak version is vague and evaluative in a circular way. The stronger version names a specific user, a specific task, and a condition that defines success.

Once the RQs are stable, use them to drive structuring a user interview discussion guide — each section of the guide should trace back to at least one RQ.

Section 3 — Methodology and Rationale

Name the method or methods, and justify each one against the research questions. “We will run moderated interviews because we need to understand the reasoning behind decisions, which a survey cannot capture” is a rationale. “We will run interviews” is not.

Moderated sessions let you probe, follow unexpected threads, and observe behaviour alongside stated preference. Unmoderated testing is faster and cheaper but removes that flexibility. Neither is superior; the RQs determine which is appropriate.

Combining methods is often sensible. A screener survey can filter a large pool to the right participants before interviews deepen understanding of the shortlisted group. When the study involves information architecture decisions, a method such as card sorting in UX research has its own planning requirements — stimuli design, category count, open versus closed format — and these belong in this section.

Use a simple three-column field: Method | Rationale | Constraints and Risks. Naming risks (low survey response rate, difficulty recruiting a niche B2B segment) lets the team prepare contingencies before they become problems.

Section 4 — Participant Criteria and Recruitment Approach

Vague criteria produce mixed samples. Mixed samples produce conflicting data.

Define primary screener criteria (the non-negotiables: role, behaviour, experience level) and secondary criteria (nice-to-haves that add diversity within the segment). For qualitative studies, five to eight participants per distinct segment is typically sufficient to reach saturation — meaning additional sessions stop producing new themes. For quantitative studies, sample size follows statistical requirements rather than saturation logic.

Always specify exclusion criteria: internal employees, participants who took part in a study within the last three months, and anyone employed by a direct competitor.

Recruitment channels include a panel provider, your own CRM or customer database, in-product intercept, or an external agency. Each has different lead times and cost profiles. Document the incentive and the person responsible for recruitment.

Use a five-column field: Criteria | N | Source | Incentive | Owner. More detail on channel selection and screening questions is covered in how to recruit participants for user research.

Section 5 — Timeline, Logistics and Roles

Break the study into phases: discussion guide design → recruitment → fieldwork → synthesis → readout. Assign a date range and a named owner to each phase, and note any dependencies (the discussion guide must be approved before sessions are booked, for example).

Assign roles explicitly. The lead researcher facilitates and carries overall responsibility. A dedicated note-taker attends every session. Observers should have clear rules — typically camera off, microphone muted, no chat to the researcher during the session. A stakeholder reviewer should be named for the discussion guide sign-off stage.

Teams consistently underestimate recruitment lead times. One to two weeks is realistic for a consumer panel; B2B recruitment through a CRM or agency often takes longer. Build that into the timeline.

Use a five-column field: Phase | Start | End | Owner | Dependency.

The most common timeline risk we see: sessions booked before the discussion guide has been finalised or approved. Avoid this by making guide sign-off a hard dependency on the first session date.

Section 6 — Analysis and Synthesis Approach

Leaving synthesis undefined until fieldwork ends is the most reliable way to create a findings graveyard — recordings nobody watches, notes nobody codes.

Specify upfront how data will be captured (recordings, written notes, transcripts, or a combination), who has access to raw data, and which synthesis method you will use. Thematic analysis suits most qualitative studies. Affinity mapping in qualitative research works well when you have a large volume of discrete observations to cluster. More structured studies may use a coding framework derived from the research questions.

Name how themes will be validated. A second researcher reviewing the same data independently and comparing findings reduces the risk of confirmation bias.

The analysis plan should loop back to the RQs directly. If a data point cannot be linked to a research question, it was not worth collecting — or the RQs need revising. Either way, the plan reveals the mismatch early.

Section 7 — Outputs, Deliverables and Sharing Plan

Specify the format of each deliverable: slide deck, written report, one-page summary, video highlight reel, or some combination. Match the format to the audience. An executive team needs a one-page synthesis with clear recommendations. A product team needs enough detail to make scoping decisions. A design team may benefit most from annotated session clips.

Set a repository location — a shared drive folder, a research management tool, or a wiki page — so findings stay discoverable. Research that cannot be found six months later has no long-term value.

Agree on what “done” means. Some teams consider a study complete at the readout meeting; others require written recommendations to be formally accepted before the study is closed. Either is fine; the ambiguity is not.

Use a five-column field: Deliverable | Format | Audience | Due Date | Repository Link.

Section 8 — Ethical Considerations and Consent Requirements

Document where data will be stored, how it will be anonymised, and how long it will be retained before deletion. Reference or link the consent form that participants will sign before any session begins.

State clearly whether sessions will be recorded, and confirm that observers will be disclosed to participants before the session starts. For any study involving sensitive topics — financial difficulty, health conditions, recent bereavement — note what additional safeguards are in place (a debrief protocol, or the option to withdraw without explanation).

