Last week, I dove into the economics of the WordPress open-source ecosystem with my call to stop the toxic scorekeeping of Five for the Future, the eight-year-old initiative to encourage companies to dedicate five percent of their resources back to the open-source WordPress project.
Against all odds, that article was popular! So this is the second chapter in a multi-article series to answer the question I posed at the end of my post last week: How do we ensure WordPress lasts forever, and that our work helps increase the reach and value of all open-source software?
In this post, I will present an alternative view of the Free-Rider Problem, which is an economic concept that considers how we should fund public goods from which everyone can benefit, regardless of whether they pay for the production or maintenance of the good.
Later in the series, I will propose a radical, economically efficient new way to fund the future development and growth of the open-source WordPress software.
I appreciate and have been inspired by the work that Josepha Haden Chomphosy has done to improve Five for the Future (and WordPress in general) in her role as Executive Director of WordPress. I hope this series helps move her mission forward – and proves that people really do like economic deep dives!
Understanding WordPress as a ‘public good’
One example of a public good with a free-rider problem is an “asteroid protection” system. If we build an asteroid protection system, someone has to pay for the raw materials and contribute the labor. But everyone on the planet, even those who don’t directly or indirectly pay for asteroid protection, will benefit if the system indeed prevents an asteroid from hitting Earth. The asteroid protection system is a public good, and the people who pay nothing but still avoid death-by-asteroid are the free riders. (If you’d like to read more about this, check out this Investopedia entry as well. I do not recommend the Wikipedia entry, it is a mess.)
We can agree that WordPress is a public good – it is free for anyone to use, there is effectively no limit to how many people can benefit from it (the software can be copied an infinite number of times for free), and it requires work to produce and maintain. Where I diverge from WordPress co-founder Matt Mullenweg’s analysis is in the definition of who in the ecosystem is the “free rider.” Mullenweg frequently cites the Free-Rider Problem to talk about web hosts and other large companies that could contribute more, as he did in this June 24 post in the Post Status private Slack, as part of his follow-up to the GoDaddy tweetstorm:
“One of my favorite areas of study in economics was the Free Rider Problem, which I encourage people to learn about, and then read the rest of my comment here: https://en.wikipedia.org/wiki/Free-rider_problem [Rob’s note: the Wikipedia entry is very confusing, use the links above instead!]
It’s what drew me to open source, actually, because it seemed in technology you could trade the scarcity of atoms for the abundance of bits. Failed open source projects usually succumb to the free rider trap — the parasites kill the host, which ultimately hurts the parasites as well but they can’t think beyond short-term. Successful open source projects escape the free rider problem, as WordPress has so far, largely because of awareness of it and people voting where to invest their talent and their dollars in organizations that contribute to the shared resource in a way that keeps it sustainable.Matt Mullenweg on GoDaddy and the Free Rider Problem in Post Status Slack, June 24, 2022
Let’s say there were two companies that made [Search] Engine Optimization plugins, when you spent a dollar with Company A, they put 5-10 cents back into WP in a way that benefitted not just them but the entire community; when you spend a dollar with company B they put .01 cents back in, and used the other 4.99-9.99 cents to buy more ads, sponsor more newsletters, etc to compete and take customers from Company A. Do you see the problem? If the 10 cents of contribution from Company A isn’t supported by customers more than the 9.99 cents of advertising of Company B, eventually A goes away. Fortunately, people make decisions on who to [do] business with based on more than just advertising spend. (Though at some level, it’s insurmountable if the product is halfway decent.) And many people choose to work at companies like Company A which hopefully allows them to execute better. But this requires broad awareness of the problem.
I don’t want to talk about GD [GoDaddy] anymore, but I really encourage people to consider the above. One question I like to ask is to assume a company’s strategy works, fast-forward ten years, what does the world look like.”
My takeaway is that Mullenweg believes that Company B is a free-rider and Company A is not. That is possible – the definition of “free rider” includes someone who doesn’t pay “enough” – but I think Mullenweg is conflating his arbitrary Five for the Future threshold with the actual economic definition of free riding. Economics has no five-percent threshold; it wants the market to decide what is “enough.” That is not to say that five percent is the wrong amount, but there is no proof it is the ideal amount.
If an economist looked at this problem, they would say: Company A and Company B are both contributors. They are both receiving value from the public good and giving back to its production and maintenance, albeit at different ratios. But there is another huge group that gets tons of value from WordPress and does not contribute anything back to it – the actual end-users!
