Skip to main content

Running your own site is painful. Hosting Nazis is worse

A stock photograph of someone writing on a laptop

I’ve spent much of my career telling organizations that they should publish in a space that they control, on their own domain name.

My usual argument is that it shields you from major changes in company policy, or even from the platform you’re depending on from going out of business. If you’ve built a media business on a platform knowing that it’s been designed as a place for text-heavy media businesses, but one day it decides it’s going to be more profitable if it becomes a place for 3D immersive video, and your whole audience is locked into that platform, your media business is in trouble.

Publishing on your own site and your own domain name avoids that issue: you have full control over your underlying platform, and because you can change where your domain name points to, you can change your platform without losing any of your audience. In terms of Michael Porter’s Five Forces, this tactic reduces the power of suppliers — the platform that underpins your site — to dictate the terms of your business.

The trouble is, running your own site isn’t easy. If you’re a developer, you probably have the skills and knowledge to select an underlying platform, install it on a server, secure your infrastructure, customize it, and optimize it for discovery and sharing. If you’re a writer, you might not — and even if you are a developer, if you’re trying to start a site from scratch, all that technical administration is time taken away from working on all the stuff that makes your site special. Having a well-running platform for your site is table stakes; the core value of any site is the content itself. As any entrepreneur knows, you should spend as much time as possible on your core value: focusing on the thing that makes you unique and special will give you a better chance of success.

All of this doesn’t mean that the characteristics of a platform aren’t important. They’re very important, and can make or break a media startup, which is why Substack seemed like such a good choice for a great many people. It had everything: integrated payments, a solid recommendation engine that accelerated subscriber growth, support for using your own domain name, customized branding, optimizations for search engines and social sharing, and it just felt really good to use.

On a technical level, Substack was clearly a very good choice for independent writers trying to make a living on their own. It also stood in contrast to Medium, which had similar goals but was firmly optimized around helping people earn a living from individual articles even if they didn’t have a built-in audience (I’ll say more about Medium in a minute). Both were free to get started on, relieving writers from the technical or financial burden of setting up their own platform, but each had a different focus. On Medium, great pieces stood alone, so you could gain an audience for a thought even if you didn’t have a following; on Substack, you could build a following for your ongoing work.

Which is why it was incredibly disappointing when it became clear that the platform actively embraced and funded bigots. First came the transphobes, which a few people made a fuss about, but not enough. (If we rewrote Martin Niemöller’s famous poem for today, trans people would be the first line.) Then, most recently, it became clear that Substack was rife with actual flag-waving Nazis.

As Casey Newton pointed out in Platformer:

To be clear, there are a lot more than six bad publications on Substack: our analysis found dozens of far-right publications advocating for the great replacement theory and other violent ideologies. […] The company’s edgelord branding ensures that the fringes will continue to arrive and set up shop, and its infrastructure creates the possibility that those publications will grow quickly. That’s what matters.

That’s what made it untenable for me and a great many other publishers. Platformer, Garbage Day, Citation Needed, and ParentData are some of the titles that have moved away (or announced that they’re moving away) over the last week, and these are the high-profile tip of a much larger iceberg.

So where should writers go?

Unfortunately, I think a platform that’s completely right as an alternative to Substack doesn’t exist, which I’ll talk about in a moment. Where writers have been going fall into a few buckets:

  • Buttondown, an independent newsletter platform
  • Ghost, a blogging platform with built-in support for newsletters and paid subscriptions (sometimes through Ryan Singel’s excellent Outpost service)
  • WordPress, the blogging platform that powers roughly a third of the web (but has poor support for newsletters and paid subscriptions)
  • Biting the bullet and developing or self-hosting their own thing

My approach, because I’ve always had my own site running on Known, has been to move my newsletter to Buttondown. It’s worked really well for me, but these are all good approaches.

I’ve been a bit surprised to not hear about people moving to Medium, which has been undergoing a quiet transformation under Tony Stubblebine’s leadership over the last year. It’s certainly worth considering: revenue is up, the platform has reorganized itself around publishers and followings, it supports custom domains, writers can take their content and subscribers with them if they choose to leave, and the strategic thrash of the mid-2010s is gone. Tony and his team are genuinely hunkering down and doing the work to support writers.

