Research as a short sprint
Research does not need a budget or a report of fifty pages. It needs a question and a deadline. A good beginner question is "Why do people choose or avoid this kind of service?" Give yourself a weekend, collect what you can, and write one page of findings. Then decide what to change.
The sprint has four parts: talk to people, look at alternatives, read what people write, and explore search language. All can be done free.
1. Talk to a few people
Aim for five conversations of fifteen to twenty minutes with people who match your target customer. Past customers, friends of friends and members of relevant communities (with their permission) are fine starting points. If you pay or reward people, say so honestly.
Good questions ask about the last time something happened:
- "Tell me about the last time you needed this. What made you start looking?"
- "What did you try first? What happened?"
- "What almost stopped you?"
- "How did you decide in the end?"
Avoid selling during the call, and avoid questions that suggest the answer ("Wouldn't it be great if...?"). Take notes and, with permission, record. Mention what you will do with the notes. If you are in a country with data protection rules, keep personal details minimal and secure.
2. Review the alternatives
List five to ten ways your customer could solve the problem: direct rivals, substitutes and doing nothing. For each, note in a table.
Alternative: Template website builder
Promise: "Build your own site today"
Price signal: monthly subscription (check current plans yourself)
Strength: quick, cheap start
Weakness seen: owner must do all the work
Gap for us: done-for-you set-up with plain-language supportYou are looking for gaps and for expectations that you must meet. You are not looking for text to copy. Copying competitor wording is poor ethics, weak positioning and may breach copyright. If you want to track how a competitor's page changes over time, the Text Diff tool can compare two pasted versions for your own notes.
3. Read what people write
Customers leave clues in public: reviews of similar products, questions in forums, replies in comments. Read thirty or forty and note the recurring words and worries. Use them to understand language, not to quote individuals without permission. If a phrase appears again and again, such as "I did not know where to start", your page should speak to it.
4. Explore search language
Type a few phrases into a search engine and see what it suggests and what questions appear. Search results pages often show related questions and related searches. These are clues about what people want to know. Keyword tools exist, many with free tiers, but their volume numbers are estimates, so treat them as rough guides. Search Console, once you have a site, shows the queries that actually led to your pages.
Write one page of findings
Summarise in a fixed structure so you can act on it.
Question: Why do small clinics delay setting up online booking?
Evidence: 5 interviews, 12 competitor pages, 40 forum posts
Findings: 1) fear of patient data leaks 2) no time to learn tools
Surprises: three clinics already use a free messaging app for bookings
Decisions: add a privacy section to the page; offer a setup call
Still unknown: how clinics choose between us and a built-in booking toolThe last two lines matter most. Decisions turn research into action, and "still unknown" keeps you honest about its limits.
Bias and limits
Five conversations are not a statistical survey. You will hear loud opinions and polite ones, and you will tend to remember what supports your idea. Reduce bias by writing notes before you interpret, by asking neutral questions and by looking for disagreement. When findings matter a lot, such as a large spend, test them with a small experiment, covered later in conversion optimisation.
Next you will turn the findings into a compact picture of who you serve in audience personas.
Keeping a research log
Create one simple document for the sprint and add to it as you go: the date, the source, what you saw and what it made you think. Keep your observations separate from your interpretations, because they get mixed up quickly. At the end, highlight the three findings you are most sure of and the three you are least sure of. The first list guides your decisions now; the second tells you what to test next. A research log also stops you from repeating the same work six months later because you have forgotten what you learned.
Common mistakes
- Asking friends who will be polite. Include some people who do not already like you.
- Researching forever. Set a deadline and a one-page output.
- Copying competitors. Learn from the market; write your own words.
- Treating tool estimates as exact. Keyword volumes are modelled, not counted.
- Not acting on findings. If nothing changes, the research was wasted.
Exercise
Write four interview questions for owners of small bookshops (fictional research for a bookkeeping service). Each must ask about a past event rather than a hypothetical.
Show one possible answer
1. Tell me about the last time you had to prepare your accounts. What was the hardest part?
2. What did you do the last time you could not find a number you needed?
3. When did you last ask someone for help with bookkeeping, and how did you choose them?
4. What happened the last time you missed a filing date, or nearly did?Each asks for a specific memory. None suggests a product, so the answers are less likely to be politeness.
Key points
- Small, focused research beats a big study you never finish.
- Customer conversations about past behaviour are more reliable than opinions about hypothetical features.
- Competitor review is for understanding choices customers have, not for copying.
- Search suggestions and community questions show language and worries, but they are clues and not measurements.