SEO competitor analysis: a 6-step method to prioritize your actions

An SEO competitor analysis compares your site with the sites holding your queries (keywords, content, links, AI answers) to rank your actions by priority.

SEO competitor analysis steps shown as icons linked by arrows, in front of three web pages ranked 1, 2 and 3

Your competitors hold queries you have been targeting for months. The usual reflex: export thousands of keywords from an SEO tool and end up with a spreadsheet nobody uses.

A useful SEO competitor analysis answers one question: which actions should you launch first to gain visibility? Now that Google shows AI Overviews and ChatGPT, Perplexity or Gemini recommend brands, that visibility plays out on two fronts: Google's classic results and AI-generated answers.

This 6-step method takes you from choosing queries to a scored prioritization grid, covering both fronts.

Key takeaways

  • An SEO competitor analysis compares your site with the sites that hold your target queries, not only with your business competitors.
  • It covers 4 dimensions: keywords, content, authority (links and brand mentions) and presence in AI answers.
  • A top 10 ranking no longer guarantees AI visibility: according to Ahrefs (March 2026), only 37.9% of pages cited as sources in AI Overviews also rank in Google's top 10 for the same query.
  • The deliverable is not an inventory: it is a list of actions ranked by a scoring grid (potential, proximity, effort, AI stakes).

What is an SEO competitor analysis?

An SEO competitor analysis (or SEO benchmark) compares your site's visibility with that of the sites ranking for the same queries. The goal: understand why they outrank you and decide what to do. It relies on data you can observe from the outside: rankings, ranking pages, content formats, link profiles, brand mentions and presence in AI answers.

It does not replace an audit, it puts it in order of priority. An SEO audit with clear priorities examines your site from the inside (technical, content, indexing). Competitor analysis sets those findings against the market: content that looks fine is much less so if the top 3 pages answer 3 times more questions.

In the SERP analyses Getfluence runs before selecting media for its clients, the same finding keeps coming back: part of the top positions is held by media, comparison sites or forums the brand had never listed as competitors. That is why the method starts from queries, not from a list of names. Links follow the same logic: backlinks still carry weight on Google and in AI answers, and comparing competitors' link profiles reveals the sites that talk about them and not about you, one of the most concrete leads in step 5.

Key point: SEO competitor analysis compares your visibility with that of the sites holding your queries, to decide which actions to launch first.

Step 1: choose the queries that matter for your business

Everything starts with scope. Comparing two domains on all their keywords produces thousands of rows unrelated to your offer: 50 to 200 well-chosen queries give a more useful analysis.

Three sources are enough to build this list:

  • Google Search Console: queries where you already get impressions between position 5 and position 20. These are the closest to a gain.
  • SEO tools: queries in your field with real search volume, filtered on your offer.
  • The field: the questions prospects ask your sales and customer service teams.

Then group these queries by topic and by intent (informational, commercial, transactional). Remove branded queries.

To cover AI answers, pair each group with 5 to 10 conversational prompts that express the same need. A payroll software vendor targeting "payroll software for small business" will also test "which payroll software should a 30-employee company choose?".

Key point: limit the analysis to 50 to 200 queries linked to your offer, grouped by intent, and pair each group with a few prompts written as real questions.

Step 2: identify your real competitors, query by query

An SEO competitor is a site that holds your queries, whether or not it sells the same thing as you. For each group of queries, record Google's top 10, then count how many times each domain appears. The ratio between that number and the total number of queries in the group gives a simple, comparable visibility share.

Keep 3 to 5 competitors per group, then sort them into 3 families:

  • Direct competitors: they sell an offer close to yours. The goal is to outrank them.
  • Editorial competitors: media, specialist blogs, comparison sites. They mostly capture informational and comparison queries.
  • Platforms: forums, Reddit, YouTube, marketplaces. They hold more and more positions on review and comparison queries.

A direct competitor is there to be outranked. A media outlet that outranks you on an informational query can also become a relay: an article mentioning you on that outlet holds the position in your place and can feed AI answers on the same topic.

Key point: your SEO competitors show up in the SERP, group by group, and split into direct competitors, editorial competitors and platforms, which call for different responses.

Step 3: measure the keyword and content gap

The keyword gap lists the queries where your competitors rank and you do not. SEO tools calculate it by comparing several domains. Isolate 2 lists: queries where a competitor is in the top 10 and you are absent, then those where you rank beyond position 10 while a competitor is on page one. Then filter on the scope from step 1: a query that does not lead to your offer is not a gap, it is noise.

The content gap is measured page by page. For each query you keep, compare your page with the top 3: questions answered, format (guide, list, comparison, tool, video), figures, examples, last update date.

Then apply a simple decision rule:

  • Create a page if no page on your site matches the intent.
  • Improve a page if it matches the same intent and ranks between position 11 and 30.
  • Ignore the query if the intent does not match your offer or if the SERP calls for a format you cannot produce.

The goal is not to copy the competitor's page. In its documentation on helpful, reliable, people-first content, Google asks: "Does the content provide substantial value when compared to other pages in search results?" It also asks whether content that draws on other sources avoids simply copying or rewriting them. The content gap shows what is missing, so you can bring what nobody provides: your own data, field experience, a method.

Key point: the keyword gap tells you where you are absent, the content gap tells you why; each query leads to a decision: create, improve or ignore.

Step 4: read the SERP and AI answers, not just rankings

Two queries with the same volume do not have the same potential. For each one, note the SERP features: AI Overview, People Also Ask, videos, local pack. An AI Overview reduces clicks to classic results: according to a Pew Research Center analysis (July 2025, browsing data from 900 US adults in March 2025), users clicked a classic result link in 8% of visits when an AI summary appeared, versus 15% without one.

