An experiment with Donationware

I'd like to share the results of a six-month experiment with a Donationware software title that began under a traditional share-ware model of (Demo, Fee) title. There are, no doubt, quite a few success stories under both models. Owing to the immense learning opportunity from my experiment, I wanted to share the other side to all of the happy-feel-good stories of Donationware.

First thing is first. Donationware is when you provide software with the expectation that a certain percentage of users will donate a certain amount of money to play the game. The reasons vary. If you want to look at this brutally, it is users subsidizing users.

There are several examples on the internet of reasons why to switch your platform or application to a donationware model. However, they fail to mention a couple key bits that are important in the consideration of the activity.

Theory of Community Destruction

This is a theory. Something rather remarkable happened when I switched to Donationware, which is that the overall quality of feedback visibily declined. While there will always be a certain segment of the userbase that you simply cannot please, what can be quite surprising is to see the champions of your application begin to fade away over time. When an application has a fee for a full version it serves to gate the horde out. When anyone can and does gain access, it becomes a commons and only a certain percentage of users try to keep the place kept up--in a manner very similar to a public or shared space. My theory is this. Donationware is a negative pressure on a community that seeds the destruction of the application licensed under this model.

Evidences of Community Destruction

Now, what I am about to say with bring a flurry of comments to my moderated comments section with the counter-point: it's the app, not the users. However, after running a parallel experiment with a completely different application and conducting the two at the same time in different ways, the similarities and differences are quite telling.

Let me show some crude outlines of this by presenting numbers that explain what I see in the spreadsheets:

Application A (Demoware Model)
  • 1,000 demo users
  • $1.00 a day conversion
Application B (Before Donationware Model)
  • 1,000 demo users
  • $1.00 a day conversion
To make a little more sense out of the conversion number, that's the money per day you get for users buying your full version of the application. The dollars here are imaginary to help illustrate where I'm going, as the actual numbers are property of RTS.

So, on any given day for Application A and B, there are $2.00 a day worth of new revenue coming in. Then, users begin pushing to open up one of the applications to a donationware model.

Thus begins the experiment. What do you suspect happens?

The first week after converting to Donationware the application makes an exhilarating amount of money and downloads shoot up fast. This causes new users who weren't targets of the original niche to show up and begin their interaction.

Then, just as suddenly, the bubble pops, but the gates are still wide open.

Donationware is Prayware

You pray your users will pay you. This isn't a reliable way to do business.

Today, Application A consistently brings in revenue. However, Application B is in Donationware-stalemate. The surrounding community fell victim to trolling and, lacking an appropriate counter-troll defense, the gated community was smashed to pieces and fed to the commons. This leads to developer apathy and lack of traction on the part of any technology leads you may have hired for the project.

Making Donationware Work

Contrary to the above, I do believe Donationware can and will work as evidenced by the other success stories on the 'net. However, if you've ever been to a start park with an 'honest drop' box for money for admission, I'd be willing to bet it's highly dependent on the niche. It must be possible to let people build up their own spheres of influence with your product. Be careful not to poke holes in those spheres by dorking around too much with the business model behind it, especially if your application is one of those unfortunate types that require frequent updates to inspire the user base.

If you'd like a discussion point on this for your group, consider this: Compare and contrast donationware software to the tragedy of the commons.