Keyword difficulty is an estimate of how hard it will be to reach Google's first page for a keyword, usually shown by SEO tools as a score from 0 to 100. Each tool builds it from its own link data, so scores disagree. Use the score as a first filter, then judge the live results against your own site.
Key takeaways
- Keyword difficulty is a third-party estimate; Google's Keyword Planner Competition column measures advertisers, not organic rankings.
- As of October 2026, Ahrefs bases KD only on the referring domains of the top 10 pages; Semrush blends links, authority and results-page features.
- Scores from different tools aren't comparable, and none knows your site.
- A five-factor check of the live results (authority, content depth, intent match, freshness, SERP features) is the more reliable test.
- New sites should start with long-tail sub-questions; in Ahrefs' March 2026 study, 31.2% of AI Overview-cited URLs ranked 11–100 and 31.0% beyond 100.
What is keyword difficulty?
Keyword difficulty (KD) is a score that estimates how much work a page needs to reach the top 10 organic results for a keyword. Tools compute it from the pages already ranking: the stronger those pages, usually measured by the links pointing at them, the higher the score.
Keyword difficulty describes the ranking pages, not you: a new blog and a 15-year-old publisher see the same score, although their chances differ. It is also a snapshot that changes when the results do.
Keyword difficulty vs search volume vs Keyword Planner competition
These three numbers get mixed up in keyword research:
Metric | What it measures | Where it comes from | Use it to |
|---|---|---|---|
Keyword difficulty | Strength of the pages ranking organically | Third-party SEO tools, each with its own formula | Shortlist keywords you might win |
Search volume | Estimated monthly searches | Google Keyword Planner and SEO tools | Compare demand between keywords |
Keyword Planner Competition | How many advertisers show ads (Low, Medium, High) | Google Ads | Gauge commercial value |
Google defines the Competition column as "the number of advertisers that showed on each keyword relative to all keywords across Google" (Google Ads Help). That is a paid-search measure; Google publishes no organic keyword difficulty score. Our guide to Google Keyword Planner for SEO covers the volume side.
How do SEO tools calculate keyword difficulty?
SEO tools calculate keyword difficulty from data about the pages in the current top 10, mostly their backlinks, but each vendor picks different inputs. Here is what Ahrefs and Semrush document, as of October 2026.
Ahrefs Keyword Difficulty (KD)
Ahrefs explains: "we analyze the search results for a keyword and look at the number of referring domains the top 10 ranking pages have" (Ahrefs). Three details on the same page change how you should read the score:
- Links only. KD "doesn't take into account any on-page factors", so content quality, intent and freshness sit outside it.
- A non-linear scale. Each value maps to the estimated referring domains needed for the first page: about 10 at KD 10, 22 at KD 20, 56 at KD 40, 202 at KD 70 and 756 at KD 90. Going from 70 to 90 means almost four times as many linking sites.
- Top 10, not the top spot. KD "evaluates the chances of getting into top 10 of search results (not top 3 or top 1)".
Semrush Keyword Difficulty (KD%)
Semrush gives difficulty as a percentage from 0 to 100, based on factors such as "the median number of referring domains pointing to the URLs that are ranking", the median ratio of dofollow to nofollow links, the median Authority Score of the ranking domains and "the SERP-related qualities of the keyword itself" (Semrush Knowledge Base). Its six bands:
Semrush KD% | Label | What Semrush says it takes (paraphrased) |
|---|---|---|
0–14 | Very easy | The best chance for new pages, least effort |
15–29 | Easy | Achievable for a new domain with quality, intent-matched content |
30–49 | Possible | Well-structured content optimized for the keyword |
50–69 | Difficult | Optimized content plus some quality backlinks |
70–84 | Hard | Well-optimized, unique content plus a lot of work on better-quality backlinks |
85–100 | Very hard | Major effort on content, link building, on-page SEO and promotion |
Moz and other tools
Moz Keyword Explorer, SE Ranking, Mangools and other keyword tools publish their own difficulty scores on their own data. Before relying on one, read its method page: which ranking pages it looks at, whether it counts links only, and what each band means.
Why do keyword difficulty scores disagree?
Keyword difficulty scores disagree because each tool measures different things with different data, then maps the result onto its own scale.
- Different inputs. Ahrefs counts referring domains; Semrush adds authority, link types and results-page characteristics.
- Different link indexes. Each vendor runs its own crawler and sees a different set of links to the same pages.
