4 ms·

Ask HN: Should I make my project open source to get users?

15 points by SunghoYahng 7y ago

I recently launched a web app(B2C). But people say they're worried about putting their data in and being locked it in, and say that if you make it open source, I'll use it. The problem is that this is a commercial project and I have to make a living with it(otherwise I'll starve). So I have to find a way to make money while being open source. If I turn the project into open source, making money will be quite limited. But it doesn't seem wise that I'm more afraid of failing later so that not doing the thing to get more users now, considering that most startups just fail. Should I make the product open source in order to grow now in any way? (Perhaps one way is to get paid only for hosting the web app. But if I make it open source, hosting it is so easy that anyone can do it. If this will be really successful, someone could even distribute the hosted version for free (I'm working on the expectation that this will really be successful, so this assumption makes sense). I'd like to know if there is another way to make money while making the product open source.)

patio11 7y ago
They probably won't use it if you make it open source. Find people who will pay for it.
- SunghoYahng 7y ago
  Thank you. But people didn't seem to hate paying, but to worry about locking their data into the product in its infancy.
  - Raed667 7y ago
    How about a data export tool?
hollasch 7y ago
This may be a model to consider: https://github.com/sponsorware/docs https://github.com/sponsorware/docs Also, this may help you: https://opensource.guide/getting-paid/ https://opensource.guide/getting-paid/
- SunghoYahng 7y ago
  This is interesting. Thank you. But I wonder whether getting sponsorship is a business model that startups can have (and the way to succeed this way seems too long, so I wonder if I can sustain it). In the Github guide, the case of Travis CI is interesting. I'm going to dig it.
gaoryrt 7y ago
> worried about putting their data in and being locked it in Your product got a problem and your consumer offered a solution. AFAIK, consumers' solution solves nothing. You should focus on fixing your problem instead of making it open source.
- SunghoYahng 7y ago
  That's so clear. Thank you. The words "consumers' solution solves nothing" touched my heart. However, I have no idea about how to solve the data retention problem except for being open source.
  - troydavis 7y ago
    > However, I have no idea about how to solve the data retention problem except for being open source. Make it good enough and few people will have that concern, and even fewer will decline to use it because of that. Look at GitHub, Trello, Todoist, and many others; they all started as tiny web apps without much trust.
    - SunghoYahng 7y ago
      Maybe that's right. Thank you very much
    - SunghoYahng 7y ago
      This seems right, provided that I can make the product good enough. It would be a much harder way to make it good enough than to make it opensource and gather users. I should consider whether the early startup even without funding can afford to choose that path.
  - lonelappde 7y ago
    Let the users download their data in a readable format. That's all you need.
rs23296008n1 7y ago
The model I use for some for-profit things I do: People who pay are your customers. Everyone else is free advertising / marketing / product or service testing / feedback group. Note that "who pay" doesn't necessarily include "those who might pay" but can include "those who would pay". The ones wanting your site open source more likely fall into the "might pay" group. I've used a plugin architecture to test the resolve of "might pay" vs "would pay" groups. By exposing APIs for at least import/export I removed the downside of "data held hostage". "Might pay" also frequently turns out to be "won't pay" most of the time. I wouldn't recommend you open source your site unless you see a lot of potential for open source contribution or where review[1] is expected. I would suggest considering some sort of plugin or API for users. This is a good enough bridge for most people. Especially business to business linkages or even b2c. [1] crypto is an example where review is worthwhile.
- SunghoYahng 7y ago
  Thank you for your case. I'm still wondering what is the ground for people who want open source to fall into 'might pay' group. I would like if you explain that part more.
  - rs23296008n1 7y ago
    "Might pay" means no actual guarantee they will turn into a customer. They are a risk you need to investigate. You'll need to dig into actually why open source is a factor for them and what it really means. Are they just wanting open source as a gimmick or are they really wanting to contribute? Is there an actual benefit to you? That benefit needs to outway the cost since you would be bankrolling development. A lot of people mistake open source as some form of insurance. It can be but having the source doesn't mean you can just fork it. Especially when the original tree is under constant development. Bitrot can occur when you don't track from mainline particular features that are seen as essential. Bitrot also occurs when the codebase doesn't change fast enough to match its surrounding ecosystem. If you're running a site and you open source you definitely need to make sure your core benefit transcends that source code. Your staff, your experience etc. Otherwise you are setting yourself up for basic competition from others using your technology. None of this stops open source being useful or a good idea in general. Just make sure there's an actual benefit and you know exactly what that is and what it implies. As mentioned before, I've had customers want the entire product open source but on further inspection I discovered they really just wanted the ability to take their data out. They really wanted interoperability.
