18
votes
How do you think software services should be monetized?
A year ago, I asked if people would pay for social media platforms and search engines if they could guarantee no data collection and no ads (although in hindsight, I wanted to ask people for basically all software services) and people overwhelmingly said no. Given how Facebook is dealing with the election and YouTube has taken control of monetization for the sake of more advertisements, I wonder what do people think is the right way for software makers to make money.
You might find some of Nadia Eghbal's research on open source software interesting, funding is one of the issues she has looked at.
Are you interested in services specifically or software in general?
There are a lot of properties and goals one may like in a funding model, depending on one's personal taste, but I think a goal one would want from any model would be "generates a steady stream of money sufficient to cover operating costs". I would break down operating costs into the labour costs of the engineers, operators, and other personnel which are essential to producing and distributing the software ('cos people gotta eat, don't they?) -- which would include sales and marketing staff if their efforts are required to generate enough money to cover that baseline [1] --; and the material costs of the products and services which are used to support the labourers, their work, and provide distribution for the software (which would include bandwidth, storage, servers, ...).
If you can meet this minimum viable outcome with a few potential models you can choose the model that best suits your other goals, and tweak the parameters accordingly. Businesses usually want to maximise profit, so they are inclined towards SaaS models (which justifies charging a subscription while giving them maximum control over access to the system and its features), and they will set their pricing to maximise
revenue = num_customers * price_per_customer(ignoring customer acquisition costs, but let's keep things simple).Other goals you might want to optimise for include 'fairness' (open to discussion what that means, usually it means some people get the software for free or very cheap, their use being subsidised by other users who pay more), accessibility (by which I mean, the widest number of people can get the software with the least amount of effort), contributor base, user base (particularly in early stage VC backed startups, profitability is not as important as proving product-market fit, finding and generating demand, and capturing as much of that market is possible -- VCs want big payoffs from unicorns, not just a steady stream of capital (yay)). I'm sure there are more, would love to hear if people have any suggestions.
Here are some funding models I can think of. I won't go into any analysis yet, although I would love to create a thorough document on (software) funding which lists all the models, what outcomes they (don't) work well for, how to tune them for your goals, etc. Shit is hard and we need more resources, and we definitely need more ideas for _new_funding models.
And of course, Nadia has already put together a good resource like this: https://github.com/nayafia/lemonade-stand
[1] I'm not trying to be callous or cold or anything, I'm just trying to elucidate the absolute minimum requirements for a funding model. Sales people are people too :D
The opening question, "How should software services be monetized?", implicitly assumes that monetization is among the goals of software development. There's one model that keeps getting missed from discussion, mainly because most of us have been steeped in the rhetoric and propaganda of free markets for so long. That's tax-based public subsidy, partnerships, or fixed investments. It's a long-standing observation that free markets fail to actualize a number of different public goods.
The public goods we'd like to have are:
Open-source projects, largely staffed by uncompensated volunteers, have only rarely come close to delivering on these goals. The best of these projects get packaged up by private companies who monetize services around configuration/UI, that otherwise can't be refined and released quickly enough under the charity model.
We know there are many software and hardware projects which receive public subsidies in the form of research grants, military and education spending, and so forth. The results aren't always directly available to the general public, but it's a truism that much of the Internet as we know it wouldn't exist without U.S. public spending on research, standardization, and deployment.
Why is something as fundamental to the global operation of the the Internet as Network Time Protocol struggling for funding? Arguably, it's so difficult to monetize in a purely free market that it absolutely should have public subsidy for ongoing development.
There are non-profit foundations which develop software for public use. We've talked about Mozilla often. Whether their spending and funding sources are optimal is related to the question, why not publicly and transparently fund something as fundamental as a browser, so it's (relatively) free from built-in adware and spyware?
Yes, there are potential issues of trust, graft, waste, national economic and military strategic policy, etc., as with any public spending. The larger the polity funding the investment, the greater the complexity and opacity of public oversight. So it might be best for smaller communities or institutions to fund public software projects, even at the risk of duplicated effort. We already have infrastructures for shared public code.
I would like to see important, infrastructural services get passed into the protection of some sort of international non-profit or conservancy in the long run, with funding coming from nation states alongside ethical advertising and paid account schemes.
In general, I find the model of purchasing a version of the software with some number of updates free to be the best model for everyone involved. The "some number of updates free" can be all minor revisions, or all revisions for the next year, or some other reasonable method. So long as the software doesn't stop working on my machine and lock my data in after the free updates run out, then everyone (software creator and software user) should be happy.
I'm vehemently opposed to using ads to support anything. It perverts incentives and makes everything worse. I'm all for software creators trying out other methods besides users directly paying for their software, so long as ads are not involved. Few of these alternatives have worked out for either software writers or users, but honestly I don't think there's been much innovation in this area.
I am personally fine with free tiers subsidized by advertising with a premium subscription option that disables those ads for web services.
For things like video streaming, I don't particularly see anything with subscriptions; an alternate would be usage based pricing, but to be honest it seems a little excessive. For "pro" software, I think subscriptions or a one time payment are fine.
For social media: A subscription-subsidized public commons. Any service requires regular income to function to pay the bills, even if it's a federation of self-hosted instances. Everybody needs to keep the power on and the doors open. I think if this could be done, but like many "solutions," we would have to start a service for it somehow.
For software, I like the model Kjetil Matheussen uses for Radium, his next-generation tracker. You subscribe for support priority, but if you drop your subscription, you still get access to builds, and can build the source-code for free. I don't know what sort of numbers he's pulling, but it seems to work alright for him. The thing is, I think this could work with many other programs, even proprietary ones.
I think the major goal should be direct user monetization, however. If your site has 1,000,000 users, and you can get 1/10 of them to put up $5 a month, you've got 500,000 coming in, which would be an incentive for, say, Facebook to change their monetization, but this would only work with a widely popular grassroots site, not a publicly traded behemoth like Facebook. If your site is only a link aggregator, you can pay for a couple of engineers and a significant amount of server infrastructure.
As a freegan, I am against monetization of my projects on principle.
But how is your project to grown, then, you may ask?
Alternatively: How would someone who cares about it support it?
My answer to those questions is the same: I want growth to happen through users setting up and maintaining their own instances which connect together.
This will bypass myself as a bottleneck in resource management.