I totally agree you have to have something to sell that users value. And it has to be something they value reasonably highly and that they can only get by paying. I am currently setting up the infrastructure for selling extensions. Time will tell if this strategy works where selling support or seeking donations failed.

I am toying with the idea of writing up what I call the TOAST test (The Open source Application Suitability Test) (see http://lists.nzoss.org.nz/pipermail/openchat/2012-February/009316.html). Are you interested in collaborating? Even though I have ideals it is important to be realistic.

All the best, Grant

That’s your problem right there.

If a user can easily understand and learn the app then they aren’t going to pay for support. Especially if it’s being used at the desktop where organisations aren’t scared of external-facing failure.

Financially successful open source needs to be obscure, complex and under-documented. Preferably, it should run in a business critical space where problems turn quickly into crises. That way even when customers don’t pay for support (and books, (un-)conferences, etc) at the start, they will when things turn bad.

I’d recommend destroying much of your doco, or changing the system so it doesn’t match. Then write an expensive book with the lost information buried in it. Also, build a network of extreme geeks who can charge $$$ to implement the package and hence are ultra keen on it.

Cynical but true.

My experience also suggests that business models based on paid support do not always work well. A lot depends on the product (complexity, installation needs, integration issues), users (students? enterprises? people used to getting all their software for free or people used to paying?), usage pattern (mission critical? downtime costly? heavily relied upon?), and the community/organisation behind the product (already with paying customers? credible international sales and support network?).

Your point about potential customers for the pro version is sobering and a friend has suggested the same. I guess I will find out. But it would be hard for donations to dry up as there are so very few anyway . At least now there will be a reason for Universities, companies and non-government organisations to make payments. If the income is worthwhile then I will keep the community and pro versions separate and probably continue development. If not, I will reluctantly shift my attention elsewhere. I suspect there will be little ongoing development after that point unless the contracting based on SOFA as a reporting tool starts to pick up. There is some promise in that direction so you never know.

I considered extensions but it gets quite complex in terms of setting up the infrastructure to make it all work and to create the market for it. I can’t afford to sink that much time and effort on a gamble. At least with a simple two version model I will soon find out how much users are willing to pay.

As for Software As A Service, and integration with Limesurvey, those could also work but the amount of development time I would need to put in would be considerable. And once again, it would be a gamble as to whether it paid off.

I am hopeful that the pro version approach will generate more income than the existing donation/support approach and allow the project to continue. The amount of work required to trial this approach is manageable so I will get started soon. There have been over 80,000 downloads of the free version so there is reasonable potential here.

From the post it is quite clear that you put a lot of thoughts in this. I am not overly surprised that donations won’t help you to survive. I was always wondering about how many donations come in at open source projects. I am a total fan of open source software for various reasons. But I have also to admit, that I haven’t donated money so far to any of the projects I use. What I have done so far is helping out in 2 of the projects I find most important and where I feel I have something to contribute, as I am not really a programmer (translations, support in the user forums, etc.).

Like many others, I would prefer SOFA to continue as it is. But of course all the work you are putting in, needs to have something coming back on the long run.

I am wondering who the main users are of your software. I guess there are a lot of universities out there using it. They should be able to make a donation once a year for SOFA, as the alternatives might be quite expensive.

I can also understand that private persons won’t make a donation, as they also don’t earn money with the software and might use it quite sporadically, but companies that use the software on a regular basis should be able to make a donation.

A problem with the open core model that I could see is, that the number of persons who are willing to pay for the professional version might be as high as the number of people who so far were willing to donate. And when there is a “professional” version and a free version, users of the free version maybe won’t be willing to donate anymore. But that you will only know, once you tried.

Maybe “extensions” could also be an interesting model. So there could be an extension for specific calculations and analysis, which you pay for, if you want to add it. This might even attract some other developers to get in the boat. There could be a charged a fee for every sold extensions, which is going to the core development. I guess developers would be find with it, because without the core there would be no extension anyway. I don’t know the potential of this of course. It is just a thought.

I don’t know if it is possible, but maybe you can offer SOFA also as SAS? I think some members of the Limesurvey project are doing this. I don’t know how successful this is though, and if you can apply this to SOFA. By the way: I guess that Limesurvey and SOFA would be a great team. Integration of Limesurvey with SOFA would have a lot of potential. Maybe SOFA could do on the fly analysis of the data collected by Limesurvey and offer the reporting backend. I think Limesurvey has quite a user base and there are quite some overlaps in the target group, without any cannibalism for both projects doing something totally different.

Then an additional income source would be courses. Statistics in general with the use of SOFA. You seem to be someone who can make Statistics “simple” (based on my impression of SOFA). There your location might be them main problem. The market in New Zealand is probably too small. Would you be based in Europe, things for courses might be easier, as you could travel around for relatively cheap, or people could reach you relatively cheap.

I hope you find a good solution to keep SOFA alive and if you think the open core model is the only way to go, then be it. Better an open core SOFA than no SOFA I guess.