Skip to content
Back to ideasThe spinout

Search infrastructure that changes what search engines see, with no new code. The idea that took form in the lab and became its own company.

Built by Simon Torngren

Sketch of a web page: many signals run in through a layer in the middle of the page and gather on a single target.
  • inthe site as it stands
  • the layerreads and reshapes
  • outwhat the crawler sees

What the layer changes: crawl · understanding · structure · visibility

Dynamic SEO changes how a site shows up in search without anyone touching the site itself. Titles, meta descriptions and structured data are optimised at the edge, before the request reaches the server. No new code, no deploy. Currently in private beta.

Dynamic SEO is the proof that the model holds. An observation became a product, and the product became a company of its own.

Visit site

Technical SEO is stuck in the code. Changing a title, a meta description or the structured data on a site usually takes a developer, a release and a moment of risk. For a site with a handful of pages that is manageable. For a site with thousands it becomes a bottleneck.

In practice, an improvement that takes five minutes to write can take weeks to ship. It goes into the backlog, waits for the next release and competes with everything else that has to be built. Many search improvements never happen, simply because the road out is too long.

What search engines read should be changeable without rebuilding the site. The cost rarely sits in the change itself. It sits in everything around it: the queue for the developers, the testing, the deploy and the fear that something will break under real load.

When that cost disappears, search optimisation becomes something you can work on continuously rather than in big, rare pushes. That was the idea to be tested.

Dynamic SEO puts a layer between the site and the visitor. Optimisation happens at the edge, in the network, before the request reaches your server. The site stays untouched. What changes is what gets served when a search engine or a visitor asks for a page.

The edge means the work happens close to whoever is asking, at the outer edge of the network, rather than on a central server. Because the layer answers before the request reaches your site, it can shape the title, the description and the structured data right there in the response. The site under the layer has to do nothing extra.

The platform is URL-first. It finds the addresses that already exist, categorises them and optimises each one. Nothing has to be recreated from scratch and no content is moved. The starting point is the site exactly as it looks today.

The chain is short. The whole thing takes milliseconds.

Every step applies as much to a search engine's crawler as to a flesh-and-blood visitor. Both meet the same optimised response, the moment the page is requested.

  • A request lands, from a crawler or a visitor.
  • The edge catches it before the server can answer.
  • The right version is served, with title, description and structured data tailored to the query.
  • The result shows up in the dashboard, in the rankings and in the traffic.

A page doesn't have to be presented the same way for every search. The same address can meet different queries in different ways, without duplicating the content or rebuilding the site.

A product page can meet one search on its model name and another on what it's used for. The title, the description and the structured data are shaped by the query, while the page itself stays the same.

The optimised response is generated the moment the page is requested. No CMS edit, no deploy, no risk of something breaking along the way.

A marketing team that wants to improve how its pages show up in search no longer has to book time with the developers or wait for the next release. The change is made, measured and can be rolled back from the dashboard.

It also changes who owns the search work. When every adjustment took a developer, the pace sat with engineering. When the adjustment is made in a panel, the people who understand search and business can work directly, and the developers can spend their time on the product instead of on metadata.

Above all it removes what usually slows search work down: the dependency on a deploy and the fear of touching a site that works. You try a wording, see what it does to the ranking and keep what works.

The gain is largest where discoverability feeds directly into the business and the number of pages makes doing it by hand impossible. That is where a layer that optimises continuously makes the most difference.

ecommerce

Stores with thousands of products, often across several languages. Every product page needs to be found on the right query, and the sheer count makes optimising them one by one a full-time job in itself.

marketplaces

Platforms where new inventory arrives every day. Pages have to be indexed and optimised as the catalogue shifts, without anyone touching them by hand.

content platforms

Sites with deep archives. Old articles get discovered again when they show up right in search, instead of sitting and gathering dust far down the structure.

saas and docs

Products where good discoverability means users find the answers themselves. The documentation does part of support's job, if it shows up when the question is asked.

Because the layer sits at the edge, it matters less what the site is built in. Dynamic SEO works on top of what already exists, not instead of it.

It works with just about any stack:

  • Publishing tools like WordPress, Shopify, Webflow, Wix and Squarespace.
  • Frameworks like Next.js, Astro and Gatsby.
  • Content systems like Contentful and Sanity.
  • And custom-built systems, through the same route.

How a result looks in search often decides whether anyone clicks. Structured data is what lets a page show up as more than a blue link: with price, rating, common questions or other information right there in the results.

Dynamic SEO generates that data for the pages, instead of it being maintained by hand in every template. A richer result takes up more space and draws more attention, and the visibility in the result itself is often what separates a page that gets traffic from one that doesn't.

Every change is an experiment you can follow. A new title or a new schema goes out, and the effect shows up in rankings and traffic. If it doesn't work it comes off again, straight from the dashboard.

Being able to roll back matters as much as being able to change. What doesn't work doesn't have to be argued away or waited out. It just gets removed. That lowers the bar for trying, because no change is permanent until it has shown that it holds.

Taken together, search work becomes measurable rather than a guess. You try a change, watch the numbers move and keep what delivers. The decisions rest on what the measurement shows.

Dynamic SEO came out of exactly this way of working. An observation that technical SEO should be changeable without rebuilding the site. Tested in the lab, built into a real product.

The engineering is the same craft that holds the rest of the ecosystem together: rendering, performance and infrastructure that must not fail under real load. Building something that answers at the edge, in milliseconds, for both crawlers and visitors, sets a high bar. When the idea held, it got its own name and its own home.

Formise is the lab, Memorise is the agency, and Dynamic SEO is what spun out. How the three belong together, and why this is not a corporate group, is under About.

That's why we point to Dynamic SEO. The path from observation to company has already happened once.

Dynamic SEO is in private beta. Seats are limited, and those who come in early get founding pricing that disappears at the public launch. No card is needed to reserve a place.

Private beta means the product is being tested against real sites before it opens to everyone. A search infrastructure that sits between a site and its visitors has to be stable before it scales, and the lab would rather spend that time now than later.

Taking a product that far is the proof this page is about. The next step is the public launch.

Dynamic SEO is live on its own site.

  • Does the site need to be rebuilt?

    No. The layer sits at the edge and shapes the title, description and structured data before the request reaches the server. No new code, no deploy, and the site under the layer stays untouched.

  • Does it work with our platform?

    With almost any: WordPress, Shopify, Webflow, Wix and Squarespace, frameworks like Next.js, Astro and Gatsby, and custom-built systems through the same route.

  • What happens if a change doesn't work?

    It gets rolled back from the control panel. Every change is an experiment you can follow in rankings and traffic, and no change is permanent until it has shown that it holds.

  • Can Dynamic SEO be used today?

    The product is in private beta. Seats are limited, early users get a founder price, and no card is needed to reserve a seat.

  • What does Formise have to do with Dynamic SEO?

    The idea was born and built in the Formise lab, and when it held it was spun out into its own company. What that road looks like step by step is under Process.