- Different scales. A Semrush 30 is the bottom of "Possible"; an Ahrefs 30 corresponds to roughly 36 referring domains.
- Different snapshots. Tools refresh results on different schedules, per country database.
- Shared blind spots. No score knows your site, and link-based scores miss what Google weighs beyond links. Google says its ranking systems "look at many factors and signals", from link analysis to "query deserves freshness" systems and passage ranking, which identifies individual sections of a page (Google Search Central).
The practical rule: pick one tool, compare keywords within it, and never compare scores across tools.
How to check keyword difficulty yourself from the live SERP
To judge keyword difficulty yourself, search the keyword in your target country, open the first-page results and score them on five factors.
- Search in a private window, set to the country and language you target.
- Note the SERP features above and between the organic results.
- Open the top 10 organic results, or at least the top 5.
- Score each factor from 0 to 2 and add up the points.
The manual keyword difficulty scoring table
Factor | What to look at | 0 = opening | 1 = contested | 2 = hard |
|---|---|---|---|---|
Authority | Who owns the ranking pages | Forums, Q&A threads, small sites in several slots | A mix of niche sites and known brands | Major brands, government or big publishers in most slots |
Content depth | How fully pages answer the query | Short, partial or off-topic answers | Decent pages with clear gaps | Complete guides covering every sub-question, with original data or tools |
Intent match | How well results fit the searcher's goal | Results only loosely match | Most fit, some don't | Every result is the right page type and angle |
Freshness | Page age vs how fast the topic changes | Outdated facts, prices or screenshots, or a topic where recency barely matters | Some pages recently updated | All pages current on a topic where recency matters |
SERP features | What sits above the organic results | Mostly plain organic results | One or two, such as People Also Ask | Ads, an AI Overview, shopping or local packs pushing results down |
Read the total:
- 0–3: target now. A better page can rank with modest promotion.
- 4–6: winnable with a plan. Close the gaps you found and support the page with internal links.
- 7–10: long-term. Build topical depth with easier sub-topics first.
Two caveats. If every result is a page type you can't build (product pages when you have no shop), skip the keyword whatever the total. And SERP features affect clicks more than rankings: a top-3 spot under an AI Overview and four ads can bring fewer visits than expected.
Tips for scoring authority and content depth
- Authority: With a backlink tool, check the referring domains of the ranking pages, not only their domains. A strong domain with a barely linked page is beatable.
- Content depth: Compare the ranking pages' headings with People Also Ask and related searches. Google's self-assessment asks whether content provides "substantial value when compared to other pages in search results" (Google Search Central), and the same page says Google has no preferred word count, so treat length as a clue, not a target.
Check content depth with RankTera's KW Research
RankTera doesn't calculate keyword difficulty or search volume; it shows how deep the competing content goes. In KW Research, enter the keyword and country: RankTera reads the top Google pages (up to 5, chosen in the Select SERP Results picker) and shows their average word count, heading structure and the terms they use under their headings with each term's average count (AVG COUNT). That fills the content-depth row with evidence instead of a guess. Take volumes from Keyword Planner and link counts from a backlink tool.
Worked example: scoring three keywords for a new composting site
This example is illustrative: the scores describe typical results pages, not a real check. A new home-composting site with no backlinks compares three keywords:
Factor | composting | compost bin for a small apartment | why does my compost smell like ammonia |
|---|---|---|---|
Authority | 2 (government, university and big publisher pages) | 1 (retailers and two niche blogs) | 0 (a forum thread and small blogs) |
Content depth | 2 (long, complete guides) | 1 (product lists without testing) | 1 (short answers, no fix steps) |
Intent match | 2 (broad guides, exactly what the query wants) | 1 (some general bin reviews) | 1 (half explain compost smells in general) |
Freshness | 0 (evergreen topic) | 1 (some list discontinued products) | 0 (evergreen topic) |
SERP features | 2 (AI Overview, videos, People Also Ask) | 1 (shopping results) | 1 (People Also Ask) |
Total | 8: long-term | 5: winnable with a plan | 3: target now |
The site writes the ammonia troubleshooting article first, then a tested comparison of small-apartment bins, then a broad guide linking both; "composting" waits until that cluster ranks. A link-only score would miss what makes the ammonia query winnable: results weak on depth and intent.
