Comment by sikozu
18 hours ago
I didn't expect this.
While I do believe that selling to DigitalOcean was a mistake ultimately, I'm concerned about the long time viability of this move considering the current landscape of web development.
Don't get me wrong, I think CSS-Tricks is a great resource and it has helped me over the years, but if I was in Geoff's shoes right now I'd be pretty scared.
Look at TailwindCSS for example, they had to reduce the size of their team and ended up joining Shopify to continue surviving. It is hard out here these days.
I don't want to sound too doom and gloom, so I wish Geoff the best of luck and I hope he succeeds.
> I'd be pretty scared.
Scared of what, exactly? It looks to me like the site is primarily static content, so operational costs should be reasonable. He could scale down to just static content, and even if traffic decrease, it is not an existential crisis. Or am I missing something?
I can only assume DO took a bath on this. Buying CSS Tricks from Coyier to largely sit on it silently before rolling it back off the books seems like a terrible investment strategy.
The cost of purchasing the property from digital ocean and the cost of his team.
Have we seen a price tag?
1 reply →
How does one pull a 6 figure salary from a blog site in 2026?
How does one pull a six figure salary from anywhere in 2026!? Might as well make a go of it
Attract an audience of people who make purchasing decisions, then attract sponsors who want to be put in front of those people. Quality of readers is more important than quantity.
7 replies →
Run a relatively small political Substack and just bide your time till you get picked to run CBS News.
2 replies →
Scared of wasting one's limited time on an ambition that can't pan out.
I published a magazine for a few years. Off the top of my head:
- publishing is fucking hard. Early on, you need a constant stream of dependable writers who do not value their time. Later on, you have a constant stream of writers but have to figure out who will make you hate your life the least.
- good writers are flaky and they always have options. A little less than half of good pitches get published. 100% of good pitches take work.
- there is a difference between good and dependable writers. Early on, you need dependable even though each one eats a lot of time. Layer on, you need good. You won’t know when to switch until your readers tell you that you suck. Then it’s too late to save your publication.
- I published before generative AI was possible but at each stage of popularity, I dealt with more PR shit bags. Add in generative AI and I’m not sure my naïveté would have survived as long as it did.
- at its best, publishing is a very expensive way to harm yourself. Back in the day, Jim Rose would have paid me to fuck myself up in equivalent ways. Naïveté convinces you that you’re changing the world through self flagellation, but when that’s gone you’re just beating the shit out of yourself.
- my co-founder handled editorial and I handled publishing. The gap between them is so massive that we really needed double the headcount. Alas, publishing is totally fucked so our budget was so low that we needed slave labour yet couldn’t even afford the slave in the first place. Without the editorial/publication divide, I would have had a heart attack five years before I did.
- there’s this thing in publishing called the trough of zero ideas. You won’t believe it’s possible in the beginning but it will hit you hard. It’s not so much there aren’t ideas; it’s that you’re tired and ideas stop feeling good… they feel like work because they are.
- in publishing it’s really easy to come up with business models. Unfortunately, every single business model requires consistently exceptional content, a reality distortion field and a lot of hand holding.
- all those membership plans etc will be invaded by sales or PR fucks as soon as you launch them. $25 a month makes a lot of people feel entitled to explain your dreams to you.
Edit - This sounds very negative and it is… but I want to publish again.
Agree the market is weird right now, but on that specific example - TailwindCSS was built around the notion of reducing keystrokes per style (basically). That's not as useful in the age of AI, of course. CSS-Tricks is about learning what modern CSS can do, which is arguably still useful. An uphill battle, for sure, but I think there's still a place for hand written human content on the web, so I plan to support it even if it's just with my clicks.
> TailwindCSS was built around the notion of reducing keystrokes per style (basically).
That "basically" is doing a lot of heavy lifting.
I don't even use Tailwind and I know the point is to localize basic style information and reduce the amount of weird CSS you have to write via functions and utilities.
Especially because Tailwind came well before CSS started natively having these features (let alone them becoming baseline widely available).
Both of those goals definitely help the current era of agents by making it easier to find the rules that affect an element without needing to reliably search all of the project CSS, and reducing required context use and the need to keep track of selector lists.
> That "basically" is doing a lot of heavy lifting.
Yea. To me Tailwind was to make CSS act predictably for someone who was not an expert and just wanted to make some tweaks in a good expertly-crafted CSS theme design without seeing the strangest unexpected things happening layout-wise.
By localizing basic style information and reducing the amount of weird CSS you have to write, you have thus reduced your keystrokes per style. We're talking about the same thing.
