In the broader sense, both are probably correct. The core point here is that if people are buying a solution in general (somewhere!), it is a good starting point for own product. In the case of app stores, if there’s a group of apps that are solving a certain problem and are doing well, you can see what you can add to the space. As with any product, differentiating yourself by adding value is always a good strategy. straight mimicking nothing I’d ever suggest, but raising the bar much higher is always a good idea. Everyone wins except the lazy developer in that game.

So, I’d say look to the apps that are doing well in the store first. That will tell you what’s actually being bought and sold today. After that, there could be other solutions out there that do well to solve the problem that just don’t exist in the store at all too. For example. if no one had built a scientific calculator app, but people obviously buy millions of them, that would be an obvious opportunity. Different market, same solution. I’d look first to the app stores though since it should be the most predictive in this case.

So, if people are buying a particular app at a competitor’s store that makes it a good business strategy, and one that you should mimic (improve upon)? Or are you saying that if people are buying a solution for a product in a medium *other* than an app, then we should create an app for that? (Or both ideas)?

I should do a post on App Store rankings because there’s quite a lot to talk about. For any search in the app store, the sorting order corresponds to popularity. The app at the top will be the most popular (and therefor highest selling). There is a “long tail” or logarithmic drop off of sales for each position. For example a simple one would be if #1 is 100 sales/day, #2 will typically be 50 of that, #3 <25 , #4, <12.5… etc. and so on. Some drop off faster, some less so, but that’s a good basis to work from.

If you look at the overall categories, each one is its own vertical or business niche. Each search is its own vertical. Find an app you think is selling well, imagine the keywords people may use to search for it, execute those searches and score each app returned in order according to a logarithmic scale to estimate total sales in that vertical.

I’m advising for an app store market research / analytics company that’s playing chess way above this level. Fascinating stuff.