Talk to us Run a free check

Structured data

Structured data adoption is rising. Here is how to measure it yourself

In short

Do not take anyone's word for adoption rates, including ours. Fetch the homepage and one deep page from ten competitors, count the JSON-LD blocks on each, and check whether they resolve into one connected graph or several unconnected ones. In most sectors you will find that presence is now common and correctness is still rare. Presence is becoming table stakes; the shape of the graph is where the remaining advantage sits.

Get AI Visible ·

We get asked a version of this constantly: is everyone doing this now, or are we early? It is the right question, because it decides whether this is an advantage you are building or a deficit you are closing.

The honest answer is that it varies enormously by sector, and that you should not accept a number from a vendor — including us — when you can measure your own market in an afternoon.

Measure your own sector in twenty minutes

Take ten competitors. For each, fetch the homepage and one page that actually sells something — a product, a property, a trip, a service page. Then count.

for host in competitor-one.co.uk competitor-two.co.uk; do
  n=$(curl -sL "https://$host/" | grep -c 'application/ld+json')
  echo "$host: $n JSON-LD block(s)"
done

That gives you presence. Presence is the easy half. The half that matters is shape, and for that you need to look at what the blocks contain:

curl -sL https://competitor-one.co.uk/ \
  | grep -o '"@type":"[^"]*"' | sort | uniq -c | sort -rn

Now you can see whether they have one considered graph or five plugins arguing. Three things to note as you go:

  1. How many blocks per page. One is a decision. Five is an accident.
  2. Whether @ids are present. No @ids means no connected graph, which means each block is an island a parser has to reconcile by guessing.
  3. Whether more than Organization and BreadcrumbList appear. Those two arrive free with most CMS plugins. Anything beyond them — Product with a real Offer, FAQPage, Service — is someone having actually thought.

What the sweep usually shows

We run this before every audit, and the pattern is consistent enough to describe, though your sector will differ and that is exactly why you should run it rather than trust the generalisation.

Presence is no longer a differentiator. Most established businesses have some JSON-LD, because it arrived with a theme, a plugin or an SEO module, and it has been sitting there for years doing a partial job.

Correctness still is a differentiator. The common state is several unconnected blocks, an Organization that disagrees with the one two templates over, prices in the markup that no longer match the prices on the page, and @type values chosen by a plugin author who did not know the business. All of that parses. None of it resolves cleanly into an entity.

That gap — between having markup and having markup that means something — is the part still worth competing on.

Why the direction of travel is one way

Two things are pushing adoption, and neither is going to reverse.

The first is that the cost has collapsed. Structured data used to be a specialist task; it is now a template change in most stacks, and increasingly a default in frameworks and CMS platforms. Anything that becomes a default gets adopted by everyone eventually, including the people who never decided to.

The second is that the incentive has changed shape. Structured data used to be justified by rich results — stars, prices and images in a search listing. That was a marginal, contested benefit. Being resolvable by a model that is answering a customer’s question directly is a different order of argument, and it is landing with people who ignored the first one.

What this means for timing

Be clear-eyed about what kind of advantage this is.

If you do this now, you are ahead on shape, not on presence. That lead is real but it decays, because the floor is rising underneath you. The businesses that benefit most are the ones that treat it as infrastructure — generated from the same data that renders the page, so it stays true — rather than as a project that ships once and is accurate for a quarter.

The ones who will struggle are those who wait until it is universal and then discover that their catalogue data was never structured well enough to express in the first place. That is the expensive version of this problem, and no amount of markup fixes it after the fact.