Market research for an app idea: 7 steps, templates and a worked example

What is market research for an app idea, and how is it different from validation?

Market research is desk work. You look at what already exists and what is already being said online, and you do some arithmetic. Validation is people work: you talk to those who have the problem and watch what they do. You need both, and the order that saves the most time is research first, validation second, because research tells you whom to talk to and what to ask.

When you finish this page, move on to the guide to validating an app idea. If you want to check how complete your description is first, the free Idea Checker points out gaps.

How do you do market research for an app idea?

Work through these in order. Keep everything in one document so the steps can feed each other.

1. Define the user and the problem in one sentence. “For [specific person] who [has this problem], this [does one thing].” If you cannot write it, you cannot search for it.
2. List every alternative. Not just apps. Include spreadsheets, paper, a group chat, a person, an agency, and doing nothing.
3. Mine the reviews. Read what customers say about the alternatives, especially the middle ratings.
4. Check demand signals. Search interest, community threads, and people asking for the thing.
5. Size the market from the bottom up. Count reachable people and multiply by a price you can defend.
6. Check the price. Find what people already pay for the nearest alternative.
7. Decide. Build, narrow, or stop, and write down which assumption would change your mind.

How do you find competitors for an app idea?

Search in the places a customer would search, using the customer’s words, not yours. Do all of these and note every name:

• Web search for the problem phrased as a person would ask it, plus “app”, “template” and “alternative”. Read the first two pages.
• The App Store and Google Play. Search three or four phrasings. Note the top results, the ratings count and the last update date.
• Reddit and forums. Search the problem in plain words. People often name the tool they cobbled together.
• Spreadsheets and templates. If a popular spreadsheet template exists for the job, that is your real competitor.
• Indirect competitors. What does a person do when they do not use a product? A phone call, a trip, a printout.

A crowded market is usually good news: someone pays for the problem. An empty one needs a second look. Either nobody has the problem, or nobody has found a way to charge for it. See app ideas that don’t exist for the difference between a real gap and a dead end.

Name | Link | Who it is for | What it does | Price | Last updated | Rating and count | What users complain about | What it does NOT do
-----
1. [Direct competitor] | | | | | | | |
2. [Direct competitor] | | | | | | | |
3. [Indirect: a spreadsheet or template] | | | | | | | |
4. [Indirect: doing it by hand] | | | | | | | |
5. [Adjacent product that could add this] | | | | | | | |

After filling in: the gap I can serve better is ______ for ______ because ______.

How do you use reviews to find what is missing?

Reviews are free interviews. People write them when something delighted or annoyed them, in their own words, and the words are what you will use on your landing page.

1. Open the store page or review site of your three biggest competitors.
2. Read the three-star and two-star reviews first. One-star reviews are often about a bug or a billing issue; three-star reviews usually say “good, but…”.
3. Copy twenty reviews into a sheet, one row each, with the quote, the star rating and a tag such as price, missing feature, hard to use, unreliable, support.
4. Count the tags. The most common complaint that a small product could fix is your opening.
5. Save the exact phrases. Use them in interview questions and in the first line of your pitch.

Quote | Product | Stars | Tag | What they wanted instead
------
"[exact words]" | [product] | [1-5] | [price / missing feature / hard to use / reliability / support / other] | [what the person says they wanted]

Totals after 20 rows:
- Most common tag: ______ (___ of 20)
- Second most common: ______ (___ of 20)
- Complaint a first version of mine could fix: ______
- Complaint I cannot fix and should not promise to: ______

Be careful with the sample. Twenty reviews is a signal, not a measurement, and people who write reviews differ from people who do not. Treat the count as a reason to ask better questions, not as a number to quote.

How do you check whether people actually want it?

Look for people asking for the thing without being prompted. Three free sources give you signals:

• Search interest. Google Trends scales results from 0 to 100 as a proportion of all searches in a place and time, so it shows direction and comparison, not absolute volume. Compare your phrase against a term you know is popular. A flat or rising line is a good sign; a falling one is a warning. Google itself notes that Trends is not a scientific poll.
• Autocomplete. Type the start of the problem into a search box. The suggestions show how people finish the sentence.
• Community threads. Count how many separate people described the problem in the last year, and whether the replies say “me too” or offer a workaround. A thread with fifty “same here” replies is stronger than a hundred likes.

You can also read how people describe problems on the Post your build ideas board. Each category page lists live ideas, which are one way to see problems phrased by strangers. Some ideas on the board were posted by the site’s operator to seed it, so use the board for phrasing and inspiration, not as a count of demand.

How do you estimate the size of the market for an app?

Use the bottom-up method: the number of people you can actually reach, times the share who would pay, times a price. Top-down figures from analyst reports (“the pet app market is worth billions”) are easy to find and nearly useless for a first product, because you will not sell to the whole market. You will sell to a small, reachable group.

Here is the arithmetic on an invented example. The numbers below are assumptions made up to show the method. They are not data about any real market.

What does this tell you? That the idea is a small side project, not a company, and that it needs a cheap build and a free way to reach owners. That is a useful result. If you wanted a business at USD 20,000 a month, you now know the first version must reach about 25 times more people, or sell something that costs more.

Make each assumption an item on your list to test. The percentages above are guesses that your interviews should replace with real answers. Official population and household counts are available from sources such as the U.S. Census Bureau’s data portal when your market is defined by geography.

How do you find the right price from research?

Do not invent a price. Read the one that exists. Look at the pricing page of each competitor, at what the spreadsheet template sells for, and at what the manual workaround costs in time or money.

• Anchor on the nearest paid alternative. If the closest product charges USD 10 a month, a first-version tool at USD 30 needs a clear reason.
• Count the free option. If a free spreadsheet does most of the job, you are competing against free.
• Check who pays. The user, their employer, or a parent? The answer changes the price you can ask.
• Write down the price as an assumption. Then test it by asking interviewees what they would pay, and by whether they offer to pre-pay.

For pricing a small software product after you have evidence, see how to price a micro SaaS.

What do you do with the research?

Turn it into a decision, not a report. One page is enough.

Idea in one sentence: For [person] who [problem], this [does one thing].

Who has the problem and where I can reach them: ______

Top 3 alternatives and what they cost: 1) ______ 2) ______ 3) ______

Most common complaint about them (from 20 reviews): ______

Demand signals I found (links): ______

Bottom-up estimate: ___ reachable people x ___ would pay x USD ___ = USD ___ a month (assumptions: ______)

Biggest unknown: ______

Decision: build / narrow / stop

If I build, the first version does exactly: ______

I would change my mind if: ______

If the decision is to build, paste the summary into the AI Coding Prompt Builder or into the structure from the guide to writing an app build prompt. The research becomes the context the agent needs.

What are the common mistakes in app market research?

• Quoting top-down market sizes. Nobody buys from the whole market.
• Counting only apps. The strongest competitor is often a habit or a spreadsheet.
• Treating no competition as a good sign. It may mean there is no market.
• Researching forever. A weekend is enough to decide whether to start talking to people.
• Skipping the stop decision. A research process that cannot end in “no” is a way to feel busy.