mjashanks 7y ago
You could provide an "Export Data" function, and even supply them with regular, scheduled exports. Also - you can still license your product with a closed source license, but allow your customers to run "On-Premise" ... i.e. they can still run your product for themselves, and own their data, but are also paying you a license fee for the pleasure. It's a different revenue model, but can work - afterall, this is how business was done with software, pre SAAS. Often, companies who license in this way offer $xxx.xx per year - getting them support and upgrades for the year. Then, if they cancel after the year is up, they get to keep it - with no upgrades. It's important that your pricing model reflects this scenario. One of the upsides of this model is that you have zero infrastructure costs.
- SunghoYahng 7y ago
  I didn't even know that the web app could be made on-premise. Thank you. I will look for this topic.
artembugara 7y ago
About 3 weeks ago we open-sourced PART of what our API will do. We got 700+ stars on GitHub and 200 signups for beta just from the GitHub. https://github.com/kotartemiy/newscatcher https://github.com/kotartemiy/newscatcher Try to open source some part of your solution. Charge for some additional stuff. The open-source part will make you more credible in front of the clients.
jimmyvalmer 7y ago
Your claim is dubious. People don't balk because of closed-source. They balk because they don't know whether your SaaS is worth relinquishing their personal data. Until we know what it is your SaaS does, it's impossible to advise on open sourcing. Generally, however, unless you've patented an algo, always open source. For the last 15 years, code is cheap commodity, and getting cheaper. The only companies capable of sustaining a proprietary model are big and old, which is not you. You will need to feed yourself from high-touch service level agreements.
pkrotich 7y ago
I'll say sell a virtual Appliance (think GitHub Enterprise) for those who want to self-host. The truth is it's really hard for people to pay for what they get for free. Open sourcing is an option but don't expect to make money right away - it's a longterm game. I have some experience in monetizing open source, if you want to reach out. I developed osTicket (popular ticketing system) and ended up launching hosted version - which is doing well now. Was able to hire a team and we are profitable.
DrNuke 7y ago
Very limited free version asking for very few, non-sensible, non-relevant data to try and convert from the pool of the maybes?
gwbas1c 7y ago
At least in my research, a common business problem is for a potential customer to ask for a feature, the business sinks a ton of resources to build the feature, and then the customer who asked for the feature is long gone. So, IMO, I wouldn't go open source just because someone asked for it. There needs to be a much more compelling reason. That being said: Depending on what your product is, and its architecture, there's nothing wrong with open sourcing part of your product. There's also nothing wrong with open sourcing it at a later date, or providing access to the source code to your customers under some kind of NDA. Furthermore: What is your value proposition? Why do your customers pay you? Is your product really a product, or is it something you should develop on contract for someone? (And provide the source to the customer?) Is it something you really should open source and sell support?
he11ow 7y ago
I don't think your problem is that it's not open-source. Your problem is marketing and positioning. For one thing, the copy on the website is not commercially oriented at all. I sense your earnestness, but copy needs to be clear and compelling. A reader needs to come out of it thinking - YES! this is just what I needed. You start off by mentioning a combination of two tools I've never heard of, and why would I even care, I'm reading about YOUR thing. Look up Pedro Cortés on LinkedIn - he does a lot of video reviews of SaaS landing pages. Two - and just as important - the use case is so abstract, you're counting on your users to imagine how they would use it. Instead, for a knowledge organizing tool, it should be pitched at the challenges people regularly have trying to organize knowledge. (For example, having a gazillion links and tabs. I could imagine this as something really helpful to solve that problem, especially if you can then export it all in some convenient structure [CSV, JSON].) You've got the core of a product, but it's not quite a product yet, that's my impression at least. It could be though! Best of luck
JJMcJ 7y ago
> But people say Does everyone say that? Then you have a problem. Do just a few people say that? Then ignore them, or figure out a way for them to save their data off your site.