Competitor analysis: a repeatable method that ends in decisions, not a spreadsheet

Your search competitors are the pages already ranking for the queries you want, and that list is usually not the same as your list of business competitors. Build it from the results pages themselves: take fifteen to twenty of your commercial queries, record the top ten for each, count which domains appear most often, and the three to five that keep appearing are the ones to study. Then compare a fixed set of things, and finish by turning every gap into one of three decisions: match it, beat it, or skip it. An analysis that ends in a spreadsheet instead of a queue of work has not finished.

This page covers the four steps, what to compare, how often to repeat it, and the mistakes that waste the most time.

Step one: build the list from the results

Start with queries, not with companies.

• Write down fifteen to twenty queries that would bring somebody ready to buy. Exclude your own brand name. Include the awkward ones: "X vs Y", "X pricing", "best X for Z", "X near me" if you are local.
• Search each one logged out, in the location you sell to, and record the top ten results. Note the page type as well as the domain: a product page, a listicle, a forum thread, a directory, a news article.
• Count the domains. The pattern appears quickly. Three to five domains will show up again and again, and those are your search competitors.
• Add the ones your own Search Console data points to. Export the queries where you sit between position five and twenty, then look at who is above you on those. Those are the contests you can realistically win this year.

Two results that surprise most people are worth acting on. First, a directory, a marketplace or a review site is often the main competitor, in which case appearing on it matters more than outranking it. Second, a company you consider a rival may be absent from search entirely, which means the contest is with somebody else.

Step two: what to compare

Dimension	What you are looking for
Page type and intent	Whether the winning pages are commercial, instructional or comparison pages. Matching the type matters more than matching the word count
Subtopics covered	The questions their page answers that yours does not
Which URLs earn their traffic	Usually a short list. It tells you where their investment actually went, as opposed to where they say it went
Site structure	How they split a category into pages, and whether they have a page per use case, per location or per integration
Internal linking	Which pages they link to most from navigation and body copy, which shows what they are pushing
Referring domains	Who links to them that could plausibly link to you: associations, suppliers, press, partners
Freshness	Whether the winning pages are updated, and how visibly
Result features	Who owns the images, videos, reviews and FAQ style features on the query
Proof and specifics	Prices, numbers, named clients, photographs, specifications. This is the gap you can usually close fastest
Conversion path	What the page asks for, and how many steps it takes

For the data you cannot see directly, the usual sources are Ahrefs, Semrush, Similarweb and Moz, each with its own estimates, plus free checks that cost nothing: their sitemap.xml for the full page inventory, the Wayback Machine for what changed and when, Google's Ads Transparency Center and Meta's Ad Library for the paid messages they are running, and their job ads, which tell you what they are about to invest in.

Step three: the content comparison, done concretely

Pick one target query and spend thirty minutes on it.

• Open the top five results side by side and list every heading on each page.
• Build the union of those headings. That is the set of things a reader on this query expects to be answered.
• Mark what your page covers, what it covers thinly, and what it omits.
• Note the format that wins: a table, a calculator, a template, a short video, original photographs, a downloadable checklist.
• Write down what all five get wrong or leave out. This is the most valuable line in the whole exercise and the only one that produces a page better than theirs rather than equal to it.

Repeat for your top three queries. Three done properly beats twenty skimmed.

Step four: turn findings into three decisions

Every gap gets exactly one label, and nothing goes on the list without one.

• Match. Table stakes you are missing: a price range, delivery terms, specifications, a comparison page, structured data for reviews. Cheap, fast, do it this month.
• Beat. One or two places where you can be clearly better: real data, original photography, a tool, a template, your own case numbers. Pick a maximum of two per quarter and do them properly.
• Skip. Things that only work because of who they are: a twenty year old domain, a newsroom, a national brand budget. Naming these out loud stops a team copying a strategy it cannot fund.

Cap the output at ten items for the next ninety days, each with an owner and a date. That cap is what makes the analysis repeatable.

How often to repeat it

• Monthly: your own query and position movement, and the top ten for your three most valuable queries.
• Quarterly: the full exercise above, and refresh the competitor list, because it changes.
• Yearly: structure and positioning. Whether the category language has moved, and whether your page set still matches how people describe the problem.

The mistakes that waste the most time

• Analysing the biggest name in the category instead of the site ranking just above you.
• Using a tool's automatic competitor list as the answer. It is a starting point, and it mixes in sites that share a topic but not a buyer.
• Comparing word counts. Length correlates with covering more subtopics, and copying the length without the coverage changes nothing.
• Tracking a hundred keywords. Twenty that map to revenue will tell you more and will actually get read.
• Auditing twenty competitors. Four, properly, in a document somebody acts on.

This page covers competitor analysis for search and content. For paid media benchmarking, pull the ad libraries and your own auction insights reports, which answer a different question with different data.

Asked and answered: competitor analysis

How do I find my competitors on Google?

Search your commercial queries logged out and in the right location, then count which domains keep appearing across the top ten. That frequency count is the whole method. Confirm it against the queries in your own Search Console where you already rank between five and twenty.

How many competitors should I track?

Three to five for search, chosen because they appear repeatedly on the queries you care about. More than that and the analysis stops producing decisions.

What if a directory or marketplace outranks everybody?

Then getting listed well on it is part of the work. Claim the profile, complete it, gather reviews there, and build your own page for the same query in parallel. Both routes send the same buyer.

Do I need a paid tool?

Not to start. The results pages, your own Search Console, their sitemaps and the Wayback Machine will carry the first full round. A paid tool saves time on link and traffic estimates once you are running this every quarter.

How do I compare against a much larger brand?

Compete on specificity rather than breadth. Pick the queries where your sector, location, size or use case is named, and answer those better than a national page can. Label the rest skip and move on.

Rivals the assistant picks for you

Your method starts with search competitors, the sites that rank for your terms. An assistant picks its own competitor set, and it can include firms that never show up in your keyword reports. When a buyer asks for options, those are the options. Add the list to the match, beat or skip table and apply the same decisions to each. Some will be firms you had never listed. AI Knows Us returns the names for your questions.