How to Blow Up a Category - Netlify's New Era and The JAMstack Endgame

[swyxio]

---

category: note
slug: netlify-era-jamstack-end

note - this is a hasty written braindump of feelings as emotions as I don't have the time to polish this essay up to my usual standards, but still wanted to capture this important moment and end in my life. pardon any poorly phrased and organized thoughts. The JSJam guys did a roundup of reactions here and Ricmac did one on The New Stack.

A lot of ink is spilled about category creation in devtools, and a lot of people try, but rarely does a company successfully create a category, and then voluntarily bury it/let other companies draft and overtake it. When I think about the recent history of Netlify and the JAMstack, it reminds me of a smaller form of Eugene Wei's excellent writeup on Elon Musk's ongoing deconstruction of Twitter: How to Blow Up a Timeline.

So that's how I'll title this post, but trust me that I had no idea what to title it - my feelings are complex and contradictory, i don't have enough information to make any sort of authoritative judgment one way or the other, and I just wish my friends well as far as possible.

The Story So Far

It's been 3.5 years since I left Netlify shortly after the Series C. Since then it's been a rollercoaster:

• CTO left in early 2021, no replacement
• a $105m Series D in 2021
• acquiring Onegraph in 2021 (shut down in 2022)
• acquiring Quirrel in Feb 2022
• launching a $10m innovation fund in mid 2022
• ending the year with a 16% layoff (48 people)
• acquiring Gatsby in Feb 2023 (Gatsby raised $47m in its life)
• offloading Netlify CMS to its community in Feb 2023
• acquiring Stackbit in June 2023 (Stackbit raised $14m in its life)
• laying off more - rumored to be 25% (~48 people including the COO, with a rumored 50 more natural attrition in the intervening period) in Jul 2023 (UPDATE: an insider comments that it was less than 25%, and other numbers in the HN comment I linked to were also significantly inaccurate. The current size is around 250.
• killing the Jamstack Community Discord in Jul 2023

JAMstack is not Netlify, Netlify is not JAMstack. But the two are tied at the hip, and recent discussions from former JAMstack community members Brian Rinaldi and Jared White gave me some occasion to examine my own complicated feelings about the whole thing. I guess I lay out the timeline as a form of demonstrating how closely I feel every bump and divot in the road despite no longer working at the company; Netlify is not only my second largest shareholding, but also where I got my professional start in devrel and devtools because many friends and colleagues welcomed me into their industry for the first time, for which I am forever grateful.

Here was Netlify's website the day I left:

And here it is today:

You now have to scroll alllllll the way down to the footer to find the word "JAMstack".

Netlify's new focus on the Enterprise

This is big vibe shift as complete and total as the logo change 4 months ago. Gone are the days where competitors would launch using the JAMstack label, Japanese fans would produce fan manga about JAMstack, and Netlify would hire industry heavyweights like Sarah Drasner and Laurie Voss and Cassidy Williams and Jason Lengstorf and Zach Leatherman on an almost monthly basis; that momentum shifted to Vercel in 2021.

The old game used to be "bottoms up"*. In 2020 we celebrated reaching 1m registered users. In 2023, when I talk to people about how Netlify is doing, they sigh and say "I hear they're doing great in enterprise?".

Aside: *in theory, it never was bottom up only; I remember sitting down with the then-CCO in my first week on the job and him working out a conjoined triangles of success-style story of how we could have both the bottom-up cake and the top-down go to market dessert with no tradeoffs. Many such ZIRP cases!

Interestingly, this doesn't necessarily mean Netlify has slowed down in the bottoms up adoption - Mathias actually disclosed that the registered user number is now up to 4 million, which is a quadrupling over 3 years. It just isn't the story anymore - enterprise is the Story, and some dude at Gartner decided that The Future of Business is Composable, hence now Netlify and Hightouch and Monte Carlo and Sanity.io and anyone else seriously pursuing The Enterprise happily lather themselves all over with Composability despite it being overused to the point of meaninglessness for the developer.

