Start with one clear question, use free sources first, talk to a handful of customers, and then run a short survey once you know what to ask. Pay for research only when the free methods leave important questions open. This order keeps costs low and helps you learn the basics before you spend money on panels or outside firms.
Start with one clear question
Before you look at any data, write down the decision the research should inform. For example: “Should we add a monthly plan?” or “Which of these three names do customers remember?” A single question keeps the work focused. Without one, it is easy to collect a lot of information that does not change anything.
Step 1: Use free sources first
Free sources are slower to read than a survey, but they show how real people describe their problems in their own words.
Public discussions and forums
Online communities in your industry are a good place to hear about frustrations. Quora, Reddit, and Facebook groups often contain detailed questions and complaints. Read threads before you post anything. Note the exact words people use, because they are useful later for writing survey questions and marketing copy. Keep in mind that people who post publicly are not a random sample of your market.
Reviews, including lukewarm and negative ones
Reviews of competing products show what customers value and what they dislike. Negative and mid-range reviews are often the most useful, because they point to needs that the product did not meet. This works for books, software, services, and physical products. Sort reviews by rating and by helpfulness, and read a range of them.
Competitors’ websites and emails
Look at how competitors describe their products, what they promise, and what offers they make. If you can, subscribe to their email lists to see the messages they send. This tells you what they think their customers care about. Do not assume they have done thorough research, though. Treat their choices as a hypothesis to check. For a fuller method, see these competitor analysis techniques.
Your own data
Your website analytics, email engagement, and sales records already show what people do. Check which pages get the most visits, which emails get clicked, and where people leave the purchase process. Behavior data is often the cheapest and most reliable starting point, because it does not depend on what people remember or say they will do.
Step 2: Talk to a handful of customers
Interviews go deeper than any free source. Even a small number of conversations can reveal patterns, and it is often possible to talk to around ten customers without a budget. Some tips (the guide to conducting in-depth interviews covers more):
- Ask about the last time something happened. “When did you last try to solve this problem?” gets more useful answers than “Would you use a tool that does X?”
- Keep your questions open. Ask “What made you choose that?” rather than “Did our price persuade you?”
- Listen more than you talk. Resist the urge to pitch. Your goal is to learn, not to sell.
- Take notes or record, with permission. Tell people what the conversation is for and how the notes will be used.
The leading and biased questions guide explains how wording can change what people say in an interview as well as in a survey.
Step 3: Run a short survey
Once you know what to ask, a short survey lets you measure how common a view or behavior is. Two groups are especially useful:
- Recent customers, asked about their experience and what almost made them choose something else.
- People who did not buy, asked what stopped them. Even two or three questions can show whether a barrier is price, trust, timing, or fit.
Keep the survey short. Use a free tier of an online survey tool if it covers your needs, and check its limits first. The online survey tools guide explains what to look for. Before sending the survey to everyone, test it with a few people. The pilot testing guide shows how.
Types of research in plain terms
Beginners often meet three types of market research. Here is what each one means:
- Syndicated (secondary) research uses data that already exists, such as government statistics and industry reports. It gives you background and context, and it is not written for your business.
- Qualitative research involves a small number of people in depth, through interviews, focus groups, or observation. It explains why people act the way they do, but it cannot tell you how common a view is.
- Quantitative research involves many people, usually through a survey, and measures how many people think or act in a certain way. It is useful for sizing a market, but the answers only describe the people you reached.
Most beginners need all three, in roughly that order: background, then depth, then measurement.
Free public data
Government statistics are free and often detailed. National census data, household surveys, and business counts are useful for sizing a market. Public libraries often have reference librarians who know business research well, and some libraries provide access to paid databases at no charge. Ask a librarian before you buy a report, because the information may already be available.
When to pay for research
Paid options range from self-run panels to full-service firms.
- Self-run paid panels let you run your own survey with access to a panel of respondents. They cost more but reach people who are not already on your list.
- Full-service firms handle the whole study, from design to report. They are the most expensive option, and the quality of their panel, method, and analysis should be checked before you hire them.
Before you pay for either, be clear about the question and the decision it supports. A paid study with a vague question still gives a vague answer.
A simple four-week starter plan
This plan is an example. Adapt it to your question and the time you have.
- Week 1: Listen. Read forums, reviews, and competitor pages. Write down the words customers use.
- Week 2: Check your data. Review website analytics, email results, and sales records for the patterns that matter to your question.
- Week 3: Talk to people. Hold around five to ten short interviews with customers or people who fit your target group.
- Week 4: Test a short survey. Write three to six questions based on what you learned, pilot them with a few people, then send them to a wider group.
At the end of the month, write a one-page summary. State the question, what you did, what you found, and what you are still unsure about. The unsure part tells you what to research next.
Common mistakes
- Relying on forums alone. Online communities show what loud voices think, not what your whole market thinks.
- Asking leading questions in interviews. “Don’t you love the idea?” gets a polite yes that tells you nothing.
- Sending a long survey. Shorter surveys get more complete answers.
- Skipping the pilot. A few test responses can catch confusing wording before it costs you responses.
- Treating free data as representative. Know who the data comes from before you generalize from it.
For a fuller picture of the process, see the market research survey questions guide. For a checklist of what a complete study covers, see what should be in market research.