For teams operating in the UK, GDPR applies. US-based teams should check CCPA requirements for Californian participants. If the study spans both jurisdictions, confirm which standard governs and document the decision. Consent forms should reflect the applicable regulation explicitly.

How to Adapt the Template for Different Study Types

The eight-section structure is a default, not a prescription.

For a rapid sprint — three days from kick-off to readout — trim to sections 1 (context), 2 (RQs), 4 (participants), and 5 (timeline). Skip formal deliverable planning and run a live synthesis session with the team immediately after the final session.

For longitudinal studies or complex B2B product research, expand every section. B2B participant criteria need a firmographic layer: company size, industry, job function, and technology stack all affect whether a participant is genuinely representative. A firmographic mismatch is as damaging as recruiting the wrong consumer segment. Teams studying retention and churn research for SaaS products will typically need this level of screener precision.

In stakeholder-heavy organisations, add a sign-off field to sections 2, 3, and 4. Named approval prevents the situation where a senior stakeholder objects to the sample composition after fieldwork has concluded.

The division of ownership between an in-house researcher and an external agency also shifts which sections each party writes. Agencies typically own the methodology and recruitment sections; in-house teams retain ownership of context, RQs, and the sharing plan.

A Real Plan in Practice: What We Changed Mid-Study (and Why)

A product team came to us mid-study. They had skipped the research questions section entirely, written a discussion guide based on features they were considering, and recruited participants broadly — essentially anyone who matched a single demographic criterion.

After six sessions, the findings contradicted each other. One cluster of participants was evaluating the product as a primary tool; another was considering it as an occasional fallback. The discussion guide asked the same questions of both groups, which produced data that could not be aggregated or acted on.

We paused the study. Working with the team, we defined two clear research questions — one about decision-making triggers, one about the minimum viable feature set for the primary use case — and rewrote the screener to separate the two user types. The remaining sessions ran with tighter criteria: eight participants across two well-defined segments rather than a further open recruitment.

The revised study produced fewer sessions overall, but the findings were coherent enough that the product team had clear direction within a week of the final session. Stakeholder sign-off, which had stalled on the earlier inconclusive data, came through in a single meeting.

The lesson is not that the original team planned carelessly. Without written RQs, there is no mechanism to notice when a recruitment brief is misaligned with the study’s actual purpose. The plan is a decision log. Each section either holds or it does not — and discovering it does not hold on paper is far cheaper than discovering it during fieldwork.

Which section do you most often skip when time is short?

Common Mistakes Researchers Make When Planning a Study

Writing discussion guide questions as research questions. “What features do you use most?” is a discussion prompt, not a research question. If your RQ section reads like an interview script, revise it.

Setting participant N based on budget rather than saturation. For qualitative work, four participants recruited carefully will usually outperform twelve recruited loosely. Decide sample size by segment logic, then work backwards to budget.

No assigned note-taker. A researcher who is facilitating and capturing simultaneously does both jobs poorly. Note-taking is a named role; it belongs in the plan.

Leaving synthesis until after all sessions are complete. Synthesis quality deteriorates when observations are not processed promptly. At minimum, write up key themes within 24 hours of each session.

Sharing the plan with nobody. A plan only the lead researcher has read provides no alignment. Send it to stakeholders before recruitment begins, and record their responses.

Frequently Asked Questions

How long should a user research plan be?

For most studies, one to two pages covering the eight sections is sufficient. Larger multi-method or longitudinal studies may need more detail, but length should match complexity. A plan that runs to ten pages for a five-session usability study is unlikely to be read, which defeats its alignment purpose.

What is the difference between a research plan and a discussion guide?

A research plan defines the study: what question it answers, who participates, how data is collected and analysed, and what is delivered at the end. A discussion guide is a session-level document — the prompts a moderator uses during an interview or test. The plan comes first; the discussion guide is derived from it.

Can I use this template for quantitative research?

Yes, with adjustments. The methodology section will name a survey instrument or analytics approach rather than an interview format. The participant section will describe a sample size based on statistical requirements. The analysis section will reference descriptive statistics or regression analysis rather than thematic coding. The core structure — context, questions, method, participants, timeline, analysis, outputs, ethics — applies regardless of method.

How detailed should the participant criteria section be?

Detailed enough to write a screener survey directly from it. If a recruiter or panel provider could not build a screener from your criteria section alone, it needs more specificity. For B2B studies, include firmographic criteria (company size, sector, role seniority, relevant tools in use) alongside behavioural criteria.

Who should sign off on the research plan?

At minimum: the lead researcher, the product owner or commissioning stakeholder, and anyone whose team will be expected to act on the findings. In larger organisations, legal or data protection review may also be required before fieldwork begins, particularly if the study involves sensitive topics or new data collection methods.

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

Author

About Vadim Glazkov

Vadim Glazkov is the founder of Glasgow Research and a product research expert working with founders and B2B SaaS teams on customer interviews, JTBD, market validation, and decision-ready research.

View author page