slorber
commented
1 year ago 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
remotesynth
category: note slug: netlify-era-jamstack-end
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:
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?".
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.
(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!)