# Pilot Study

A trial run of a research procedure used to find and correct problems before the main study begins.

- Type: method
- Domain: Research & behavior
- Level: Foundational
- Last reviewed: 2026-09-25
- Original: [Soroush’s Product Wiki](<https://wiki.soroush.uk/topic/pilot-study/>)

## What it means

A pilot study rehearses a research session, instrument, task set, prototype, data capture, and logistics under conditions close to the planned study. It evaluates whether the research design works as intended. It differs from a pilot program, which exposes a product or service to limited real-world use before broader rollout.

## Why it matters

A broken link, ambiguous task, inaccessible consent form, unrealistic session length, or missing prototype state can waste participants and compromise an entire study. A pilot catches these problems while the team can still revise the protocol and protects the consistency needed for credible evidence.

## When to use it

Run a pilot before expensive, remote, quantitative, sensitive, multi-facilitator, or operationally complex research and whenever the team uses a new method or tool. A lightweight internal rehearsal may be enough for a familiar low-risk round; use a representative pilot participant when comprehension and realism need testing.

## How it works

1. Prepare the complete protocol, materials, technology, roles, timing, and data-capture process.
2. Run the session from recruitment handoff through closing without skipping familiar steps.
3. Observe ambiguity, leading prompts, technical failures, accessibility barriers, timing, and missing states.
4. Debrief the pilot participant and research team, then revise materials and responsibilities.
5. Repeat critical portions when changes are substantial and exclude pilot data from the main dataset unless inclusion was planned and justified.

## A practical example

Before an unmoderated checkout benchmark, a team pilots the study with two relevant users. One task exposes a coupon code that reveals the intended route, mobile recording omits error messages, and the survey takes longer than estimated. The team repairs all three before recruitment opens.

## Where it breaks

A successful pilot cannot expose every participant, device, facilitator, or field condition. Using colleagues may miss jargon and accessibility problems. A very small pilot is not evidence that the product hypothesis is true, and frequent protocol changes during the main study require separate judgment.

## Common mistakes

- Testing only the prototype while skipping consent, instructions, recording, and handoffs.
- Using a team member who already knows the intended answers as the only pilot participant.
- Treating pilot responses as substantive findings rather than evidence about the study design.

## Visualization

[View the visualization on the website](<https://wiki.soroush.uk/topic/pilot-study/#visualization-pilot-study>)

## Sources

1. [Pilot Testing: Getting It Right \(Before\) the First Time](<https://www.nngroup.com/articles/pilot-testing/>) — Nielsen Norman Group · Professional guidance
2. [Checklist for Planning Usability Studies](<https://www.nngroup.com/articles/usability-test-checklist/>) — Nielsen Norman Group · Professional guidance

## Related topics

- [User Research Plan](<https://wiki.soroush.uk/topic/user-research-plan/>)
- [Usability Testing](<https://wiki.soroush.uk/topic/usability-testing/>)
- [Discussion Guide](<https://wiki.soroush.uk/topic/discussion-guide/>)