The End of the Road for JAMstack

In the same way, JAMstack isn't the story anymore. The party line from Netlifolk in response to the JAMstack decline commentary has been taking it as a kind of victory lap:

That's certainly one way to view it, that JAMstack "won", and now we move on to make The Enterprise Composable. At the end of the day, money is money, and if this strategy is working for them then honestly your or my opinions straight up don't matter, except to the extent that you pine for an idea of a company that no longer exists (perhaps because it was unsustainable in its prior form in the first place).

I'll just say what I feel - as a former JAMstack advocate (check this blog's history), this isn't what victory feels like.

React's new push into Server Components and tighter integration with Next.js has made things MORE tightly coupled, not less.

The rebrand from JAMstack to Jamstack felt like a distinction without a difference (which is why I continue to capitalize the old way). When the frontend metaframework metagame moved from 100% static rendering to mostly-static to mostly-serverless-and-sometimes-edge, JAMstack's advocates pivoted adroitly, claiming this fit JAMstack all along, which means it fit nearly everything, which means it stood for nearly nothing. When even a VP is saying "Jamstack is a feeling", what he doesn't say is that feeling is most often ambivalence.

But above all, what has been shipped in the past 3 years may feel appealing to enterprises, but insofar as I have any read at all on developer mindset, it has been unappealing to developers. What has been launched wasn't followed through on, what is promised hasn't driven excitement, and JAMstack started being associated more with the trailing edge rather than the bleeding. (it is now doing both).

Most tech doesn't die, it just starts denying that it is dead. JAMstack has reached that point, and it's finally okay to say it. Netlify must forge a new path ahead to take back industry leadership, because selling devtools to the C suite without recapturing the hearts of developers belongs more in the 1980s than the 2020s.

---

Update - response from an insider: Going from solving the problems of an individual web team, helping them ship faster, leave operations behind, build better experiences, is one thing. Solving a company sized problem of a larger enterprise is really different and has taken a lot of investment on our end, and I realize that's not personally features that would be on your wishlist, but they're really important to our customers (and first of all the developers at those customers).

(good. I'm many years away from the story at this point, and recognize that I don't have anywhere close to all the relevant information. I wish my friends at Netlify the most luck!)

[slorber]

I have similar feelings.

Somehow Netlify remains a great platform that I still like to use, but the business direction and recent marketing hasn't been exciting to me. It looks like a pivot, maybe because it's too hard to compete with Vercel on the exact same field without a popular framework like Next.js?

IMHO the Jamstack term is kind of dead but the concept of a fully static site is not. Many people still like to deploy on GitHub Pages, despite its inferiority in terms of server-side features and redirects.

My hope is the adoption of Rust tooling will make building large fully static sites much faster, leading to a SSG revival. SSG doesn't mean trailing edge, and React static site generators can still leverage useful new React features like Server Components at build time. We need modern SSGs that do not compromise the build speed.

[swyxio]

"modern SSGs that do not compromise the build speed" is a misnomer imo. that battle is lost - vercel was right to bet harder on serverless than netlify did (with a fullly vertically integrated framework) and I myself have switched permanently to that (for swyxkit) after years of trying every framework i could in pure static SSG land.

[remotesynth]

I really enjoyed this post and it adds a different perspective to this argument. I personally don't think this is about Vercel and I suspect Vercel's "pivot" is not far behind. Did they win on the framework? Yes. But Netlify is a case in point that popularity (which is what your chosen graph measures) and profitability do not always go hand in hand. I'll be honest that I don't think they have the answers on how to turn that popularity into a profitable enough business for the type of exit VCs expect, even if the current level popularity of Next is indefinitely sustainable (note: it isn't).

[swyxio]

@zachleat tries to pick up the jamstack slack: https://www.zachleat.com/web/jamstack-future/

this is fine but there will probably never be a big company built around jamstack again in the next 10 years