Month after month after month, 2018 was full of examples of the risks of platform overreliance. The problems became harder to ignore, both for organizations and for end users.

We started 2018 with Facebook cutting reach for Pages yet again — another sign of the love/fury relationship news outlets have always had with that social network. We ended 2018 seeing Tumblr make a dramatic, hail-Mary move to remove significant amounts of user-generated adult content from its platform. Both moves highlighted the worst of what happens when platforms define the ways we communicate.

Lots of smaller trends touched upon these same weaknesses. For example, controversial political personalities such as Alex Jones and Laura Loomer found themselves “deplatformed,” each making public scenes in last-ditch attempts to salvage their public voice. (We can agree or disagree about whether they should have been kicked off, but the fact we’re even having the conversation is telling.) Meanwhile, small media companies with outsized influence — like The Awl, Lenny Letter, and Rookie — said their goodbyes, their business models running into the challenges of scale that come with running a modern media outlet.

The undercurrents behind many of these shifts during 2018 were clear: Platforms are the key to influence in the modern era. We’ve spent years being burned by them and complaining about them for either doing too much or not enough.

But what if, in 2019, we take a step back and decide not to let the platform decide how to run the show?

We already have the seeds to get things started. In 2018, email marketing became an increasingly important part of the media diet — and even though platforms abound there, you don’t need them to create something great, because email isn’t platform-dependent. The one social network that seems to have gained any real momentum since Snapchat is Mastodon, in large part because it has explicitly sold itself as an anti-platform of sorts that will never see an IPO. Tech-savvy users who have seen it all before want to see something else. There are more of those users than ever — and they’ll still want things to read after they deactivate their Facebook account.

And there seems to be an increased interest in the strategic advantages of a good content management system that is hosted on your own server and based on your own nuanced needs. No longer are we talking about the CMS in terms of whether it’s dead, or whether we should let Medium eat all our content and share it out based on some algorithm. (We already tried that. It didn’t work.) We can own our technology — even sell it to other media outlets without the specialization — and define more of our own destinies. I’ve been working on a redesign of my site recently, using a more robust CMS, and the advantages of controlling the structure of the platform soup-to-nuts are obvious, even if it requires more upfront work.

These platforms promised reach, but they came with a lot of other things we didn’t actually want, and the scale is tipping in favor of the do-not-want category. 2019 is the year when publishers — whether big ones like Axios or the Los Angeles Times or tiny ones like mine or Judd Legum’s Popular Information — move away from letting someone else call all the shots. Or, at least, they should.

We’ll never be rid of social networks and other digital gatekeepers, but in 2019, perhaps we should right-size their influence on our media businesses. Let’s replace the shadows that Twitter and Facebook and Google have been on the media with some business-model fundamentals. As 2018 has shown, they’ve offered us a lot more heartache than it feels like they’re actually worth.