When I inherited the search function, it looked like most in-house SEO does: a long list of tasks with no line connecting any of them to money. Pages were being published. Keywords were being tracked. And nobody in the room could tell me which of it had produced a single inquiry. That is not a marketing problem. It is an operational efficiency problem, and it is the exact problem consulting is built to solve.
I run search inside the luxury real estate brokerage I run search for, a business selling branded residences and villas to cross-border buyers. The stakes per inquiry are high, which made the absence of measurement even harder to accept. So I stopped thinking like a marketer and started thinking like someone auditing a factory line. Every station had to justify its existence by what came out the other end.
SEO consulting starts by tying every page to an inquiry
The first thing I did was unglamorous. I mapped every page on the site to a single question: does this exist to earn a qualified inquiry, and can I prove whether it does? Most pages failed the second half. We had no clean way to see which URL a buyer touched before they filled out a form.
Fixing the measurement came before writing a single new word. I set up proper tracking so each inquiry could be traced back to the page and the query that produced it. Only then did the picture become honest. A handful of pages were doing almost all the work. A long tail of older content was producing nothing and quietly costing us in maintenance and crawl budget.
- Producers: pages that mapped directly to a real inquiry in the tracking. These got more investment.
- Passengers: pages with traffic but no inquiries, usually the wrong intent. These got reworked or merged.
- Dead weight: pages with neither traffic nor inquiries after a fair window. These got pruned without ceremony.
That single classification changed how the team spent its hours. We stopped feeding the long tail and put the time where inquiries actually came from. It is the same move a good consultant makes on any process: find the stations that produce, and stop paying for the ones that do not.
Killing work that does not produce
The hardest part of this discipline is discipline. It is easy to keep publishing because publishing feels like progress. I had to be willing to look at a content calendar the previous team was proud of and cut half of it because none of that half traced to an inquiry.
One rebuild paid for the whole exercise. Our structured data was a mess, which meant the pages I most wanted buyers to find were showing up as plain blue links while thinner competitors got rich results. I rebuilt the site's structured data from the ground up, following Google's structured data guidance rather than the half-implemented markup we had inherited.
Within weeks the listing pages I had prioritized were rendering properly in search, and the inquiry numbers on those pages climbed. That work sits behind our portfolio of branded residences and villas, which is now the single strongest source of qualified contact.
Building a system the business can repeat
A one-time win is not a pipeline. The reason the number held at 31 inquiries and did not fall back is that I turned the wins into a repeatable system rather than a series of heroics.
The insight that made it repeatable was about the buyer, not the technology. Our buyers search in French and Arabic, and they start searching months before they ever travel to view a property. Most of our competitors published only in English or produced weak local-language pages. So I built genuine buyer guides in Arabic aimed at the questions people ask early in that journey. Because the competition was thin, those pages ranked fast, and they pulled in exactly the cross-border buyers we sell to. When I benchmark what earns links and rankings in this space I lean on data from Ahrefs rather than guesswork, so the content decisions are grounded in what the demand actually looks like.
What made this consulting rather than tactics is that I documented it. There is now a written system for how a page earns its place, how we measure whether it produced, and when we retire it. Anyone joining the team can run it without me in the room. The function moved from a cost line the finance side eyed with suspicion to a measured pipeline with a number attached, and that number is defensible because every part of it traces back to a source.
The lesson generalizes past SEO. Any function that cannot connect its work to an outcome will eventually be treated as a cost and cut. The way you protect it is not by producing prettier reports. It is by building the measurement first, killing what does not produce, and turning the survivors into a system. That is the whole of operational efficiency, applied to search.
