Ask HN: Should I release a SaaS with double license or not?

[atmosx]

Hello,

I started a side project for our family business. Now I reckon that it might have value as a SaaS, for others.

I don’t know if I should make the project Open Source or not. I’m keen to opt for a solution where there is an open source version, which you could install it on your server and a ‘pro’ version with additional features which I run on my servers (with API, integrations, etc.).

I would like to have some advice on the matter from people who had to make the same choice before.

Another issue that might arise in case I opt for the double-license is what license should the open source version be released under. I took a look at sidekiq[1] which I think uses this model successfully and uses a proprietary license for ‘pro’ and LGPL instead of BSD or MIT licenses. I don't know why they went with LGPL over the other two though.

Comments and thoughts are more than welcome, thanks!

[1] http://sidekiq.org

[logn]

I made this choice and went with Affero GPL. This theoretically satisfies: (1) other devs building libre software, (2) non-tech-savvy customers who can still buy service, and (3) tech savvy entrepreneurs who want to limit lock-in and also need a compelling reason to spend money (limit licensing liabilities).

Some people are adamantly opposed to AGPL and prefer GPL or BSD. In my view, it's either because they fall into bucket 3 above or they simply prefer freedom to make software non-free. However, my outlook is that AGPL is the only actually free license and it also coincides with my business model (which involves essentially selling extra license rights or SaaS).

[atmosx → logn]

Thanks I didn't look at AGPL, will do!

[mindcrime]

So here are a few thoughts:

1. A "pro" for going open-source: It is a comfort to your customers, knowing that if you go out of business, that they can continue using the software - either by hosting it themselves, or paying someone to do so.

2. If your product is OSS, it is also a near-perfect guarantee against "vendor lock-in" which is a big deal to some people. As the old saying goes "one of the best ways to get people to use your product, is to make it easy for them to stop using your product." IOW, people will be scared to risk investing in using your thing, and getting locked in, when they aren't yet assured of the value it offers. If you make it easier to leave, it's easier for them to make the initial plunge. It sounds paradoxical, but it seems to hold true.

3. Another "pro" is that you may actually get some contributions from the community, which could help you advance your project faster than you can do it alone.

4. A "con" is that some customers may elect to host it themselves and not pay you, instead of using your SaaS. My feeling is that those people were never really your customers to begin with, and represent no big loss. People who pay for SaaS aren't really paying for the software, they're paying for the idea that you will take care of making it Just Work, keep the servers running, handle scaling, backups, etc.

5. Another "con" is that a competitor can use your code and setup a competing service.

All told, I'm a fan of the OSS route, but I'm also an Open Source ideologue, so I'm not exactly unbiased. So take these comments for what they're worth. :-)

[atmosx → mindcrime]

Hello,

Thanks for sharing your thoughts. I'm more keen to open source, because like you, I love the open source ideology.

Point '4' was all I needed to read actually. I'm not worried about point '5' I don't believe my idea is that good on one hand and I'm not afraid of the competition :-)

Thanks again!