The Journal

Too Big for Squarespace, Too Small for Adobe: An AI CMS That Fits

Enterprise CMS platforms can be too complex and too expensive for a small marketing team, but the basic builders are too limiting. Our middle ground is an intuitive AI-supercharged CMS you actually own.

Andrew Fairlie
Written by
Andrew Fairlie
Technical Director · 13+ years with Craft CMS
Craft CMS Published 2 Sept 2026 6 min read

Sound familiar? The enterprise platforms like Sitecore, Adobe and Optimizely are too complex and too expensive for a team your size but the website builder you started on is too limiting to grow with.

I think most UK companies fall into this spot. There's around 5.7 million SMEs (making up roughly three-fifths of private-sector jobs and over half of all turnover!). But most of the software that would help them grow is built and priced for the higher end. The good stuff wears an enterprise badge, an enterprise contract and an enterprise onboarding fee, while everyone else gets the cut-down version or nothing.

Meh!

What's out there?

The enterprise DXPs

Sitecore, Adobe Experience Manager and Optimizely sit at the top end. These are sprawling Digital Experience Platforms with AI woven deep into them. Sitecore has in-editor content generation and machine-learning personalisation that segments audiences in real time. Adobe's AI auto-tags and smart-crops your assets. Optimizely's assistant has moved past writing copy into genuinely agentic territory, running experiments and building content models on your behalf.

It's mature, serious technology built for large organisations, and the licence costs run into six figures a year. Wildly out of proportion for most businesses.

The fully headless platforms

Contentful, Sanity and Storyblok represent the modern, API-first approach. Their in-editor AI is decent, and in one respect they're ahead of the big enterprise suites: several already ship support for emerging agent standards, so AI tools can read and act on your content the way it's meant to be read.

The catch is that headless is developer territory by design. You have to bring the capabilities the enterprise platforms include as standard, like analytics and personalisation, meaning stitching on yet more third-party tools and costs. Great flexibility, but only if you've got the engineering resource and budget to wield it.

The all-in-one marketing suites

HubSpot is an obvious example, with its own AI layer spanning writing assistants, agents and enrichment. If HubSpot is already your entire world, having everything in one place is appealing.

But the AI that actually that is appealing is often behind the Professional and Enterprise tiers, metered by usage credits on top of your subscription, with onboarding fees to match. The model rewards large teams with predictable, high-volume usage. For a smaller team the bill climbs quickly and unpredictably, and it only really works when that one platform is your single source of truth for everything.

The website builders

At the more accessible end sit Wix, Squarespace and the like. Cheap, quick, and increasingly AI-assisted for building a site from scratch. For a simple brochure site, fine.

But you're renting a closed platform and there's a ceiling on how far you can grow, limited control over how things actually work, and your content and data live on their terms rather than yours. The moment your ambitions outgrow the template you have to re-think everything.

Wordpress

WordPress powers a huge share of the web, and for good reason: it's open, self-hostable and endlessly extensible. Like Craft, it's a platform you genuinely own rather than rent. On AI, WordPress 7.0 brought AI foundations into the core, with a single place to connect providers like Anthropic, OpenAI or Google, and a vast ecosystem of AI plugins sits on top, from content assistants to SEO tools.

The side-effect of that is that WordPress's AI story is a build-your-own affair: you pick the plugins, wire up the providers, reconcile a handful of separate tools with their own interfaces, pricing and update cycles, then keep the whole stack patched and secure. Assembling a coherent, well-maintained AI setup from that is entirely possible.

But instead of weilding the tools to make your work better, you can easily spend your time wrangling the tools to just work.

Our way

We build on Craft CMS. It allows us to model content properly instead of forcing everything into blog-shaped boxes (which matters a lot for structured content / GEO), it's API-first so nothing is off-limits when you want to extend it, and you own it outright: your site and your data are yours, not leased back a tier at a time.

Because it's open, you can add precisely the AI you want without ripping anything out, paying for a hundred features you'll never touch, or migrating into someone else's ecosystem. You keep the ownership of an open platform without the second job of stitching a dozen tools together and keeping them patched.

That's what we built Mutual One to do. It's our managed support service and marketing platform, built directly into Craft. It lives where your team already works and it's aimed at improving the marketing outcomes an SME cares about. A few examples:

  • An assistant that speaks CMS
    Editors can talk directly to a built-in assistant, right inside the CMS. Ask it to make sense of your analytics, dig into a form submission, explain a security alert or sort out an SEO question, and get a straight answer in plain language, without raising a ticket or waiting on a developer to translate.
  • Writing and optimisation as you go
    Descriptive alt text written automatically on every image upload. Spelling and grammar checked in real time. An SEO keyword optimiser that shows how your content maps to search intent while you're still writing it. AI summaries that double as meta descriptions.
  • Audience Lens
    Give the AI a persona and a goal, and it reviews your page from that person's perspective, then hands back surgically precise fixes rather than vague advice.
  • Search that understands intent
    A vector database of your content means site search returns the right page even when visitors don't know the exact term for it. The same engine powers a chatbot that answers your visitors only from your real content, so it stays accurate and on-message.
  • Analytics that live in the CMS
    Privacy-first, no cookie banner, and your data never leaves your server (unless you decide to export it, or ask AI about it). You get real Core Web Vitals from actual visitors, navigation paths, and a view of which AI crawlers are hitting your site and which AI engines are sending you traffic.
  • Optimised for AI search
    Automatic generation of an LLMs.txt file and support for the emerging Web MCP standard mean AI search engines and agents can properly understand and act on your content.

There's loads more: accessibility reviews, bulk translation, content approval workflows, redirect management, weekly digests to your inbox. And these are the capabilities you'd otherwise pay enterprise money for.

Own the foundation, add the intelligence

I think that the businesses that power most of the UK economy shouldn't be the ones locked out of the tools that help them grow. Capable, AI-enhanced marketing tooling shouldn't be the preserve of six-figure contracts and enterprise onboarding fees.

If you're interested in how Craft CMS paired with Mutual One can help bridge the gap between SME and enterprise, reach out - hello@mutual.agency

Andrew Fairlie
Andrew Fairlie
Technical Director · 13+ years with Craft CMS

Andrew is Technical Director at Mutual, a Craft CMS Partner agency. He has been building with Craft CMS since its public beta in 2012 — working through every major version from Craft 1 to Craft 5 — and has delivered over 100 sites for clients including Apple, Transparency International, and Arts University Bournemouth.

He writes about Craft CMS on the Mutual blog and has contributed to net Magazine. At Mutual, he leads development of Mutual One, a marketing platform built on Craft CMS as its foundation.

He has spoken about Craft CMS to undergraduate students at the University of Brighton and Canterbury Christ Church University, and appeared on the Devmode.fm podcast. He has also trained other developers on working with the platform.

More from The Journal