In other words, my clients are free riders, in the economic sense. They get tons of value from the public good of WordPress, and they give nothing back! You are also a free rider if you build sites with WordPress and do not contribute back to the WordPress open-source project. The only people who are not free riders in this economic scenario are the people who are producing and maintaining WordPress. Maybe you can add people who produce open-source plugins and themes to that list, but you definitely can’t consider premium software part of the package – those are club goods, not public goods. And hosting doesn’t count either – that is a private good, because the host has to rent or buy physical servers.
Don’t blame me – the economists are the ones who write all these definitions!
Whatever gripes you may have about GoDaddy, it is not at all clear that they fit the economic definition of a free rider, and they are clearly not the worst free rider. If you want to take the economists’ view, the free riders are the millions of people who get value from the WordPress public good without ever giving any time or money back.
How valuable is WordPress?
I know, your mind is blown, but let’s keep going. I spent the weekend reading all the economics books on the Free-Rider Problem I could find, and the experts agree that the first step toward solving the problem is figuring out how to accurately determine the utility of WordPress for each person that uses it. Unless you can reliably figure out the utility people are receiving, there is no way to know if they are contributing enough back to the project. (That said, for people who use WordPress but contribute zero, we can safely assume that they are free riders because their utility is almost certainly above zero.)
The users: One challenge here is that many WordPress end users (e.g. my clients, or my clients’ interns who are required to update their company website) aren’t educated enough about what WordPress is to accurately determine how good it is compared to other options. They rely on me to make the right software call. But I think we can still make some reasonable guesses about how much value a person gets from having a WordPress site. You can look at competitors who sell similar club goods: an ExpressionEngine license used to cost $299 and a Squarespace Business Plan costs $1,380 if you keep your site for five years. So I think it is safe to say that the value of WordPress to an end user is at least $300. This means that anyone who gives back less than $300 is a free rider, according to the economists.
The agencies: Now let’s look at web design freelancers and agencies, like my award-winning team at Howard Development & Consulting. (Yes, I’m linking to my own stuff – it’s the only economically rational thing to do!) We would need to define the value of using WordPress over competitive software, and one way to do that is by assessing how much more quickly we can build a web site in WordPress versus Drupal, ExpressionEngine or Squarespace. I have been doing this for 20 years, and the answer is a lot faster. WordPress is just an objectively better software product than Drupal for my needs. ExpressionEngine was awesome but petered out due to company issues. Craft is really good, but we’ve never bothered to explore it much because the benefits over WordPress aren’t significant enough to justify the switching cost and the negative effects of a smaller community. I send my relatives to Squarespace so I don’t have to build them sites for free, but it’s not ideal for anything beyond a very small business.
So, moving away from WordPress would require me to use a worse piece of software and pay my employees more to accomplish the same goals. Let’s assume WordPress saves me 20% on payroll – that means it’s worth about $250,000 per year to me. (Damn you, economists!)
The hosts: The final category is The Big One – web hosts. There are five or ten hosting companies that seem bent on acquiring pretty much everything in the WordPress universe. I think this is a questionable business strategy, but they are prominent players in the ecosystem nevertheless. In his GoDaddy tweetstorm, Mullenweg calculated the value GoDaddy receives from WordPress as a function of how many WordPress hosting plans it sells – but I do not think this is the correct way to think about this. (I will take a moment to reiterate that I do not have any particular love or hate for GoDaddy, nor are they a sponsor – I only use them as an example because Mullenweg did.)
Web hosts don’t really have a huge stake in which CMS software their customers use. They are not selling WordPress to customers because they believe it is better than Drupal or ExpressionEngine. They are responding to market demand for WordPress because customers have already determined that WordPress is better before they start shopping for web hosts. In other words, I think Mullenweg’s calculations are backwards – GoDaddy sells WordPress hosting because the market demands that it sell WordPress hosting. If WordPress disappeared tomorrow, people would still want to host the same number of websites, and GoDaddy would just switch to selling Managed Drupal or Managed ExpressionEngine hosting plans instead.
How much different would a host’s profit margin be if all of its users switched from WordPress to Drupal tomorrow morning? Perhaps it would change the server load per pageview or increase support requests – but certainly it’s possible that there could be a faster CMS out there that actually provides more utility than WordPress on these metrics. I am sure a web host could anonymously feed me some numbers on this, but I suspect that the value of WordPress over a replacement CMS is actually less for hosts than it is for agencies. It really hurts quality of life at an agency if you have to build a site on a crappy CMS, but hosts are commoditized and mostly agnostic to software. Even hosts like WP Engine, who have optimized their systems for WordPress, could have been successful as Drupal Engine in an alternate universe where WordPress didn’t exist.
