Keyword difficulty (KD) is a 0 to 100 score that estimates how hard it will be to reach the first page of Google for a keyword. Each SEO tool calculates it differently: Ahrefs counts links to the top 10 pages, Semrush blends links with domain authority and SERP features, and Moz weights the authority of the top results by how many clicks each position gets.
That means a keyword can score 20 in one tool and 60 in another, and both can be "right". The score is a filter for long keyword lists. The real answer to "can we rank for this?" comes from looking at the search results yourself and comparing them with what your site has.
This guide explains each tool's method from its own documentation, where the scores mislead, and a 15-minute manual check you can run on any keyword.
What keyword difficulty measures
Every major tool frames KD the same way: the effort needed to rank in the top 10 organic results. None of them can see Google's ranking systems, so each picks measurable proxies, mostly backlinks and authority scores, and turns them into a number.
Ahrefs is direct about this. Its glossary says keyword difficulty "is always only an estimation because Google does not disclose all its ranking factors" (Ahrefs, Keyword difficulty glossary). Its help centre adds that it does "not recommend you base all your SEO decisions on our KD score alone" (Ahrefs Help Center).
Do not confuse KD with the "Competition" column in Google Keyword Planner. Google defines that column as "the number of advertisers that showed on each keyword relative to all keywords across Google" (Google Ads Help, About Keyword Planner forecasts). It describes paid ad competition and says nothing about organic rankings.
How Ahrefs, Semrush and Moz calculate it
| Tool | Metric | Scale | Inputs, per the vendor's own docs |
|---|---|---|---|
| Ahrefs | KD | 0 to 100, logarithmic | Number of referring domains linking to each of the top 10 organic pages. No on-page factors. |
| Semrush | KD% | 0 to 100% | Median referring domains to ranking URLs, median dofollow to nofollow ratio, median Authority Score of ranking domains, SERP features on the keyword |
| Moz | Difficulty | 1 to 100 | Page Authority and Domain Authority of page 1 results, weighted by projected click-through rate, with adjustments for homepages, query term use and new pages on strong domains |
Ahrefs KD
Ahrefs pulls the top 10 ranking pages, counts how many websites link to each, then plots the result "on a logarithmic scale from 0 to 100" (Ahrefs glossary). Its help centre confirms the score "does not take into consideration any 'on-page SEO' factors" (Ahrefs Help Center). Ahrefs says it keeps the method simple because backlinks are the easiest confirmed ranking factor to measure.
The logarithmic scale matters. Moving from KD 10 to KD 20 is a small step. Moving from 70 to 80 means competing against pages with far more linking sites.
Semrush KD%
Semrush says KD% "measures the SEO effort it would take for your content to rank organically in the top 10 results" and uses four inputs: median referring domains, median dofollow to nofollow ratio, median Authority Score of the ranking domains, and SERP-related qualities of the keyword (Semrush Knowledge Base, What is KD%?). It publishes bands:
| KD% | Semrush label |
|---|---|
| 0 to 14 | Very easy |
| 15 to 29 | Easy |
| 30 to 49 | Possible |
| 50 to 69 | Difficult |
| 70 to 84 | Hard |
| 85 to 100 | Very hard |
Moz Difficulty
Moz says Difficulty "takes into account the Page Authority (PA) and Domain Authority (DA) scores of the results ranking on the first page of Google," modified for projected click-through rate, so higher, more visible positions carry more weight. The score "roughly correspond[s] to a weighted average of the PA of the top 10" (Moz Help Hub, Difficulty).
Moz is the most candid about interpretation. It says few keywords score below 20 or above 80, treats 20 to 35 as low, 36 to 50 as middle, 51 to 65 as tough and 66 to 80 as very difficult, and notes that keywords get "significantly tougher" around 40.
Why the same keyword gets different scores
Three reasons, all visible in the table above.
- Different inputs. Ahrefs looks only at linking domains. Moz looks at its own authority scores. Semrush mixes both with SERP features. A results page full of big brands with few links to the specific URL can look easy in Ahrefs and hard in Moz.
- Different link indexes. Each vendor crawls the web separately, so the referring domain counts behind the score differ.
- Different SERP snapshots. Scores are cached. Ahrefs notes cached KD scores "may not be the most accurate" until refreshed (Ahrefs Help Center). A results page that changed last week may still carry last month's score.
The practical rule: pick one tool and compare keywords against each other inside it. A KD of 30 means "easier than a 50 in this tool", nothing more.
What the scores leave out
KD is a measure of the competition's links. Several things that decide whether you rank are missing.
- Search intent. If Google shows comparison articles and you have a product page, the link count is irrelevant. Ahrefs' own analysis of why most pages get no search traffic lists mismatched intent as one of three main causes, alongside no search demand and no backlinks (Ahrefs, 96.55% of content gets no traffic, 2023).
- Your own site. Moz says a powerful site can rank for keywords "in the 60s and 70s", while a brand-new site may find even scores "in the 20s or 30s" tough (Moz Help Hub).
- Topical relevance. Moz's example: Moz.com trying to rank for "pizza recipes" would struggle whatever the score says.
- Page quality. Google's ranking systems aim to reward helpful, reliable, people-first content (Google Search Central, Creating helpful content). No KD formula measures that.
- Time. In a 2025 study of 2 million new English pages, Ahrefs found 6.11% reached the top 10 within a year, and the average #1 result was 5 years old (Ahrefs, How long does it take to rank, 2025). A low score does not mean fast results.
