Every time someone asks me which tool to use for their website, what they're really asking is something else: how much is this decision going to cost me in the end. Because the cost of building it shows up in the quote, and the cost of living with it doesn't.
I work with all three. These are the questions I settle before recommending one.
1. Who is going to edit the site, and how often?
This carries the most weight and gets asked the least.
If someone non-technical will be editing it every week — posting offers, changing prices, uploading photos — WordPress or Framer win, and not narrowly. Building a custom content system for that is expensive and adds nothing.
If the site gets edited twice a year, paying monthly for a content system is money thrown away. Code wins there.
2. How much does speed matter?
On a corporate site visited by people who already know you, performance matters little: they'll wait.
On a site taking search or campaign traffic it matters a lot, because every second of load time costs you visits you already paid for. Custom code has an edge platforms can't quite match: you control exactly what loads and when.
That said, a badly built Next.js site can be slower than a well-built Framer one.
The tool sets the ceiling; the work decides where you land.
3. Does the site have to talk to other things?
A form, a newsletter and analytics are fine on any of the three.
A catalogue that syncs with your ERP, a private area per client, a quote calculator, content that changes by language and country — on a closed platform those get solved with patches, plugins or external workarounds, and every patch is a piece that can break on the next update. In code, it's simply what gets built.
4. How long is this site going to live?
A three-month campaign and a site that will be your face for five years don't deserve the same decision. For the first, fast wins every time. For the second, count the accumulated cost: licences, paid plugins, the content system, the redesign due when the template starts to look its age.
What I recommend, in practice
WordPress when there's a lot of content managed by non-technical people, when they already use it happily, or when they need a specific ecosystem (a simple shop, memberships, an industry plugin). Its weak point isn't the tool, it's maintenance: an unattended WordPress is a security problem waiting its turn, and that cost has to be budgeted from the start.
Framer when the project is mostly design, there's a deadline and it'll change often. It's very good at that. The price is the lock-in: what you build there stays there.
Code, with Next.js, when the site has to do more than display information, when performance or SEO are part of the business, when there are genuinely several languages, or when you want the site to be an asset rather than a subscription.
It costs most up front and least at two years.
What I never let the tool decide
That the domain and hosting are in the client's name. That documentation exists. That it can be migrated. That's independent of the technology, and it's what really decides whether the site is yours.
If you're not sure which one is yours, tell me your case in four lines: who will edit it, how often, what it has to talk to, and how long you want it to last. That's enough for a recommendation — even if the recommendation is that you don't need me.