A big difference with Medium is that the audience’s financial relationship is with the platform rather than a writer. On Substack, you subscribed to outlets like Platformer and Citation Needed individually; on Medium, you’re paying one of two tiers to access the full network. For readers, that’s clearly far better: you’re getting a world of writing for the same flat price. For writers, getting frictionless access to Medium’s aggregate paying userbase may help grow followers; it is clearly true, though, that you can’t bring your paid customer relationships with you if you choose to leave. On Substack, those relationships are made directly with you and depend on you having your own account with Stripe, which means you can leave and keep servicing your subscriptions without asking anyone to re-enter their payment details.

In order to really support independent media startups, particularly individual writers looking to make money from their work on their own, there are three categories of service that I wish existed:

A fully-managed direct relationship platform for writers

Any writer should be able to sign up to a service, configure their platform, and begin selling their writing directly to an audience without worrying about their writing showing up next to, or appearing to endorse, hate speech. It should be beautiful, easy to use, and friction-free.

They should be able to own their relationship with their audience to the point where if they choose to change platforms, they can take their audience with them and seamlessly start writing somewhere else. They should never have to deal with technical configuration: everything should just work. And each writer should be able to gain from network effects as the platform grows, allowing them to gain a following and build revenue more quickly.

That’s almost Medium, aside from the direct relationships. That’s almost Ghost, aside from the network effects. So close!

Indie recommendations

A lot of the focus of the indieweb and of the fediverse has been to provide an independent alternative to social media. I have no criticism of that approach! We need that! But there’s also something I’d like to see that goes beyond it.

I think we’ve assumed that social media is how we learn about new websites and articles. That’s a user experience pattern that has been inherited directly from Twitter and Facebook, which always wanted to be the way people discovered news and information. When I built the first version of Known, I had the idea of owning your own versions of social media platforms in mind.

Long before social networks, personal websites had links to other sites the authors enjoyed. Sometimes it was just called a links page. Blogs called them the blogroll. Substack’s version of this was for an author to recommend other newsletters, so when someone subscribed, they would also be asked if they’d want to subscribe to these other author-endorsed outlets. It was a superb way for writers to rapidly build a following outside of their own established networks.

The following requires some underlying protocol work, but here’s how it would work from the user side:

  • As a user, I want to subscribe to an author.
  • I visit their site and click to subscribe, entering my details.
  • The site shows me a selection of other blogs or newsletters the author recommends.
  • I agree to subscribe to those. (Not as a paying subscriber, but as someone who will receive new content as it is published.)
  • I am instantly fully subscribed, without having to enter any further information on those third party sites.
  • The authors of those sites know that they gained new subscribers that were referred from the recommending site.

The net result: every author can have the freedom and ownership of publishing on their own site, but with the network effects of a closed network.

Of course, even without this infrastructure, any site can already create a links page or a blogroll. I’m actively working on that.

Self-hosting that works like an iPhone

You should be able to pay for server hosting and have access to a fully-managed App Store that, just like an iPhone, lets you install new services with one click and keeps your software up-to-date. Some of those services will be free; some will be paid-for. It should be no more complicated than that, with zero exposure to the underlying server processes and scripts.

Shared hosting still hasn’t really evolved since the nineties: it’s a world of (S)FTP access, dubious control panels that don’t do much to help the user, and appalling user experiences. Virtual hosting is newer and more powerful, but you need to be a very sophisticated user to deal with containers, package managers, and so on. The former are designed for hobbyists; the latter are designed for software engineering teams. A self-hosting environment that’s optimized for non-technical individuals to own their own websites and data does not exist.

In summary

We do need a way to support great writing. It’s how we learn about the world, quite often in a way that helps inform our democratic decisions and the way we see the society around us. A world where everyone is writing for free and independent journalism has no means of financial support is not tenable or desirable, in my opinion.

I also believe that not providing financial support to literal Nazis is non-negotiable. I can’t believe that’s an argument we even have to make. This isn’t ambiguous: Nazis are bad.

The indie web should be a place where independent writers can thrive. I believe it will be. We just need a little bit more infrastructure: network effects, easy payments, removing the need for non-technical people to get involved in technical administration. The Ghost ecosystem in particular has shown us that there’s a great opportunity for this to be done well for writers.

Unlike many open source / indie web folks, I also don’t draw a hard line about hosted services like Medium, given the right features. The important thing for me is that writers can write and be heard. Anything that makes that easier — while not, again, literally funding Nazis — is fantastic in my book. The writing is what matters. Letting people connect and learn from each other — reaching people with ideas — is what the web is all about. The trick is giving them the tools and freedom to do that sustainably, on their own terms.

· Posts