Ranking well is not enough to appear in these summaries either. The Ahrefs study on AI Overview sources (March 2026, updated May 2026, 863,000 queries and 4 million cited URLs) shows that 37.9% of pages cited as sources rank in Google's top 10 for the same query, 31.2% between positions 11 and 100, and 31.0% beyond.

Test your step 1 prompts in ChatGPT, Perplexity, Gemini and Google's AI Mode, and record 2 separate pieces of information:

  • Mentions: the brands named in the text of the answer.
  • Citations: the pages shown as sources for the answer.

A competitor can be named without its site being cited, because media or comparison sites talk about it. Citations also underestimate how pages are actually used: the study The Attribution Crisis in LLM Search Results (Strauss et al., June 2025) finds that Perplexity Sonar visits about 10 relevant pages per query but cites only 3 to 4. Repeat each test: answers vary.

Key point: read each SERP (AI Overview presence, formats shown) and record separately the brands named and the pages cited as sources in AI answers.

Step 5: compare authority, between links and brand mentions

Links remain central to Google rankings. Google's guide to ranking systems states that PageRank "continues to be part of our core ranking systems". Compare each competitor's referring domains, their topics and the pages receiving those links. Volume alone says nothing: a link counts through its context (a media outlet in the field, useful content, a natural anchor).

On the AI side, the brand's presence across the web also matters. An Ahrefs study on mention types (November 2026, 10,000 US brands, 260 million prompts) measures the correlation between several types of mentions and how often ChatGPT, AI Mode and AI Overviews name a brand: 0.75 for mentions in Reddit threads that rank in Google, 0.70 for mentions on web pages, 0.67 for press articles, 0.64 for YouTube. These are correlations, measured on US brands: they do not prove that a mention makes a brand appear.

The most useful cross-check is to list the sites that talk about your competitors and not about you, then compare it with the pages cited as sources in step 4. Domains present in both lists are your authority priorities, on Google and in AI answers alike. In Getfluence, you do not have to build this list by hand: the Backlinks filter and the Getfluence assistant natively identify, at no extra cost, the domains in your field where your competitors are present and your brand is not, and suggest them as referring domain opportunities. A contextual article on one of these media outlets brings both a link and a mention, two levers that complement each other.

Key point: compare link profiles by context, not volume, and spot the media that mention your competitors and are cited by AI: they are your priority targets.

Step 6: prioritize actions with a scoring grid

The first 5 steps produce possible actions (pages to create or improve, media to target). This table gives each action a score out of 12, based on 4 criteria scored from 1 to 3.

Criterion1 point2 points3 points
Business potentialInformational query far from the offerComparison or choice queryQuery close to purchase or a quote request
ProximityNot in the top 50Position 21 to 50Position 5 to 20
Effort requiredNew page and link acquisitionNew page onlyImproving an existing page
AI stakesNo AI Overview, no competitor named in testsAI Overview shown for the queryCompetitors named or cited in AI answers, you absent

Reading the score: from 10 to 12, the action goes into the current quarter; from 7 to 9, into the next quarter; at 6 or below, the query stays under watch.

The final deliverable fits on one line per action: query or group, action (create, improve, earn a link or a mention), target page or media outlet, owner, deadline and tracking indicator. Track rankings, Search Console clicks and the brand's presence (mentions and citations) in your prompt tests every month, and run the full analysis again every quarter.

Key point: a 4-criterion grid (potential, proximity, effort, AI stakes) turns the analysis into a dated action plan, tracked monthly and reviewed quarterly.

An SEO competitor analysis is worth the decisions it triggers, not the volume of data it gathers. By starting from queries and adding AI answers to the comparison, you see who really holds your ground. Start with a single group of 30 queries: the scoring grid will give you your first 5 actions in a few hours.

Your questions about SEO competitor analysis

How do you run an SEO competitor analysis for free?

Start with Google Search Console to list the queries where you get impressions between position 5 and position 20. Then record the top 10 for each query in a private browsing window, and test the same needs as questions in ChatGPT, Perplexity or Gemini. Free versions of SEO tools complete this data, with volume limits.

What is the difference between an SEO audit and an SEO competitor analysis?

An SEO audit examines your site from the inside: technical setup, indexing, content, internal linking. Competitor analysis compares your visibility with that of the sites holding your queries, using data observable from the outside. The two complement each other: the audit lists the problems, the competitor analysis shows which ones really matter against the market.

How often should you redo an SEO competitor analysis?

A full analysis every quarter is enough in most sectors. Between two analyses, track the rankings of priority queries every month and rerun your prompt tests. Also redo the analysis after a major Google update or when a new player appears on your queries.

How do you analyze competitors in AI answers?

Write 5 to 10 prompts per group of queries, as real questions, and test them in ChatGPT, Perplexity, Gemini and Google's AI Mode. Record separately the brands named in the answer (mentions) and the pages shown as sources (citations), and repeat the tests, because answers vary from one session to the next.

Sources

  • Google Search Central, "Creating helpful, reliable, people-first content", Google Search documentation, November, 2026
  • Google Search Central, "A guide to Google Search ranking systems", Google Search documentation, November, 2026
  • Ahrefs, "Update: 38% of AI Overview Citations Pull From Top 10 Pages", Louise Linehan, March 2026 (updated May 2026)
  • Pew Research Center, "Google users are less likely to click on links when an AI summary appears in the results", July 2025
  • Strauss I., Yang J., O'Reilly T., Rosenblat S., Moure I., "The Attribution Crisis in LLM Search Results", arXiv 2508.00838, June 2025
  • Ahrefs, "Which Types of Mentions Correlate the Strongest With AI Visibility? (260M AI Prompts Analyzed)", November 2026