The premium plugin developers: The same can be said of premium plugin and theme developers (who sell club goods, not public goods). Their utility comes from WordPress being popular and thus creating an audience to whom they can sell their plugins or themes. (Our extraordinary Understrap Premium Child Themes are a great example!) To some degree it may be easier to develop add-ons for WordPress than for other systems, so there’s utility there too. But if WordPress didn’t exist, we could have built Understrap Premium Themes for Drupal instead.
Since WordPress already exists and is widely used, we will also eventually need to consider switching costs, which will play into my funding recommendations later in this series. It would be costly for everyone if WordPress were no longer maintained – either you’d have old and potentially insecure software, or you’d have to spend time and money to switch your site to another CMS.
So, let’s look at the utility of WordPress across this spectrum of participants in the open-source economy. There are hundreds of hosts and premium software developers who receive some utility from it – but for the most part these companies would still be doing the same thing in a world where WordPress didn’t exist, just with different software. There are probably tens of thousands of agencies that use WordPress, and they get significantly more utility than hosts because WordPress makes their day-to-day lives easier and reduces payroll costs. Some hosts, plugin developers and agencies give back to the WordPress project in various ways. But clearly, the largest cohort is the millions of end users who pay nothing, contribute nothing, and get at least $300 in utility from WordPress for each site that exists.
If we believe that there are 450 million WordPress sites on the internet and conservatively cut that in half to 225 million, that’s $67,000,000,000 (sixty-seven billion dollars) in utility that free riders received and never paid for. Even if you just look at the 2 million new WordPress downloads per year, that’s still $600 million in utility per year for the free riders. To put that into the context of the Five for the Future initiative, I estimate that Automattic has 100 employees working on WordPress Core and spends about $10 million on payroll – at that rate, it would take them 670 years to spend $67 billion.
Introducing a magical thing called ‘money’
Let’s say I turn all of Matt Mullenweg’s deleted tweets into NFTs and suddenly find myself with $1 million burning a hole in my pocket. I love WordPress as a public good and want to make sure that its future production and maintenance are fully funded. How do I give that money to the WordPress open-source project?
The short answer is that I can’t – no system exists for people to donate money to WordPress open-source development. The closest thing you can do is sponsor a WordCamp, but that just funds the event itself rather than sending money back into the WordPress open-source project. To use my money to improve WordPress, I’d have to pay individual people to work on the project – which is effectively what Mullenweg proposes in his vision for Five for the Future.
This is horrendously economically inefficient, and I suspect this inefficiency is why Five for the Future is not really working. There is a mismatch between what I have and what WordPress wants, so there is no way for us to make a transaction.
This is the classic economic parable about the creation of money – you grow wheat, I raise chickens. I want your wheat, but you’re not interested in my chickens or eggs. Without money, I would have to make a series of complex transactions and hope I eventually end up with something I can trade for your wheat. But in a system with money, I just sell my eggs to the highest bidder and use that fungible cash to buy your wheat. Dinner is served!
But I can’t do this for WordPress, because the Five for the Future ethos cuts money out of the equation. Instead of being able to accept money from people who are receiving utility from the WordPress public good, we have a complex system where I have to actually go hire a developer and then direct that person to get involved in the WordPress open-source project, at which point their skills may or may not efficiently match the project’s highest-priority needs. In practice, most money enters the system via hosting sales or when users purchase club goods, such as premium plugins, and the creators of those goods then pay people to work on the open-source project – but that is extremely complex for no good reason, except for perhaps the desire to establish a moneyless Star Trek utopia. This has been a known issue since a few hours after Mullenweg announced 5ftF in 2014. It seems a bit silly that we are fascinated with obscure concepts like the Free-Rider Problem, but we don’t have a way for the WordPress project to directly receive money.
There are some legitimate reasons that The WordPress Foundation, which is a non-profit that owns the WordPress trademark, doesn’t want to be the clearinghouse for funding the open-source project (the IRS seems to frown on this), but there’s no reason another entity could not take on that role. In fact, as you will see in my next post, doing this would solve a lot of the other governance and inclusion problems that WordPress currently struggles with.
In my next post, I’ll propose a new system that allows WordPress to receive money – and clearly defines how to use that money to allow for more contributors to work on the open-source project. Stay tuned!
This is the second post in my series, Funding the Future of WordPress.