What Is Keyword Overlap? And Why the Percentage Misleads

Two websites. One list of search terms that both of them already rank for. That shared list is keyword overlap. Every competitive research tool will sell it back to you as a single percentage.

Sixty percent overlap with one rival. Eight percent with another. Ranked that way, the report reads like a list of who you are actually up against. It is not that list. The number measures something narrower than most people assume. It also hides one detail that changes the entire reading.

Keyword Overlap Is a Count, Not a Verdict

Pull the report and you get an intersection. Terms on your list. Theirs in the second column. Whatever sits in the middle is the overlap.

Nothing in that middle section refers to money. A term lands there because two domains both rank somewhere inside the tracked results. Position twelve qualifies. So does position ninety-four. Neither site may have earned traffic from that term all year.

That is worth sitting with for a moment. Your keyword overlap total includes plenty of terms generating nothing for anybody.

Volume does not rescue the figure either. A shared term with heavy monthly demand still splits across ten organic listings. Then a map pack takes its share. Whatever Google stacks above that takes another share. Two sites can both rank for that term. The traffic still lands near zero for each of them.

So the count answers one question only. Do these two domains show up for the same searches? Useful information. Just not the information most people think they bought.

Run the Percentage in Both Directions

Here is the part a dashboard rarely shows you. Keyword overlap is asymmetric.

Say the shared list runs to four hundred terms. Your site ranks for six hundred in total. The national brand you compared against ranks for twelve thousand. Same four hundred terms either way.

For you that intersection covers two thirds of everything you have. On their side it is barely three percent of the portfolio. You sit inside their shadow. They have not noticed you exist.

Run the arithmetic yourself instead of reading the headline figure. Open the tool. Find the raw keyword count for each domain rather than the percentage. Divide the shared total by your number. Then divide the same shared total by theirs.

Two figures now, instead of one. The gap between them tells you who is exposed. A near-identical pair means genuine rivals of similar size. Anything lopsided means one of you is a rounding error on the other’s report.

Then ask which of those two figures your strategy has been built on.

Somebody Will Say High Overlap Proves They Are the Competitor

Fair objection. It is also the reason this metric gets misread so often.

High overlap proves shared vocabulary. That is the whole of what it proves. Two Calgary businesses can share six hundred terms without ever competing for a single customer. Every shared row turns out to be a definition or a how-to that both sites happened to publish.

Meanwhile the contractor two streets over takes the phone calls you wanted. Your overlap with him might be nine terms. Nine terms carrying real buying intent behind them. He outranks you on every one.

So the objection lands, then confirms the point. Keyword overlap does identify a competitor. It identifies whoever writes about your subject. Revenue lives in a separate column. The report does not contain that column.

Most of the Shared Terms Are Filler

Sort the export by search volume first, the way everyone does. Then sort it by intent instead. The order changes what you actually see.

Informational phrasings dominate almost every overlap list. What is. How to. Why does. Both sites published something. Each one ranks. Neither one sells from it.

Strip those rows out and watch the list collapse. A four hundred term overlap can fall under thirty once you keep only the terms a buyer would type. Thirty is workable. Four hundred was noise with a percentage attached to it.

Proper keyword research does this filtering long before any competitor enters the picture. Overlap analysis simply runs it late.

The Overlap That Actually Costs You Money

Filter the export down and the useful list appears quickly.

Keep the rows that name your service. Drop the general ones. Keep the rows carrying a city name, because local intent narrows a search down to your service area. Flag every row where a competitor sits above position ten while you sit below it. Those are live losses, happening this week.

What survives is your competitive surface. Usually small. Always specific.

That short list is also the only part of the report worth acting on this quarter. Everything else in the export is context.

Two of Your Own Pages Chasing One Term Is Different

Same word, different situation entirely. Call it cannibalization when two of your own URLs chase one keyword. Competitive overlap tools compare domains against each other. They will never see two of your pages fighting internally.

Google resolves that fight by picking one URL. The other starves. Both bounce around the search results week to week because the choice keeps getting revised.

Diagnosis starts with a proper website audit rather than an overlap export. Merge the two pages or split them by intent. Then write down which page owns which term. Otherwise the problem quietly rebuilds itself over the next two years.

Where Keyword Overlap Data Stops Being Useful

Third-party tools estimate. They sample a keyword universe, then guess at positions inside it. Your overlap figure can move because a database refreshed. Nothing on either site changed.

Feature layout goes unrecorded too. Both domains rank for a term. One of them sits under an AI summary that answers the question first. The overlap row reads identically in both cases.

Paid results fall outside the report completely. A competitor can own the top of the page through Google Ads management while your organic listing technically ranks higher. Overlap will not mention it. Neither will your rank tracker.

Treat the number as a prompt for questions, then. Never as an answer by itself.

Checking Keyword Overlap for a Calgary Business

Geography shrinks the intersection fast. A Calgary plumber and a Toronto plumber can post an enormous overlap on paper. They share no customers whatsoever.

SEO Company To-The-TOP! has been running competitor comparisons for Calgary businesses since 2007. Solo practitioner. The specialist reading your overlap export is the same person building the strategy out of it.

Order matters here more than the tool does. Geography filter first. Intent filter second. Volume last, if it factors in at all. Most reports get read in the reverse order, which is exactly why they produce long lists and short results.

Businesses arriving for Calgary SEO work usually bring a competitor list assembled from overlap percentages. Plenty of those names come off within the first week. The rest get compared properly, both directions, on the terms that actually matter to the business.

Common Questions About Keyword Overlap

What is a good keyword overlap percentage?

No benchmark exists for this. A high figure against a small publisher just means you write about the same subjects. Low numbers against a direct rival can still cost you every sale in the city. Judge the terms themselves, never the ratio.

How do I find keyword overlap between two websites?

Any competitive research tool carries a domain comparison report. Enter both domains. Export the shared list. Then pull the total keyword count for each domain separately. The headline percentage hides which side of the ratio you are standing on.

Is keyword overlap the same as a keyword gap?

No. Gap analysis looks at terms the competitor holds while you do not. Overlap covers what you both already rank for. Two different columns out of one export.

Does keyword overlap hurt my rankings?

Not between separate domains. Overlap there is ordinary competition. The damaging version is two pages on one site chasing an identical term. That one already has its own name. Cannibalization.

Should I target every term in the overlap?

No. Most shared terms carry no buying intent whatsoever. Filter for service words first, then location words. Whatever remains is usually short enough to work through inside a month.

One column sits right beside the overlap in most exports. Nobody gives that one a name. The terms you rank for that nobody else on the comparison list does. Often the more interesting read.