Your search competitors are not your business competitors
The first mistake is choosing the wrong companies.
The businesses you lose deals to are not necessarily the pages you lose rankings to. For a productised SEO service, the competitors in a pitch might be three agencies — but the pages occupying the search results are as likely to be a software vendor's blog, a marketplace listing and an industry publication.
Search competitors are whoever occupies the results you want. Sometimes that is a rival. Often it is a publisher with no commercial interest in your market at all, which is a different problem requiring a different response.
So build the list from the SERPs, not from memory. Take ten to twenty queries you genuinely want, search each, and record who appears. The domains that recur are your search competitors, whatever your sales team would say.
Third-party numbers are estimates
Every competitive tool reports estimated traffic, estimated keywords and estimated authority. None has access to a competitor's analytics. They model from a sample of keywords, an assumed click-through rate by position, and their own index.
That makes them useful for comparison and direction — this competitor grew, that one declined, this page attracts more terms than that one — and unreliable as absolute figures. A tool reporting "42,000 monthly visits" is stating a modelled number, and two tools will disagree substantially about the same site.
Use the relative signal. Ignore the decimal places.
The four things worth extracting
1 · Keyword gaps — terms they rank for and you don't. The headline feature of every tool, and the one most misread. The list will be long, and most of it will be irrelevant or unwinnable. What matters is the subset that is *commercially relevant to you* and *plausibly within reach*. A gap analysis is a starting list, not a plan.
2 · Their best-performing pages. More useful than the keyword list, because it shows what *format* works in your market. If the top pages for your terms are all comparison tables, the market is telling you what it rewards.
3 · Their content structure. How the site is organised, what clusters exist, what they have chosen not to cover. Gaps in their coverage are often more interesting than their strengths.
4 · Where they are vulnerable. Pages ranking on old content, thin pages holding a position they do not deserve, a topic covered once and abandoned. These are easier targets than their strongest pages.
What not to copy
A strategy you cannot fund. If a competitor ranks through hundreds of earned links and a content team, copying the tactics without the resources produces a worse version of their site. Compete where you can actually win.
Their keyword list wholesale. Their business is not yours. Terms that convert for them may be irrelevant to what you sell, and chasing their full list spreads you thin.
Their site structure. It reflects their history and constraints, including their mistakes. Structure should follow your own demand and services.
Whatever ranks, on the assumption it ranks because of what you can see. A page may hold position one because of domain authority accumulated over a decade, not because of anything on the page. Copying the page copies the visible part of an invisible advantage.
Reading the gap list properly
The filter that turns a 4,000-row export into a plan:
- Is it commercially relevant? Would someone searching this ever buy what you sell? Cut everything that fails this, however good the volume.
- Is the intent right? Informational queries in an AI-answered market may bring no clicks. Comparison and pricing queries still do.
- Can you realistically compete? Look at who currently ranks. If the first page is entirely established brands with far more authority, that term is a long-term goal, not a next action.
- Do you already have a page for it? Frequently yes, and improving it beats writing another — which also avoids creating cannibalisation.
- Is there one page that could cover several of these? Gaps cluster. Ten related terms often want one strong article, not ten thin ones.
Steps 1 and 3 remove most of the list. That is the point — an unfiltered gap analysis is a way to feel busy.
Watching the paid side
Competitors' paid campaigns are visible in the same tools, and they carry information their organic pages do not: a keyword somebody pays for repeatedly, month after month, is a keyword that converts for them. Nobody sustains spend on terms that do not pay.
That makes an established paid list a useful commercial filter on an organic plan — the reverse of the ad-to-organic bridge, applied to someone else's spend rather than your own.
Frequently asked questions
How do I find my SEO competitors?
From the search results, not from your sales pipeline. Search ten to twenty queries you want to rank for and record which domains recur. Those are your search competitors, and they are frequently publishers or marketplaces rather than the businesses you compete with commercially.
Are competitor traffic estimates accurate?
They are models, not measurements. No third-party tool sees a competitor's analytics — figures are estimated from a keyword sample and assumed click-through rates, and tools disagree substantially. Use them for relative comparison and direction, not as absolute numbers.
What is a keyword gap analysis?
A list of terms competitors rank for and you do not. It is a starting list rather than a plan: most rows will be commercially irrelevant or unwinnable, and the value is in the subset that is both relevant to what you sell and realistically within reach.
Should I target every keyword a competitor ranks for?
No. Their business is not yours, and chasing their full list spreads effort across terms that will never convert for you. Filter for commercial relevance first, then for whether you can realistically compete.
See where you actually stand
Competitive analysis goes wrong in two directions: an unfiltered export that becomes a to-do list nobody can execute, or copying a strategy that depends on resources you do not have.
If you want a clear read on where you stand against the sites actually occupying your results — organic and in AI answers — the Organic & AI Visibility Benchmark measures it — $297, one-off, no retainer.