Keyword difficulty: how to read KD without being fooled

Keyword difficulty is the number most keyword lists are sorted by, and the one most often misread. A KD of 32 in one tool and 58 in another for the same query is not a contradiction: each tool measures something slightly different, and none of them measures your site. This article sets out what two major tools actually document, the places where the score misleads, and a four-question rule for choosing targets that does not depend on a single number.

What the number measures: two documented methods

Ahrefs describes its KD as a score from 0 to 100 based on the backlink profiles of the top-ranking pages: it takes the top 10 results for a keyword and counts the websites linking to each of them. Ahrefs is open about the limits. The score is a generic benchmark that does not account for your own site's authority, it can hide specific weaknesses in a results page, and it does not tell you whether the ranking pages actually satisfy the searcher.

Semrush defines KD% as the SEO effort it would take to rank in the top 10 on Google, and says the calculation uses factors such as the median number of referring domains pointing to the ranking URLs, the median dofollow/nofollow link ratio, the median authority score of the ranking domains and SERP-related qualities of the keyword itself. Its published bands run from 0–14 (very easy) to 85–100 (very hard). The help page does not say how much each input weighs, so the exact recipe stays closed.

Two consequences follow. First, scores are not comparable across tools, so a column that mixes KD values from two tools is meaningless. Second, both methods are essentially a measure of link strength around the top results. Neither reads the content, judges whether the intent matches the page you plan to write, or knows how strong your domain is.

Where KD misleads

The most common failure is a low score on a results page you cannot compete on. If the top 10 is made of product pages, calculators or marketplace listings, a low KD only means those pages have few links; an article will not replace them, because the result type is wrong. Check the page type before the number.

The reverse also happens. A keyword with a high score can have one thin or outdated result in the top 10, and that single weak page is the opening. Because the score summarises the whole top 10, it hides exactly the detail that matters. Mixed-intent queries are a third trap: a broad term can carry a high KD because several intents share one results page, while a narrower variant with a single clear intent is easier to serve.

Finally, difficulty is relative to the site. In practice a new domain and an established one need different cut-offs for the same list, which is why a threshold copied from a blog post ("go for KD under 30") travels badly.

A four-question rule for choosing targets

Use KD only as the last filter, after these questions, in order.

1. Is the result type one you can build? Open the results page. If it is mostly tools, product pages or forums and you plan an article, drop the keyword whatever its score.
2. Is there a weak page in the top 10? Look for thin, dated or low-authority results that rank anyway. One is a real opening; none means the page will need exceptional depth or links.
3. Does the score fit your own history? Calibrate on your site: pull five keywords where you already rank in the top 10 and record their KD in the one tool you use. That range is your working ceiling, more reliable than any generic band.
4. Does the page support a service or industry page? An article that leads nowhere is not worth a difficult keyword. Use KD to break ties between candidates that already pass the business test.

A worked illustration

The numbers here are illustrative, not data. Suppose your five existing top-10 keywords sit between KD 8 and 31 in your tool. A candidate at 28 whose results page includes one dated article passes all four questions. A candidate at 55 fails question three, however attractive its volume, and a candidate at 12 whose top 10 is entirely comparison tools fails question one. The decision came from the results page and your own history; the score only confirmed it.

Keeping KD honest in a content plan

Record the tool and the date next to every KD value, keep one tool per column, and re-pull the figures when you revisit the plan, since results pages change. Treat volume figures in the same list as planning estimates rather than facts. If you are still building the list itself, our keyword planner method shows how to group queries by intent before scoring them, and our content SEO service page describes how we turn a scored list into a publishing plan. The free SEO tools can help with the page-level checks that KD cannot see.

Sources

• Ahrefs — Keyword Difficulty: how KD is calculated from the top 10 results' linking domains, and its stated limits.
• Semrush Knowledge Base — Keyword Difficulty score: inputs and difficulty bands of KD%.