2 replies →
The tailwind ones slightly misleading. They laid off their staff, posted on twitter about it, opened a sponsorship program, made approximately $2 million, and THEN joined shopify.
I agree with this. In my day to day I've started AI as a glorified "look this up for me" thing and hardly look at reference sites anymore. It's a shame because I really did admire CSS-Tricks when I was starting out and referenced their stuff often. I'll be rooting for them though!
>While I do believe that selling to DigitalOcean was a mistake ultimately
using the word "ultimately" makes this "Monday night quarterbacking"
It's not a mistake to plan a picnic for a day that turns out to be rainy, even if you got paid to plan the picnic.
It would be a mistake to sell if the seller should have known at the time of the sale that the seller's goals would not be met, but something turning out badly doesn't retroactively make it having been a mistake, it just means it came out badly.
This was my initial thought but I think that if he will focus on new problems arising. For example how to build websites with AI to not look like a slop, how to approach this problem, why is it happening and what to look at. It certainly is brave decision but it could also be an opportunity.
>"I'm concerned about the long time viability of this move considering the current landscape of web development."
I owe Coyier and CSS Tricks a huge debt to getting my career started in the 2000s. It's loss will be tragic. But web development (as we knew it) has gone the way of elevator operators. Dead and gone. Good on the guys for getting their bag while it still had some value.
Llms are not good at web design or writing css, this art form will live on for a long time to come. It is art, I dont care what people say, and as we all know machines mostly suck at creating art. Anthropic claims to have started training in "taste", but im sure they've realized that cannot be distilled into vector embeddings. Taste is somewhat empheral and is dictated by society and culture that machines can't experience.
In fact, I think it will have more staying power than most, if not all of web dev domains of expertise.
Theres simply no "soul" behind llm crafted websites. There are companies out there, like lovable et al who sell the idea that web design is solved but are still hiring six figure retainer design firms to do their frontend.
I believe the best web platforms will heavily invest in human crafted UI/UX because that is what will make you stand out. The rest will get thrown on the slop pile and forgotten.
I see your point and I agree that ultimately web design will be incredibly hard to automate, probably harder than almost any other programming domain, where people will likely stay for accountability reasons.
But the thing is that for the vast majority of web pages/applications, it barely matters what does the design look like, especially for B2B/internal apps. All the dashboards, CRMs, tracking tools, etc -- these exist mostly for functional requirements, so the design just needs to be inoffensive.
B2C market might hold longer, but ultimately there might emerge some sort of higher-level design specification language which would sidestep the original problem or being too recognizable.
This is like saying the design of your mobile operating system barely matters as all you use it for is to open Youtube and a browser. Design is everything when building a product people interact with.
The thing is, there's no end to the appetite to lower quality in order to reduce personal cost. Whether there's a website or not is very "seen". Whether it's any good at evoking intended understandings or organizing information is not (directly) seen. Unfortunately, only a very small portion of humanity directly cares about design.
I shared your sentiment maybe a year+ ago when it comes to css, and I do think it's one of the areas AI in general struggles with (where art and code meet), but it's a lot better today with the latest OpenAI and Anthropic models as someone that works on design systems for a living (for now). It follows stylistic patterns, it can inspect the issue in the browser and have a better understanding of CSS upstream effects than it used to (it used to suck at this and paint itself into a corner over and over like some game of whack-a-mole), and can even tackle things using latest interop features.
It does make really silly mistakes now and then, like using fallback css variables in the wrong order, but they are fewer and farther in between.
"I believe the best web platforms will heavily invest in human crafted UI/UX because that is what will make you stand out. The rest will get thrown on the slop pile and forgotten."
Agreed!
AI model taste is not good, but taste is easy and many cheap human have soul, only hard part is memorize CSS trivia, I predict cost of UI/UX go down and CSS education business also go down
>Llms are not good at web design or writing css, this art form will live on for a long time to come.
I don't want to ruin your day, but LLMs excel at CSS, thanks to the abundance of explainer articles like those on css–tricks.com. Had a bug fixed in about 10s the other week after having stared at it myself for like 10min.
[flagged]
You can just make a mockup in whatever tool you want, and pour your 'art' and 'soul' into that.
Then hand it to the LLM and it'll make CSS to match.
Spoken like someone who's never actually put their art and soul into a mockup and asked an LLM to make CSS to match.
The guys made some money, it's not bad idea, specially knowing nobody cares now tutorials on internet with the era of agentic development. See StackOverflow, it's dead. They did the right thing when selling this.
CSS-Tricks was sold to Digital Ocean a while back. Now it's been purchased by an individual again. Someone who used to work on it back in the day
I bet they bought it for a tenth of the price they've sold it.
2 replies →