Most competitor analysis is theatre, not strategy
Twenty-five years in, I have read more “competitor analysis” reports than I want to admit.
Most look impressive in an exec deck. Most do nothing for SEO performance.
One law firm I have worked with for three years was convinced its main rival was a well-known boutique two streets away. After one afternoon of proper SERP-level checking, the actual organic rivals turned out to be three solo practitioners with twelve-page websites and very tight local on-page work.
The boutique they obsessed over barely ranked.
Good competitor analysis in 2026 starts with one shift: stop listing the companies you assume are competitors, and start listing the URLs that show up next to yours when buyers search.
Who shares your SERPs is not who shares your industry conference
Business competitors and SEO competitors are different categories. The biggest firm in the city is rarely the one ranking for the high-intent queries small and mid-size firms convert on.
I run a quick SERP scrape on the 30-50 queries that genuinely matter to a client. The pages that show up repeatedly are the real list. Sometimes a forum thread or a Reddit post outranks every “official” competitor for the buyer’s actual question.
Knowing that changes the brief. You stop trying to outrank Forbes for a generic term and start outranking the three blogs that actually capture buying intent.
The four things worth extracting
When a client asks me to analyse five rivals, this is what I actually pull.
Top pages by traffic share. Not the whole site. Just the 10 to 20 URLs Ahrefs or Semrush flag as the real organic earners. The model lives in those pages, not in the homepage.
Link velocity, not link count. A five-links-a-month flow over two years beats a 200-link spike from one PR campaign. Patterns matter. Totals are vanity.
SERP feature mix. Are rivals winning featured snippets, video carousels, or the local pack? Each one tells you what Google currently rewards for that intent. The answer guides format choices for your own pages.
Real keyword overlap. Strip brand terms from both sides. Filter “common keywords” reports to queries with buying intent. The clean list is usually one tenth of the raw export.
For the methodical version of this on traffic specifically, see my competitor traffic analysis checklist.
Why most of your rivals’ pages are decorative
One number worth keeping in mind before you spend a week mapping every page on a competitor’s site.
96.55% of all pages in our index get zero traffic from Google, and 1.94% get between one and ten monthly visits.
Source: Ahrefs, 14 Billion Page Search Traffic Study (2024)
The implication is uncomfortable: most pages your rivals publish do nothing. Find their five to ten traffic-driving URLs and study those properly. Skip the rest.
The toolkit I use, with honest caveats
| Tool | What I use it for | Reality check |
|---|---|---|
| Ahrefs | Top pages, keyword overlap, link patterns | Traffic figures are CTR-model estimates, not measured data |
| Semrush | Position tracking, paid intelligence, content gap | Stronger in US English, weaker in long-tail Spanish |
| SimilarWeb | Channel split, audience overlap | Reliable mostly above 50k monthly visits |
| Google Search Console | Where you actually appear next to rivals | Free and underused, especially the Pages and Queries cross-filter |
| Manual SERP checks (VPN) | What real Spanish, French, UK SERPs look like | Tool snapshots lag the live SERP by days |
Ahrefs themselves document the gap between their traffic estimates and Google Search Console reality. Treat the numbers as ranges. Compare relative trends, not absolute figures.
Multilingual competitor analysis is plural by default
The Valencia law firm I mentioned earlier has three different competitor sets.
A Spanish set for domestic family law and inheritance queries. A French set for matrimonial cases involving French expats living on the Costa Blanca. An English set for the Anglo expat community handling probate, NIE issues, and property disputes.
Three competitor lists. Three different content briefs. Three different link strategies.
UK and US agencies regularly copy a single English-market analysis across all language versions. The output is wasted budget and content that ranks nowhere. If you operate across countries, expect to do the analysis once per language, not once per company.
For the deeper view, see best practices for multilingual SEO.
AI search shifts the question, not the method
Yes, GEO matters. Yes, ChatGPT and Perplexity citations now factor into B2B buyer journeys. The underlying competitor logic does not change much.
What I add for clients now: a small layer of “who gets cited in AI answers for our buyer prompts”. I run five to ten prompts a real prospect might genuinely use, log the cited domains, and check overlap with the traditional SERP rivals.
Often the overlap is high. Sometimes a niche site I had not flagged shows up. Either signal is useful, but neither replaces the SERP-level work.
The mistake is treating GEO as a separate exercise. It is the same SERP behind a chat layer.
Three failure patterns I keep seeing
Reading tool numbers as truth. Estimates are not measurements. The further down the long tail you go, the wider the gap between Ahrefs’ estimate and Google Analytics reality.
Treating keyword gap exports as content briefs. A 1,200-row export is a list. A brief is what survives an editor and a subject-matter expert reviewing it. The two are not the same document.
Running the analysis once a year. Markets shift. Algorithms update. New entrants arrive. For active clients I refresh the picture quarterly. For fast-moving niches like AI tooling or fintech, monthly.
When competitor analysis earns its budget
Three scenarios where it genuinely pays back.
Entering a new geographic or vertical market, where you need to learn what good looks like before committing budget.
Diagnosing a stalled growth plateau, where your team is “doing all the right things” and rankings still will not move.
Pricing or service positioning decisions, where you are weighing a shift and want to know which space is already occupied.
Outside those moments, weekly competitor reports are mostly busy work.
A second pair of eyes on your real rivals
If you suspect your team is benchmarking against the wrong list, I can pull a focused SERP-level competitor analysis for one of your priority queries inside an hour.
I do this for clients in legal, freight, real estate, and translation across six languages, working from Valencia.
The SERP does not care about org charts. Neither should your analysis.
Get in touch or look at how I run multilingual SEO programmes.