How to judge difficulty from the SERP yourself
Run this check on every keyword that survives the KD filter. It takes about 15 minutes per keyword. Search in a private window, set to the country you care about, and work through the top 10.
Step 1: confirm the intent matches a page you can build
Write down what type of page holds positions 1 to 5: guide, list, product page, category page, tool, local business, video. If 8 of 10 results are product category pages and you planned a blog post, stop here. The keyword is hard for you regardless of its score.
Step 2: look at who ranks
| What you see in the top 10 | What it means |
|---|---|
| Forums, Reddit threads, Q&A pages | Google is short of good dedicated pages. Often a real opening. |
| Small niche sites and blogs | Beatable with a better page and a few relevant links |
| Big publishers on loosely related pages | Beatable with a focused page, though the domain strength helps them |
| Government, medical bodies, major brands on exact-match pages | Hard. Expect a long timeline |
| The same 3 to 4 domains holding several positions each | Hard. Google sees them as the authorities on the topic |
Step 3: check the ranking pages' links
Open the top 5 URLs in any backlink tool and note referring domains to that specific page as well as to its domain. If the page you would need to beat has 5 linking sites and yours has 50 across the domain, the keyword is in reach. If every top 5 page has hundreds, plan for link building.
Step 4: read the pages
Ask four questions of each top result. Is it current? Does it answer the query in the first screen? Does it show first-hand experience? Is anything obviously missing? Pages that are thin, dated or generic are the weak spots in a SERP. Google's Search Quality Rater Guidelines describe how human raters judge these same qualities.
Step 5: count what sits above the organic results
AI Overviews, ads, the local pack, shopping results, video carousels and "People also ask" all push organic results down. Google lists the result types it can show in its visual elements gallery. This affects the value of ranking more than the difficulty. Pew Research found Google users clicked a traditional result in 8% of visits when an AI summary appeared, versus 15% when it did not (Pew Research Center, 2025). For questions that trigger AI answers, also check whether you are cited there. Our guides to ranking in AI Overviews and AI Overviews cover that.
Step 6: score it
Give the keyword a simple personal difficulty rating:
| Rating | Signals |
|---|---|
| Go now | Intent matches, weak or dated pages in the top 5, link gap you can close in a few months |
| Plan for it | Intent matches, strong pages, link gap of several times your current page strength |
| Skip for now | Intent mismatch, or top 10 held by authorities with exact-match pages |
A worked example
Say you run a small accounting firm and compare two keywords in one tool.
| Keyword | Tool KD | What the SERP shows | Your rating |
|---|---|---|---|
| "small business bookkeeping checklist" | Higher | Positions 2, 5 and 7 are thin listicles from 2021; one forum thread ranks | Go now |
| "what is bookkeeping" | Lower | Dictionary sites, a government small business page and two major software brands; an AI Overview answers it fully | Skip for now |
The lower score was the worse target. The checklist query has intent you can serve, weak pages to beat, and searchers who need a firm like yours. That is why the score is a first filter only.
How to use KD in a keyword workflow
- Export your keyword list with volume and KD from one tool.
- Filter out anything far above what your site has ranked for before. If your best ranking keyword scores 25, start below 40.
- Group keywords by topic. One page usually ranks for a cluster, so judge the cluster's main keyword.
- Run the SERP check on the top 20 to 30 candidates.
- Prioritise by business value first, then personal difficulty. A hard keyword that brings buyers can beat an easy one that brings nobody.
- Track rankings after publishing, and recheck the SERP for keywords that stall at positions 11 to 20. Those are usually a links or freshness problem.
Ranking position matters because clicks concentrate at the top. Backlinko's analysis of 4 million Google results found the #1 result had an average click-through rate of 27.6%, and the top 3 results took 54.4% of clicks (Backlinko, Google CTR stats). A keyword you can reach position 3 on is often worth more than one where you will sit at position 9.
For a broader view of who you are up against, see our SEO competitor analysis guide, and for link data across the industry, the backlink statistics page.
FAQ
What is a good keyword difficulty score?
It depends on your site. A new site with few links should start with the lowest scores in its tool (roughly under 20 to 30). An established site with strong links can target much higher. Moz notes keywords get significantly tougher around 40 on its scale.
Is keyword difficulty accurate?
It is a consistent estimate within one tool, and a rough one. Each vendor says its score is an estimate, because Google does not publish its ranking factors. Use it to sort keywords and confirm with a manual look at the results page.
Why is keyword difficulty different in Ahrefs and Semrush?
They measure different things. Ahrefs uses only the number of sites linking to the top 10 pages, on a logarithmic scale. Semrush combines referring domains, the dofollow ratio, domain Authority Score and SERP features. Their link indexes also differ.
Does Google use keyword difficulty?
No. Keyword difficulty is a metric invented by SEO tool vendors. Google's Keyword Planner has a "Competition" column, and it measures how many advertisers bid on a keyword. It tells you nothing about organic rankings.
How long does it take to rank for a low difficulty keyword?
There is no fixed answer. Ahrefs found that about 6% of new English pages it studied reached the top 10 within a year, so even easy keywords can take months. Pages on sites that already rank for related terms usually move faster.
Should I avoid high difficulty keywords?
Not if they matter to your business. Target them with your strongest page, build links to it, and support it with easier pages on the same topic that link to it. Start the easier pages first so you have traffic while the hard one matures.
