I’ve never really been fond of the advertisement based business model. It’s not just that I find ads annoying, or that I hate being unwittingly turned into “the product”, or that I don’t want my privacy eaten away by companies with questionable intentions. But even more fundamentally, I simply have this unshakeable feeling that if something is valuable, then people should pay for it, directly. Maybe it’s because deep down I’m just a naive engineer, and a part me cannot let go of the romanticized “if you build it, they will come (and pay you)” world.

But people don’t just come, even if you’ve built exactly what they wanted. They are even less likely to pay for it, at least not willingly. Facebook made $34.86 per user in the US in Q4 2018. That’s more than $11 per person per month, for every user on the platform. How many Americans would pay a $11 explicit monthly subscription fee for Facebook? Definitely not everyone. Far from it probably, given that less than 45% of all listeners pay for Spotify Premium. So if Facebook had suddenly adopted a direct payment business model, they would probably be slashing their user base and revenue in half, or more.

There are many reasons for this universal “unwillingness to pay directly” phenomenon, and I will not get into their analysis here. But given this reality, and assuming that we cannot magically alter everyone’s mind to start valuing money differently, even my naive engineer brain must admit that ad based business models solve a really big problem, one that no other free market business model has been able to solve. And that is the crowd subsidization of cost at scale, for the people who really cannot afford to pay¹.

Without ads, the most straightforward way to solve this problem is to do it at the government level: subsidize the services that we think everyone should have access to with tax dollars. This is basically what we’ve done for many other things deemed as basic necessities for life. But top down solutions like this is prone to the usual problems of being slow and inefficient, which make them very ill-suited to solve problems in the technology space. Universal Basic Income is a way to help with this, since it distributes money directly, leaving the rest to the free market, so it should be faster at responding to technological changes. But we are still quite far from having it widely adopted to rely on it as the solution.

Another way to tackle the problem is via the freemium business model. With freemium, the base product becomes free, and you offer a premium version of the product (with more features and functionalities) for a fee. Under this model, the power users, i.e. those who actually find the premium features to be worth the price (and can afford it), essentially subsidize the rest. But this creates two key problems. One is that you inevitably end up with a chasm between the free and paid versions of the product, so not everyone is equal. This could cause serious issues, especially for social network type services and products. The other is that of scale: it’s very difficult for freemium products to reach truly massive scales of adoption. There are of course very large companies based on the freemium business model (e.g. Dropbox), but even these extremely successful products cannot seem to reach the level of adoption as ubiquitous as Facebook.

A variation of the standard freemium model is one where the free and premium products have no functional differentiation, so the payment is strictly on a goodwill, voluntary basis. This eliminates one of the key problems with freemium products, but creates a bigger scaling issue. Many open source projects in the software community is effectively funded by this model, relying on donations to pay for continued development, but these projects almost never grow to be massive due to funding limitations.

Recently, another alternative-to-ads business model caught my eye, which is used by Sam Harris for his Waking Up meditation app. Under this model (which is like a heavily modified flavor of freemium), the default, widely marketed version of the product is paid, but for the people who really cannot afford it, they can explicitly request for free access via email. I think this model probably has an easier way to grow initially, but will run into problems later, once it reaches a certain scale. The main problem I suspect will be the rise in the abuse of the free access system. With that said, I am interested in seeing how this will work out, and I hope that it won’t fall to rampant abuse.

This is not at all meant to be a comprehensive survey of all the ad-free business models. But even so, we can see that every business model probably comes with issues of its own. The ad based business model does seem to be uniquely capable of solving the otherwise intractable problem of crowd subsidization at scale, at least for now. Naturally it comes with its set of problems (hence the “evil” part of “necessary evil”), which has been exacerbated tremendously in recent years by certain advances in technology such as machine learning. Many people have started to focus on the search for alternative models as a result, but I think that as much as people dislike ads, they are here to stay. Therefore we should look more into solving the problems created by the ad-driven economy, rather than trying to replace the model with something else entirely.