Is your competitor actually beating you in search?

'Are they beating us?' is a question about outcomes, and outcomes are the part you have least access to. You cannot see their traffic. You cannot see their conversion rate, their margins, or whether the visitors they attract are worth having. A competitor can appear above you for a term that brings them nothing while you sit below them earning the enquiries that matter, and from the outside these two situations look identical.

The answerable version is narrower and more useful: what have they done to their pages that we have not, and is any of it worth doing. That question can be worked on with public information, because a competitor's website is published to anybody who visits it. It also keeps the exercise pointed at something you can act on. The usual failure of competitor analysis is producing a document full of estimated figures that supports no decision, when the same afternoon spent reading their pages would have produced three concrete things to fix on your own.

A competitor's page source is as readable as your own, and a great deal is directly observable rather than inferred. Their title tags and meta descriptions, for every page you care to look at, which tells you what each page is aimed at and how they position it. Their heading structure, which reveals how they have organised the subject. Their structured data, which shows what they have chosen to declare about themselves. Their internal linking — which pages they link to repeatedly from body text, which is their own site's statement about what matters commercially to them.

More is visible with slightly more effort. Their site structure and how deep pages sit, from their navigation and their sitemap, which is usually at the conventional address and public. Which pages exist at all, which frequently reveals a subject they cover thoroughly and you do not. Their robots.txt, which occasionally reveals sections they would rather not have crawled. How their pages perform on a page-speed test, which anyone can run against any public URL. Whether their pages work properly on a phone. None of this is estimated, none of it needs a subscription, and all of it is checkable in an afternoon on the five pages that compete with yours.

Then the questions everyone actually asks, which are the ones public data cannot answer. Which queries they rank for and where. How much traffic they receive. How many sites link to them and which. Which of their pages earn the most. Which terms they have gained or lost recently. None of this is in their markup, and no search engine publishes it — Google does not run a public position-lookup service and does not release traffic figures for anybody's site.

Products exist that answer these, and they are worth understanding for what they are. A competitive analysis tool reports its own estimate, built from its own crawl of the web and its own modelling of click behaviour, priced according to what that costs to maintain. Two reputable vendors will give materially different traffic figures for the same site, which is not a defect in either — it is what an estimate is. That does not make them useless: a large gap between you and a competitor in a vendor's estimate is probably a real gap, and directionally useful. What it does not support is a precise claim. 'Their estimated monthly traffic is 14,300' is a number with false precision attached to a modelled figure, and repeating it inside your business as a fact is how a guess becomes a planning assumption.

This is a habit rather than a technique, and it costs nothing. When a figure from a vendor enters a document, name the vendor beside it and call it an estimate in the same sentence. Not as a disclaimer nobody reads, but in the line itself, because six weeks later the number will be quoted in a meeting by somebody who never saw the source and the qualifier will be the first thing lost.

The same discipline applies to your own inferences. 'They rank above us for this term' — checked how, on what date, from what location, on whose connection? Results are personalised and localised, so a search you ran from your office is not a measurement of anybody's position, and a private window reduces the personalisation without eliminating the location effect. If you check three times over two weeks and see the same ordering, you have weak evidence of a stable pattern. If you checked once, you have an observation. Writing down which of the two you have is the entire practice, and it is the difference between competitor analysis that informs a decision and competitor analysis that manufactures confidence.

A concrete method that produces decisions. Pick the three to five queries that matter commercially — the ones where you would genuinely want the enquiry. For each, search it in a private window and note which competitors appear. Now take the three that recur and read their pages for those queries properly, as a customer would: what does the page answer that yours does not, what does it show that reassures somebody, how quickly does it load on a phone, what is the title doing, does it answer the price question you avoid.

Then write two lists. Things they do that we should: usually specific and unglamorous — they state their prices, they show the actual practitioners, their page loads in half the time, they answer the question about turnaround that we leave to the enquiry form. And things they do that we should not, which matters as much: a page stuffed with a phrase, thin location pages for cities they clearly do not operate in, claims they cannot support. Copying a competitor's mistakes is a real risk of this exercise, because presence in results gets read as endorsement of everything on the page. Neither list needs a single estimated number, and both produce work you can start on Monday.

None of this is an argument against paying for competitive data. It is an argument for knowing what you are buying and when it becomes worth it. Two conditions make the purchase sensible. The first is that a decision genuinely turns on the figure — you are choosing between two markets to expand into, or justifying a budget, and a directional estimate of relative size would change the answer. The second is that you have already done the free work, because a vendor subscription bought before anybody has read the competitor's title tags tends to produce dashboards nobody acts on.

The sequence matters more than the spend. The free comparison produces concrete fixes; the paid data produces context and sizing. Doing them in that order means you arrive at the subscription with specific questions, which is when these tools earn their price. Doing it in the other order means paying monthly for a number you look at, feel something about, and do not act on — while the competitor's page still states a price yours does not, which was visible for nothing the whole time.