Structured data alone moved the numbers on a 24-year-old domain.

The situation.

A national vendor of facilities-management software had a web presence most companies would envy: a domain twenty-four years old, years of accumulated content, and the kind of organic authority that only time builds. It also had the problem mature sites tend to share. Organic performance had flattened. Traffic was steady but no longer growing, the obvious optimizations had been made years earlier, and there was no active SEO program running to push it further. The site had reached the plateau that arrives once the easy wins are spent.

When I looked at what was actually there, one absence stood out. The site had no structured data. Two decades of content, and nothing in the markup told a search engine, in terms it could parse without guessing, what the organization was, how its pages related to one another, or which entity stood behind any of it. Google was inferring all of that from raw text, the same way it had been since the site launched.

Why a mature domain stalls.

On a young site, growth comes from the basics: getting indexed, earning links, publishing content. A 24-year-old domain has long since done all of that. What remains is rarely more content. It is the structural layer that tells search engines precisely what the existing content means. That layer is invisible to visitors and easy to skip, which is exactly why a mature site so often still has headroom hiding in it. The information is all there. It has simply never been made machine-readable.

What the audit found.

Before writing a line of markup, I mapped what the site did and did not already tell search engines. The answer was: very little, and what existed was incidental. No Organization markup, so nothing formally established the company as an entity. No sameAs connections tying the site to the profiles that corroborate its identity elsewhere on the web. No consistent content-type schema, so product, article, and informational pages all looked the same to a parser. No breadcrumb markup to express the hierarchy a visitor could see but a crawler could not. For a site of this age and depth, the gaps were not small oversights; they were the whole structural layer, missing. That made prioritization easy. There was no need to choose between competing fixes. The work was to build the layer that had never existed.

Why this was an unusually clean test.

The engagement had a quality I almost never get: a single variable. No content campaign running alongside it, no link building, no redesign, no migration, no other SEO work in flight. Structured data is normally one input among many, which is what makes its individual impact so hard to isolate and so easy to dismiss. Here it was the only thing changing. Whatever moved in the numbers could be attributed to one cause, and that made the result worth paying attention to.

What I built.

A complete structured-data layer, written by hand and validated, rather than generated wholesale by a plugin. The foundation was entity definition: Organization markup establishing who the company is, tied together with sameAs references to its authoritative profiles, so search engines could connect the site to a single, corroborated identity rather than treating it as an anonymous collection of pages. From there, content-type schema across the templates, so each kind of page declared what it actually was instead of leaving Google to infer it. Breadcrumb markup to make the site’s hierarchy explicit. And, where the content genuinely supported it, the supporting types that help a page earn richer treatment in results.

The discipline mattered more than the volume. Generic schema plugins tend to stamp approximate markup across everything, which can do as much harm as good when it misdescribes a page, and bad structured data is worse than none. I matched each schema type to what was true of the content it described, validated all of it against the standards search engines actually enforce, and kept it accurate. Nothing speculative, nothing keyword-stuffed. Just a precise, machine-readable account of a business that, after twenty-four years, had never had one.

How I measured it.

Because the structured data was the only change, the measurement could be simple and honest. I baselined organic impressions and clicks in Google Search Console before the work went live, then tracked the same metrics over the following months against that baseline. No attribution gymnastics, no untangling of overlapping campaigns. One change, one before-and-after.

The outcome.

Three months after the schema went live, organic search impressions were up twenty percent and clicks up ten percent, with nothing else in the mix to explain it. The structured data gave search engines a clearer, more confident read on what the site was and who stood behind it, and they responded by surfacing it more often and for more queries. The impression lift came first, which is what you would expect: the site became eligible for, and visible in, a wider set of results, and the clicks followed.

Why this is also a GEO story.

The same structured data that lifted classic rankings is exactly what positions a site for AI answer engines. When a generative engine decides which sources to trust and cite, it leans heavily on entity clarity and machine-readable content, the precise signals this work put in place. So an engagement framed as schema and SEO was also foundational generative engine optimization, building the entity understanding that earns citations in AI answers as readily as it earns rankings in the classic ones. The lesson holds well past this one client: a mature domain that looks tapped out is usually just under-described, and the structural layer that fixes it is some of the highest-leverage technical work available on an established site.