What a keyword difficulty score actually measures
Keyword difficulty is a single number, usually on a 0-to-100 scale, that estimates how hard it would be to rank on the first page of Google for a given search term. It is not a law of physics. It is a prediction a tool makes by looking at the pages currently ranking and asking, in effect, how strong this incumbent field is and how much a new page would have to do to break in. Understanding how that number is built is the difference between trusting it blindly and using it the way a strategist should: as one input among several.
The score is most useful as a triage device. When you paste a seed term into a tool and get back hundreds of related queries, difficulty lets you sort them quickly into "worth a real effort," "winnable with patience," and "not yet." What it cannot do is tell you whether your site can win, because it knows nothing about your authority or your content. Read it as a description of the competition, not a verdict on you. For a plain-language definition, see our glossary entry on keyword difficulty.
How tools build the score
Although each vendor guards its exact formula, the ingredients are well documented and broadly similar. Most models combine some version of these signals from the pages ranking on page one:
- Backlink strength of the ranking pages. The historically dominant input. Tools count the referring domains pointing at the URLs already ranking and how authoritative those domains are. A SERP of pages with thousands of referring domains scores high; a SERP of thin, barely-linked pages scores low.
- Domain-level authority. Whether the incumbents are heavyweight sites (large publishers or established brands) or smaller, beatable players. A first page dominated by Wikipedia, Healthline, or government sites carries a high score almost regardless of the individual pages.
- On-page and content depth. Some models factor how thoroughly the ranking pages cover the topic, not just how many links they have.
- SERP features and intent crowding. Ads, featured snippets, "People also ask" boxes, shopping carousels, and video packs push organic links down the page and can raise effective difficulty even when link competition looks modest.
The output is then mapped onto a 0-to-100 band. Single digits or low teens usually mean a sparse, weakly linked field; the 60s and above mean entrenched, heavily linked incumbents. Mid-range scores are where judgment matters most, because that is where authority and intent decide the outcome rather than the number itself.
Why two tools disagree about the same keyword
It is completely normal to see the same term scored 28 in one tool and 41 in another. The disagreement is not a bug; it follows from four design choices that differ by vendor:
- Different link indexes. Each tool crawls the web with its own bot and builds its own backlink database. If one index has found more referring domains for the ranking pages, its difficulty estimate climbs.
- Different weighting. A link-first model and one that leans harder on content depth or SERP features weight the same evidence differently and land on different numbers.
- Different normalization. A "40" is not a universal unit. Each vendor calibrates its own curve, so scores are only loosely comparable across tools.
- Snapshot timing. SERPs move, so two tools that sampled the ranking pages a week apart can legitimately report different competition.
The practical rule: never mix difficulty scores from two tools in one spreadsheet and compare them as if they share a scale. Pick one tool, learn how its numbers behave for terms you understand, and stay internally consistent. To see how the underlying data and scoring differ in practice, our Semrush review and Mangools review cover two very different takes — one a deep, link-rich enterprise index, the other a lighter, friendlier read aimed at smaller sites.
What a winnable score looks like
"Winnable" is relative to your site's authority, so there is no single safe threshold. The bands below are a working guide, not a guarantee — treat them as starting assumptions to test against the live SERP.
| Difficulty band | Brand-new site (little authority) | Established site (proven authority) |
|---|---|---|
| 0-15 | Primary hunting ground; realistic within months with one solid page. | Easy wins; often rank quickly. |
| 16-30 | Achievable with genuinely better content and patience. | Comfortable target; a normal part of the calendar. |
| 31-50 | Stretch; possible on standout pages, but expect months and some links. | Bread-and-butter range where authority does the work. |
| 51+ | Usually premature; earn authority on easier terms first. | Worth pursuing for high-value terms with real link investment. |
A new site should treat the 0-to-20 band as home base and resist high-volume, high-difficulty head terms. Those terms feel important, but ranking for them requires authority you have not yet built, so chasing them early burns months for nothing. The reliable path is to win a cluster of low-difficulty long-tail keywords first, accumulate the rankings and links that follow, and let that earned authority gradually unlock the harder terms.
A concrete example
Suppose you seed the term "running shoes." The head term scores very high — a first page of major retailers and publishers with enormous link profiles, so a new store has no realistic path there. But expand the seed and you surface phrases like "best running shoes for flat feet marathon training" or "how to clean white mesh running shoes" that score in the low teens, because the ranking pages are thinner and far less linked. Same seed, very different difficulty — and the winnable opportunities are the specific, lower-volume variants, not the trophy term.
Combining difficulty with intent and authority
Difficulty in isolation is half a decision. The other half is whether the term is worth winning and whether you are positioned to win it. Bring three lenses together:
- Difficulty tells you how contested the SERP is.
- Search intent tells you whether ranking will actually serve your goal. A low-difficulty term with the wrong intent is a trap — ranking first for an informational "what is X" query will not sell a product if buyers search "X pricing" or "best X for Y." Always check what the current page-one results are trying to do for the searcher; our note on search intent explains how to read it from the SERP itself.
- Your authority recalibrates the number for your specific site. A score of 35 is a stretch for a three-month-old blog and routine for a site with a mature link profile. The honest question is never "is this keyword hard?" but "is this keyword hard for me, right now?"
A simple way to operationalize all three: score each candidate on intent match (does winning it move my business?), difficulty relative to my authority (can I realistically rank?), and value (volume or commercial worth). Prioritize the terms that pass all three, not the ones with the biggest volume or the lowest number in isolation.
Do and don't
- Do open the live SERP before committing. The score is a summary; the actual results show who you would compete with and what format Google rewards.
- Do calibrate the tool against keywords you already rank for, so you learn what its numbers mean for a site like yours.
- Don't compare raw scores across two tools as if they share a scale — they don't.
- Don't let a low difficulty score override a clear intent mismatch; traffic you can't convert is a cost, not a win.
- Don't treat the number as fixed. Difficulty rises and falls as the SERP changes, so re-check before investing heavily in older research.
Used this way, keyword difficulty stops being a number you either trust or distrust and becomes what it was meant to be: a fast, imperfect filter that you finish with judgment. Pick one tool whose data you understand — our roundup of the best keyword research tools compares the options — learn how its scores behave for your site, and always read difficulty alongside intent and your own authority.