Sometimes the CMS is the strategy

When people ask me about the work site I’ve been building - the Design to Manufacturing one for the day job - the question usually starts somewhere on the surface. What does the homepage look like, what’s the navigation, what’s the design system. Fair questions. Not the most interesting ones.

The actual interesting part of that site is in Sanity.

I keep thinking the people who treat a CMS as “just where the content lives” are missing something. On a B2B site where the strategy is essentially built around how topics relate to each other, the CMS isn’t a passive container. It’s where the strategy is encoded. The content model decides what can connect to what. The references decide what shows up next to what. The shape of the schema is, in a real sense, the shape of the site.

For this project, the structure is built around clusters of related topics. Pages aren’t just standalone documents. They have relationships - parent topics, sibling topics, supporting articles, case examples - and those relationships drive both the navigation experience and the SEO behaviour. A reader landing on one page should naturally find the adjacent ones. A search engine crawling the site should be able to see, structurally, that these documents are about closely related ideas.

You can fake some of that with internal linking and good copywriting, and a lot of B2B sites do. The trouble is that linking by hand becomes brittle the moment the site grows past a certain size. Add a new article, and now you’ve got to remember every existing page that should reference it. Miss a few and your cluster has gaps.

If the relationships live in the data model instead, the site can construct those connections itself. Add a new piece of content, tag it with the right references, and the related pages start showing it without anyone going back to update internal links by hand. The navigation, the breadcrumbs, the related-content blocks, even the sitemap can all be derived from the same underlying structure.

That’s not a CMS being clever for its own sake. That’s the CMS doing the strategic job that would otherwise have to live in someone’s head and a thirty-tab spreadsheet.

What I’ve enjoyed about working in Sanity for this in particular is that the schema feels like an honest object. You write what the content actually is, with the relationships you actually want, and the rest of the build talks to it cleanly. It doesn’t pretend the data model is hidden. It treats the schema as a real piece of architecture.

Which it is. Most of the time, the schema is the architecture, and the visible site is just the rendering.

Worth saying plainly: a CMS isn’t always strategic. For some sites it really is just a place to put words. But for a content-led site where SEO clustering and topic relationships are doing real work, the modelling decisions inside the CMS are the work.

The page templates are the easy part.