Every all-in-one platform is solving 2014's problem. Here's what 2026 actually needs.

The 2014 platforms got something right that nobody talks about anymore. They figured out that a small business needed a website, and building one used to take six weeks of working with a developer who didn't reply on weekends. So they built drag-and-drop. They built templates. They built hosting. They built domains and SSL and email and shopping carts and forms and analytics, all in one place, with one bill.

It was the right answer to the actual problem. The build was hard. The platform made the build less hard.

That was 2014. The problem changed. The platforms didn't.

The mental model the 2014 generation was built around

Pick any all-in-one. Squarespace, Wix, Kajabi, Webflow without its developer mode. They share a shape.

You sign up. You pick a template. You name your site. You put your business identity (logo, colors, fonts, tagline) in. You then build pages inside that site. About, services, contact, blog, maybe a shop. Every page lives under one domain, one brand, one design system. The site is the unit. The page is a sub-unit.

The pricing tiers reflect the model. Squarespace gates by features (commerce, member areas, advanced analytics). Kajabi gates by capacity (number of admins, number of products, number of customers). Wix gates by polish (ad-free, premium templates, App Market). All three gate by site. You pay per site, not per page. Because the site is what you bought.

That mental model presumes one outcome: the customer is shopping for a business. They land on your homepage. They click around. They get a feel for who you are. They convert through your funnel of pages.

It was correct for the businesses that existed when the mental model got picked. Local services. Boutique e-commerce. Photographers and consultants and yoga studios. The business was a place. The website was a digital version of the place.

What changed

Two things, in this order.

First, the coaching economy went big. By 2022, there were more solopreneurs selling expertise online than there were local services using Squarespace to advertise expertise offline. The unit of revenue changed. A consultant selling $5K engagements doesn't need a website. They need a sales page. A coach selling a $497 cohort doesn't need a website. They need a sales page. Plus a webinar opt-in. Plus a replay page. Plus a checkout. Plus a confirmation page. Plus a delivery page.

The business unit shifted from "the company" to "the offer." And an offer is shaped like a page, not a site.

The 2014 platforms responded by adding course modules, payments, member areas. They tried to bolt the offer onto the site. The bolts mostly held. The pricing went up. Kajabi Pro is $399/month, because the bolts are expensive to maintain.

Second, and more important: AI flipped the build constraint.

Building a page used to take hours. Now it takes minutes. The "I need a designer to make this look right" wall fell over. The "I need a developer to wire the form" wall fell over. The "I need a copywriter to write the headline" wall fell over. A coach with no design or code training can describe a sales page and get a working draft back in 90 seconds.

The bottleneck moved.

It moved from build to publish. From "can I make this thing exist" to "can I get this thing live, addressable, on my domain, in front of buyers." The shape of the problem changed at a layer below the platform.

What 2026 actually needs

A coach in 2026 is going to ship 30 to 60 pages a year. Not one website with 30 pages on it. Thirty separate pages, with separate domains or subdomains or paths, each pointing at a separate offer or audience or campaign.

Lead magnet for the email list. Lead magnet for the affiliate launch. Webinar registration. Webinar replay. $497 cohort registration. $497 cohort replay. $1,200 group-coaching application. $5K mastermind application. 1:1 sales page for a single high-intent lead. Confirmation pages, thank-you pages, delivery pages, upsell pages.

Each page is its own work. Each one has a different audience, a different promise, a different conversion target. The "site" they all roll up to is mostly a fiction. The actual unit of work is the page.

The all-in-one platforms aren't priced for this. They are priced as if you have one site and you maintain it. They are priced as if pages are slow to make and cheap to host. The math is upside down for 2026.

If a coach ships 40 pages this year, the all-in-one is charging them roughly $2,400 (Skool Pro), or $4,800 (Kajabi Pro), or $200 to $600 (Squarespace, depending on tier and add-ons), for the right to host all 40 inside one shared chrome. The chrome is doing nothing for them. The actual work is the page, the prompt, the publish.

This is the structural argument for the all-in-one tax. You are paying for a model of business that doesn't describe the business you are running.

The page-first stack

The page-first version of the stack looks different.

A publishing layer that takes a description (prompt) and produces a live URL. Built for volume. Priced per page or per use, not per "site." No template selection. No tier gating on "number of products." Pure publish.

A domain layer that can map subdomains or paths or full custom domains to the pages. Without a website wrapper. Each page can stand alone.

A payments layer (Stripe, Paddle, LemonSqueezy) that the page can talk to directly. No platform middleman.

A scheduling layer (Calendly, SavvyCal, Cal.com) for high-touch offers.

A community layer (Skool, Circle) for the post-sale relationship. After the page sells. Not before.

