A short history of who builds arts websites
I made a fancy widget, and this post is my excuse to put it somewhere. You’ll have to scroll down to see it.
If you can delay gratification for just a moment, though, I’ll tell you a quick story.
When agencies ruled the land
Once upon a time, if you were an arts organisation in need of a website, you hired someone to build it.
Your budget dictated whether that was a friend’s cousin, a small agency, or a bigger outfit packed with talented specialists. Some agencies started specialising in the particular nuances (some imagined, some real) of arts and culture organisations.
Agencies wanting to break into this glamorous, riches-filled sector often cut their teeth on smaller work (maybe an independent or touring theatre, or a microsite or exhibition interactive) before anyone needing a serious ticketing integration would take a chance on them.
Then the DIY website builders arrived
Platforms like Squarespace and Wix came along and took a bite out of the smallest end of the market. For a lot of people, decent templates with a bit of customisation were good enough, and for a small annual fee, the price was right.
The agencies didn’t mind too much, because that end of the market was never particularly profitable anyway. But it was where newcomers got their start, so the way into the sector got a little narrower.
Website-as-a-service came on the scene
Next came platforms built for the sector, such as Basker, carbonhouse, CultureSuite, Loop, Saffire and Simpleview.
Some have been around for a while. Others are newer upstarts built by agencies with sector experience who saw the value in a different model. They’d probably been casting envious eyes at ticketing platforms and collections management systems, with their single codebase and subscription income.
The sales pitch is that, for an annual subscription, you get far more functionality for your budget than you might otherwise. You might lose a little flexibility and portability, but that’s the trade-off.
And now there’s agentic coding (or vibe-coding)
AI tools like Claude, Codex, Lovable and Replit give a reasonably handy person the means to knock together something simple (or not so simple). They also supercharge what in-house teams are capable of producing.
That probably means more competition at the lower end. Or maybe the speed of production makes smaller jobs worth outsourcing again, and the way in widens back up a bit. It could also leave a mass of poorly coded, unsupported websites and apps knocking around the place. Who knows? It’s early days.
It's widget time!
Here's how the market has developed over time. Click through the stages to watch it change. Wow.
The costs along the top are a guide to what people might normally pay for something bespoke.
And I know, I know. The choice of approach is going to come down to all sorts of things like complexity and so on. I've also left out in-house teams.
So what?
It’s just an observation, and I'm painting with a very broad brush here. But look…
The middle is being squeezed, just like it always ends up being. There’s more competition at the lower end, and more again in the £20-100k range, where website-as-a-service and agencies now overlap. If you’re an arts organisation in the market for a website with that sort of budget, lucky you.
I’d imagine established agencies will chase work upmarket, looking for clients with bigger budgets. Although not so big that they do everything in-house.
Come to think of it, agentic coding might mean that those in-house teams can become even more productive - outsourcing less than they might. And maybe more organisations decide to hire more coders and hybrid builder/PMs, unlocking a golden digital age of the arts. Maybe.
Anyway, that was in my head, and now it's not. Thanks for scrolling.