How to validate an app idea before you build it

01Name the job, not the feature

Write the idea as the phrase someone would type into the store search: “scan receipts”, “plant identifier”, “baby sleep tracker”. That phrase is the niche. A feature list is not searchable; a job is.

If you cannot find a phrase people would type, the app will depend on you finding every user yourself. That can work, but it is a different business from a store niche.

02Is there demand?

Type the start of your phrase into the store search and watch the suggestions. Suggestions appear because people search for them; a job with no suggestions around it is searched rarely.

Then open the top ten results and look at their rating counts. Ratings are the one public trace of how many people installed and cared. If the most-reviewed app in the top ten has a few dozen ratings, demand is unproven, however good the idea sounds.

• Several suggestions around the phrase: people search for it.
• Leaders with thousands of ratings: people install for it.
• Leaders with almost none: nobody has shown it can be sold yet.

03Do people pay?

Open the leaders' store pages and read their in-app purchases. A top tier of a couple of dollars is a different business from a yearly subscription. On Google Play the range is printed on the listing; on the App Store it sits under “In-App Purchases”.

Paid leaders are good news: someone has already proven the job is worth money. A niche where every leader is free and ad-supported usually means you will be too.

04Can you get in?

Look at when each leader launched and when it last shipped an update. A top ten full of apps that have not been updated in nine months or more is an opening; a top ten where several apps launched in the last year and a half shows new entrants can still climb.

A top ten of household names, all updated last week, is a wall. It may still be a good market, but not a good first one.

05Is it moving?

Check back over a few weeks and note how fast the leaders' rating counts grow. Growing ratings mean new users are arriving in the niche now, which is the difference between a trend and a leftover.

One reading tells you size; two readings a month apart tell you direction. Direction matters more for a new app.

06Read the one-star reviews

Sort the leaders' reviews by lowest rating and read thirty of them. Complaints that repeat across several apps are the category's gap: the thing none of them does well. That is your pitch, written by the people you want as users.

A single app's complaints are its own problem. The same complaint about three different apps is an opening.

07Traps that make a niche look better than it is

Some niches pass every check for the wrong reason. These are the common ones:

• The phrase is one app's name. When almost all the reviews belong to one app, people are searching for that app, not the job.
• Licensed businesses. Lending, cash advances, gambling and pharmacy apps need a licence in most countries, whatever the numbers say.
• Brand terms. Ranking for someone else's trademark ends with a takedown.
• One country. A niche that is strong in one small storefront may be noise; check the same phrase in a large one.

08Deciding

Set your bar before you look, or every idea will pass. A workable rule: proven demand, at least some leaders charging, at least one sign of an opening (neglected leaders, recent entrants or a repeated complaint) and none of the traps. Two out of three is a maybe; fewer is a no.