Keyword difficulty scores, compared: what each vendor is actually measuring
Ahrefs counts links to the top ten and says the result is always an estimation. Semrush names six inputs and calls its score the most accurate on the market. Both quotes are real.
The short version
Keyword difficulty is not a standard, it is a product. Each vendor invents a formula, scales it 0 to 100, and gives it the same name — so a 42 in one tool and a 42 in another are two unrelated statements that happen to share a number.
The clearest evidence is that the two biggest vendors describe their scores in flatly incompatible terms. Ahrefs says difficulty "is always only an estimation because Google does not disclose all its ranking factors". Semrush says its "Keyword Difficulty metric is the most accurate on the market". Both sentences are quoted from the vendors' own pages.
What Ahrefs says it counts
Ahrefs publishes its method in one sentence: "We pull the top 10 ranking pages for your keyword and count how many websites link to each." The result is plotted "on a logarithmic scale from 0 to 100".
It then explains the narrowness deliberately: "We intentionally keep our KD calculation methodology simple because backlinks are probably the only easily measurable confirmed ranking factor." That is an unusually candid thing for a vendor to print, and it tells you exactly what the score does not consider — content quality, topical authority, search intent, page experience, or anything else Google weighs.
The logarithmic scale matters more than people notice. On a log scale the gap between 10 and 20 is not the same amount of work as the gap between 70 and 80, so the number is not linear in effort even within its own model.
What the other big vendor says it counts
Semrush describes a broader recipe. Its difficulty and personal difficulty scores "rely on analyzing Google's top 10 organic results and several keyword characteristics", and it names six inputs: backlinks to the top-ranking sites, measured as median referring domains and the median ratio of follow to nofollow links; the authority scores of those sites; the keyword's search volume; the presence of features on the results page; whether the term is branded; and the number of words in the search term.
It also publishes the words it attaches to each band, which Ahrefs does not. Those bands run from Very easy at the bottom to Very hard at the top, with Possible sitting in the middle of the range — and they are the only published translation from a number to a decision that any vendor on this site offers.
Why the two scores cannot be compared
One score is a count of linking domains on a logarithmic curve. The other blends six factors including the keyword's own volume and word count. Feeding the same keyword into both will often produce numbers tens of points apart, and neither is wrong, because they are answering different questions.
This has a practical consequence people get burned by: a difficulty threshold learned in one tool does not travel. If you have taught yourself that anything under 30 is worth attacking, that rule is about the tool you learned it in and needs relearning from scratch in another.
What difficulty is genuinely good for
- Sorting a list you already have. Within one tool, difficulty is a reasonable way to order a hundred candidate keywords from easiest to hardest.
- Spotting outliers. A keyword with decent volume and unusually low difficulty is worth opening the results page for, which is the actual research step.
- Arguing about priorities with a colleague who can see the same screen.
What it is not good for is prediction. No score here knows how good your page will be, how well your site is trusted, or what the results page will look like in three months. Treat it as a filter that saves you time, not a forecast.
What the other tools do
Every tool on this site carries a difficulty number of some kind, and most publish nothing at all about how it is built. Mangools reports a difficulty score alongside volume and claims on its product page to have "the most accurate keyword difficulty", without describing the calculation. SE Ranking lists keyword difficulty among the metrics its Keyword Tool returns. Moz reports a difficulty figure alongside an organic click-through estimate and a priority score that blends several of its own metrics together.
Absence of a published method is the norm, not the exception, and it is worth holding in mind when a score feels authoritative. A number between 0 and 100 looks like a measurement whatever produced it, and most of these were produced by a formula nobody outside the company has seen.
The habit that beats any score
Open the results page. Every tool that computes difficulty does so by looking at the current top ten, and you can look at the same ten for free. If they are all national newspapers and product pages from billion-dollar brands, no score is needed. If two of them are thin forum threads and a six-year-old blog post, no score should stop you.
That is also the honest reason the free tools survive: a difficulty number is a convenience, not a capability. Where the free keyword tools stop sets out what you genuinely lose by not paying, and why volume numbers diverge for the same reason is the companion to this page. If you are choosing between tools, the eight tools and their entry prices is the comparison.
Questions about keyword difficulty
What is a good keyword difficulty score?
It depends entirely on which tool produced it and how established your site is, because each vendor builds the score differently. Semrush publishes bands for its own score running from Very easy through Possible to Very hard; Ahrefs publishes no equivalent table and says its score is "always only an estimation". A threshold you learned in one tool does not transfer to another.
Is Ahrefs or Semrush difficulty more accurate?
Unanswerable, and this site did not test it. Ahrefs counts linking domains to the current top ten on a logarithmic scale from 0 to 100 and says so. Semrush names six inputs and claims its metric is "the most accurate on the market". One of those is a description of a method; the other is a marketing claim; neither is a measurement anybody outside the vendors can check.
Why is keyword difficulty 0 for a keyword nobody ranks for?
Usually because the formula found very few linking domains among the current results, which is what a low score is built to report. It does not mean the keyword is easy to win — it may mean the results page is filled with the kind of result your page is not, such as a video carousel or a map pack.
Should I only target low-difficulty keywords?
As a starting strategy on a new site, largely yes, because ranking anywhere earns more than ranking nowhere. But difficulty ignores how well your page matches what the searcher wants, which is the factor you control most. A high-difficulty keyword with a badly served results page is sometimes a better bet than an easy keyword nobody needs answered.
Do free tools give keyword difficulty?
Google's free tools do not. Keyword Planner shows advertiser competition, which is about bidding rather than organic ranking, and Search Console shows your own average position. A difficulty score computed from the live results is one of the clearest things you are actually paying for when you buy a tool.
There is no partner link anywhere on this page, so nothing on it earns this site money. Where a tool named here does run a programme we take part in, it is listed — with the ones that pay nothing and never will — on the page that sets out how we get paid.
Picking a tool rather than reading about one? The eight keyword research tools with their entry prices side by side is the page to start from, and what the free tiers actually give you is worth reading first if you have not spent anything yet.