“I will not pay for your weather app!”

This article is about my experience with Nice Weather , an Android weather app available since 2013.

Imagine you are a developer who wants to create an app, and you’re just saying to yourself: “Hey, why not create a weather app!”. Hell yeah, it’s a pretty cool idea! There are many possibilities to create awesome things with the weather data: forecasts, temperatures, precipitations, sunrise and sunset hours, moon phases, etc.

Okay, so, now, what’s next?! Personally, I think it’s always a good idea to search your data provider first. It could be exciting to start by imagining your future nice and awesome UI, but if you can’t find any providers, not sure that your mockups will be very useful…

What do I want exactly? It could be great to have a provider who gives me at least 3 days (5 could be perfect) of forecasts, with, let’s say, a forecast every 3 hours (every 1 hour could be awesome). If I could have the precipitations and some other stuff to add in my app, it could definitely have all the data I need to create something special, something great.

After some researches, this is what I’ve found:

• The free provider (like OpenWeatherMap): Good news, it’s free! But bad news, the data are not very accurate, you don’t have many possibilities, and you can find some strange behaviors (like bad timezones for example…). Who wants to use a weather app with inaccurate data?
• The expensive provider (like WUnderground): Here are the best data you will found, but they are pricy, and when I say pricy, I’m not sure you want to invest at least 150$ per month to have 5 000 requests per day and an hourly 1-day forecast (except if you have a strong business plan!).
• The cheap provider (like Forecast.io or WorldWeatherOnline): Okay, it’s not free, but not too expensive though. So, most of the time, you will have not-bad or pretty-good data (but not all the time nor for everywhere), and you could have enough data to create the app that you want. But remember, it’s not free.

After this quick analysis, you will certainly have chosen the cheap provider because you don’t want to pay a lot, and the data seem to be good enough for you, right? Now, if you don’t want to lose your money for your users, you have to find a way to make money with your app.

And that’s where the main problem of the weather apps appears:

• People don’t like to pay for things that they can have freely.
• The free forecasts data aren’t very accurate (honestly, they can be terrible).
• So, you choose at least a cheap provider, and you will have to pay for the data every month.
• But if you decide to set your app at 1 or 2$, you will probably have a problem after X months, and if you put a subscription (like 1 or 2$ per year), you have to create a fancy crazy awesome app with a lot of features to be sure to have enough users to pay your data.

You probably see what’s the problem here. Even if you think you can create the most beautiful weather app on the App Store & the Play Store, you will have to deal with this: the weather data are a (very) big market. At this point, you probably understand how Yahoo! and Google do to have accurate data in their apps (or cards for Google): they have partnerships with big companies (like WUnderground) to offer the best forecasts to their users, so it’s not easy to compete with these guys…!

Creating a weather app can be a nice way to improve your skills, but remember that it’s a pretty hard category with a lot of competition. Users can be very rough with you if you can’t explain why your app is not free. Try to create something new, something cool & something great to use, and don’t forget the monetization part!