Search visibility, for companies that sell something real
Most mid-size companies do not lose search to a better competitor. They lose it to their own site — buried pages, a technical fault nobody sees, content written to fill a calendar. This is the order we work in and why.
The floor: things that must be true before anything else is worth doing
There is no point competing for a query while one of these is broken, and each one is checkable in an afternoon.
- Every page you care about actually gets indexed. Not "is in the sitemap" — indexed. Sites routinely carry a rule, a tag or a header that excludes an entire section, and it survives because nothing visibly breaks.
- Redirects go one hop to a live page. Chains and loops are where authority quietly evaporates. We have seen a whole site where every URL but the homepage looped, styles and all — and the homepage still looked perfect because it was being served from a cache.
- One canonical version of the site. With or without
www, with or without a trailing slash, one protocol. Two versions of every page is a self-inflicted competition. - The content is in the HTML. If the words only appear after script runs, you are relying on a renderer you do not control. Assume the crawler that matters most is the least patient one.
- The page is fast enough on a phone on a normal connection, which is not the same as fast on your laptop on office wifi.
- The structured data matches the page. Marking up something the page does not say is worse than no markup at all.
Architecture beats individual pages
The biggest single lever at this size is usually not a page — it is the shape of the site.
- One page per thing you actually sell. If a service does not have a page, it does not exist to a search engine, and "it is mentioned on the services page" is not a page. This sounds obvious and it is the most common structural fault we find.
- Stop competing with yourself. Four thin pages about variations of the same subject split the signals four ways and none of them ranks. One authoritative page, with the variations as sections, beats all four.
- Depth where the money is, not where it is easy to write. Most sites have forty posts about general topics and one thin page about the thing that actually pays the bills.
- Internal links pointing at the pages that convert. A page nothing links to is a page you have told the search engine not to care about. Your own linking is the cheapest ranking signal you control outright.
- Retire what you no longer sell — with a redirect, not a delete. A site full of pages for discontinued services teaches search engines and answer engines the wrong thing about you, and it takes months to unteach.
Which queries are worth the effort
Volume is the least useful number on the list, and it is the one most reporting leads with.
Intent first. Somebody searching how to do the thing themselves is not a buyer. Somebody searching for a comparison, a cost, a standard, a supplier or a fix at 4pm on a weekday is.
Winnability second. If the first page is entirely national directories and manufacturers, a mid-size company will not displace them with a blog post, and pretending otherwise burns a quarter. Pick the queries where the page that ranks is the kind of page you can credibly write.
Then volume. Ten searches a month by somebody about to buy beats a thousand by students.
And the ones nobody tracks: your own name plus a qualifier — reviews, alternatives, complaints, pricing, "is X any good". Those results are often the last thing a buyer reads before they contact you, and they are frequently owned entirely by third parties.
What we actually do
- Crawl and measure the site as it is, including the machine-readable files, and produce the list of things that are simply broken. This is week one and it is usually the highest-yield week.
- Fix the floor. Indexation, redirects, canonicals, speed, structured data.
- Fix the architecture. A page for each thing sold, consolidation of the duplicates, internal links pointed at the pages that convert, retirement of what is no longer offered.
- Publish depth where intent is, not where writing is easy.
- Build corroboration — profiles, listings, industry bodies, real relationships. Slowly, because the fast version of this is the version that causes problems later.
- Report against the baseline recorded before we started, per query, including the ones that did not move. What we report →
What we will not do
- Buy links. The upside is temporary and the downside lands on your domain, not on ours.
- Publish volume for its own sake. Twenty generic posts a month is a cost pretending to be an asset, and it now actively harms you with answer engines.
- Report rankings for queries nobody searches. It is the oldest way to make a flat quarter look busy, and it is the reason so many owners stopped believing search reporting.
- Take the budget while the handler is broken. If enquiries currently die between arriving and somebody dealing with them, more traffic multiplies the waste. Lead response →
FAQ
How long before search work shows?
Technical and architectural fixes can move within weeks because they remove a blocker. Competitive positions are a quarter or more. The honest answer is that the first month is repair and you should expect to see the repair, not the ranking.
Do we need a blog?
You need depth on what you sell and answers to what your buyers ask. Sometimes that lives in a blog; often it belongs on the service pages themselves, where the intent already is.
What about AI answers?
Related, and a separate discipline with its own measurement. See answer engine optimisation.
Can you work with our existing site and team?
Yes — and that is the default. Everything runs on your accounts, in your stack, documented as it is built, so it keeps working whether or not we are still here.
Answer engine optimisation · Ask what we would measure first · Back to marketing
Get in touch
Email solutions@ikonic303.dev to start with the measurement. ikonic303 is based in Colorado and works with clients nationwide.