That stack is cheaper, faster, more flexible, and more accurate to the work. It works the way the work works.

The all-in-one is the opposite. It's the work bent to fit the platform.

Where the 2014 generation goes from here

Three paths.

One: rebuild around pages. Drop the site as the unit. Reprice per page or per publish. This is hard because the entire backend architecture, billing model, and product brand are built around sites. The Squarespace UI tells you to start with a template, not a prompt.

Two: get acquired or wound down. The 2014 generation has been making the same hardware-from-the-software-era category transition that Adobe Creative Suite made in the 2000s, where the unit of work (the design file) outgrew the platform (boxed software). Adobe rebuilt around the cloud and the file. The 2014 platforms have not rebuilt around the page yet.

Three: stay where they are, profitable, gradually losing relevance. This is the most likely outcome for most of them. They have huge installed bases. They have switching cost. They will charge those customers for a long time. They will also stop being the place new coaches start.

What I think happens

I think Squarespace and Wix become the new GoDaddy. Capable, profitable, irrelevant to whoever is shipping today.

I think Kajabi stays in the picture because the course is a real artifact that needs hosting. But the sales pages move off Kajabi. The $399/month becomes a course-platform bill, not a marketing-platform bill. Kajabi's growth slows.

I think Skool fights the all-in-one fight with a friendlier wrapper for another two years and then has the same conversation Kajabi is having now. The community is real. The community alone isn't enough. The page in front of the community is the missing piece.

And I think the new publishing layer becomes invisible infrastructure. Like SSL. Like CDNs. The page goes live. Nobody talks about which tool shipped it.

Where I might be wrong

I have been pitching this argument inside Leadpages for 18 months. The pushback I take seriously is that customers don't think this way. They think they are building a website. They want a website. If you tell them they need a publishing layer plus a payments layer plus a scheduling layer, you've just made the buying decision harder, not easier.

That is real. There is a customer for whom the all-in-one is the correct answer because the bundling is the value, not the architecture. The all-in-one is sold as one decision instead of five.

What I think happens is the new ICP doesn't shop for that bundle. The coach who has been buying $99-499/month all-in-ones for the last five years is not the coach who is signing up this year. The coach signing up this year is more technical, more comfortable with AI, and more willing to compose a stack instead of buy a bundle. Their first ten pages cost them less than $20 in tooling, total, and they didn't watch a single getting-started video.

If that customer doesn't exist in the volume I think it does, the page-first thesis loses. That's the risk.

The signal we look for is two things. Per-customer page count, which is what HTML Pub publishes. And the source of those pages. If they came from a prompt, the work happened above the platform. If they came from a template editor, the work happened inside the platform. We see prompt-sourced pages outnumber editor-sourced pages on our side, at growing margin. That is the pattern that has to keep holding.

Why I'm not worried about 2014 platforms shipping a publish-from-prompt feature

The other pushback I get is that Squarespace or Wix will eventually ship a publish-from-prompt feature and end the conversation.

They might. There's no patent on it. The protocol is open. AI is commodity.

The reason I don't think they win the new market even if they ship the feature is that the rest of the product is shaped wrong. The pricing is wrong. The onboarding is wrong. The dashboard is wrong. They will ship a publish-from-prompt feature and slot it under the existing site builder, behind a navigation tab, gated to higher tiers, with templates as the default starting point. It will be a feature inside a 2014 platform. The customer who wants to ship 40 pages will not pick it.

The architectural choice that matters is whether the page is the primary unit or a sub-unit of the site. That choice cascades to pricing, to the dashboard, to the URLs, to the analytics, to the support model. You can't bolt the right architecture onto a product that is shaped wrong.

This is the same reason Salesforce never won the Notion-shaped market for ops collaboration, even after they bought Quip. The architecture rejected the bolt.

Where we landed

The publish button your AI was missing is what HTML Pub is. Not a website builder. Not an all-in-one. A publishing layer that takes a description and produces a live URL. Built for volume. Priced for pages, not sites.

Carrd is an editor. HTML Pub is a description. They design. We build. The shape of the answer is different because the shape of the question changed.

If you're already inside Squarespace or Kajabi or Wix and the business is working, don't blow up the system. The work is to add a publishing layer on top of what you already have. Ship the next 20 pages from a prompt. Keep the website as the catch-all if it's converting.

If you're starting today, skip the all-in-one. Stack the pieces. Use the publishing layer. Use Stripe directly. Use Calendly for calls. Use Skool or Circle once you have a community to put there. Pay for the pieces you use, not for the bundle that assumes you're a 2014 business.

The 2014 platforms were right about the problem they solved. The problem changed.

About the author

Michael Sacca is the CEO of Leadpages. He spent 2024 through 2026 betting the company on the idea that pages, not websites, are the unit of work.