While I am waiting to collect information from the survey, I started thinking about how I could make money with Pra. I thought about it before at a high level. But now I have started thinking about it in more depth.
It is important to understand that my expectations aren’t that this product be some sort of unicorn. I share the mindset of Paul Jarvis — I want to build a successful, sustainable business rather than shoot for explosive growth — and we largely operate our current boutique consulting/contracting business that way. Paul just launched a book titled Company of One where he talks about this mindset. If I boil it down, my expectations for Pra are to be something that:
After talking in depth about this with my co-workers, Ryan and Adam, it seems like I want to have a free tier that provides “real” value. Ideally I want people to find enough value in the free version that they use it over GitHub (or any of the other providers) directly and are willing to tell other people about it.
Then, to make money, I need to find compelling additional features that at least some people would be willing to pay for.
The natural organizational entities are accounts, organizations, and repositories.
There are two main stances I have thought of related to this product that I could take:
My current stance is that I am not 100% sure which route to take. My instinct is that I should take the Core Value Free stance as it facilitates growing a potentially larger following for other products in the future, and I can envision how to make it happen. Also, in the Core Value Paywalled stance I would potentially have to build out a larger more complicated version of the app prior to the paywall being a thing, as it wouldn’t make much sense for people to use it if it was a worse tool than what GitHub already provides.
The above leads us to the following questions:
After hashing it out I think the answer to this question is no. I don’t think it provides/communicates/shows the value of aggregation when it is limited to simply 3 repositories.
Well, I could. However, the largest use case of people are probably people that have one primary Organization (the org of the company they work for), and their personal organization. The question then becomes is there enough value of aggregation seen for them to then purchase a subscription just to get their personal org added in, or would they just deal with the annoyance? After talking about it with Adam and Ryan I think most people would just deal with the annoyance. So, the answer is no.
I also need to keep in mind that I want the largest possible group of people using and finding the free version valuable so that I can convince them to pay for even more great value.
Initially the thinking around this was that I charge for multiple accounts, as it is probably a niche subset of the market (consultants/contractors). However, I know personal developers who have their personal repos in BitBucket and their company repos in GitHub. So, I am not sure where I net on this at the moment. It still needs some consideration.
If I don’t charge based on core organizing principles what do I charge for? Right now the thinking is that I would charge for features. For instance, provide some of the Quick Filters for free and some of them would require you to be a subscriber. I could also make the search functionality, keyboard shortcuts, enriched PRs, etc. paid features.
One thing I don’t like is the complexity around purchases and nickel and diming people. So I think it should basically be two tiers:
One thing I have talked about as a possible 3rd tier would be Enterprise, which would potentially get you integrations with GitHub Enterprise, etc. The thinking around this is some individuals might pay for it because they have become dependent on Pra and love it. But I think the real money with that is in group licensing to companies that use Enterprise GitHub, etc.
Although I don’t have all the answers, my instinct is still to take the Core Value Free stance. I do believe thinking through some of this prior to getting the results from the survey back is a good thing. It helps me look at the data I get back in ways I may not have looked at it originally. Hopefully, the results of the survey will help clarify answers to some of these questions.