What we found: the top pages for "keyword difficulty" itself
We ran "keyword difficulty" through RankTera's KW Research on October 15, 2026, using US Google results. The five top-ranking pages it analysed came from ahrefs.com, semrush.com (two pages), seranking.com and moz.com, all SEO tool vendors. They averaged only 773 words, and the run found 59 keywords. The most-used terms:
Term | Average count per page |
|---|---|
Keyword Difficulty | 6 |
search results, Difficulty score | 3 each |
Keyword Difficulty score, search volume, difficulty scores | 2 each |
keyword difficulty checker, Authority Score, organic traffic | 1 each |
On our scoring table, these five pages rate 2 for authority, since each belongs to a major SEO tool vendor, but at that length they leave room on content depth. In our separate SERP analysis of six top Google results the same day, the median page had 2,089 words and none of the six contained a table. When strong domains rank with short pages, the opening is depth, not authority.
How to find low competition keywords as a new site
Low competition keywords are specific queries where the current results are weak, dated or off-intent, and for a new site they usually sit in the long tail. Start there; head terms come later.
Where low competition keywords hide
- Sub-questions of a head term. People Also Ask, related searches and ranking pages' H2s show what a broad keyword breaks into; our guide to long tail keywords covers seven ways to collect them.
- Forum threads or outdated pages on page one. Few dedicated, current pages exist for that need.
- Mismatched results. If searchers want a comparison and the results are generic guides, the right format can win.
- Your near-misses. In Search Console, queries where your page sits on page two or three are ones Google already connects to your site.
Low reported volume isn't a reason to skip a keyword: a query you can rank for brings more visits than a popular one you never reach, and one page often ranks for many variants.
Why sub-questions matter more in AI search
Ranking for sub-questions can also get a new site into AI answers sooner. Google says AI Overviews and AI Mode may use a "query fan-out" technique, "issuing multiple related searches across subtopics and data sources", and show a wider set of supporting links than a classic search (Google Search Central). In Ahrefs' study published in March 2026 (863,000 keyword results pages, 4 million AI Overview URLs), 37.9% of cited URLs also appeared in the first 10 positions of the query's results page (Ahrefs counted SERP features as positions), while 31.2% sat in positions 11–100 and 31.0% beyond the top 100 (Ahrefs). Ahrefs' July 2025 study had put the top-10 share at about 76%. The author suggests fan-out queries may explain the drop but doesn't prove it: Ahrefs says it also improved its parsing since the July 2025 study so it can see more AI Overview citations, and Google doesn't reveal the fan-out queries it runs. Treat sub-questions as a likely route into AI Overviews, not a guaranteed one: a page that answers one sub-question well may be cited before the site ranks for the head term.
Group winnable keywords into clusters
One winnable keyword rarely moves a new site. Ten related ones, each on the page that fits its intent and linked to a pillar page, build the depth that makes harder keywords reachable; our guide to topic clusters shows how to plan them. Commit to a keyword when at least two factors score 0 and you can add something the ranking pages lack, such as a tested fix or original data.
Keyword difficulty checker: free tools and how to use them
A keyword difficulty checker returns a difficulty score for a keyword you enter. As of October 2026, Ahrefs offers a free Keyword Difficulty Checker that estimates "how hard it will be to get into top 10 search results" for a keyword, and the same page maps each KD value to the referring domains needed (Ahrefs). SE Ranking, Mangools, Moz and most other SEO suites show difficulty inside their keyword tools; check each vendor's site for what its free version includes.
Whichever checker you use:
- Stick to one tool, so every score in your list comes from the same formula.
- Set the country database you target; the same keyword can score differently by country.
- Run your whole list at once if the tool accepts bulk input, then sort by score.
- Shortlist the lower bands and run the five-factor check on each before you commit.
Even Ahrefs says KD "should not be the only factor affecting your keyword choice" and recommends looking at the results page yourself.
Frequently asked questions
How is keyword difficulty calculated?
Most tools calculate keyword difficulty from the pages ranking in the top 10. Ahrefs counts the referring domains linking to those pages and maps the result onto a non-linear 0–100 scale. Semrush combines median referring domains, link types, the Authority Score of the ranking domains and features of the results page. Neither score looks at your own site, so confirm it against the live results.
What is a good keyword difficulty score for a new website?
A low score in the tool you use: in Semrush, the "Very easy" and "Easy" bands (0–29); in Ahrefs, a KD around 0–20, which Ahrefs maps to up to about 22 referring domains. Because these scores ignore content quality and intent, check the results page too: weak, dated or off-intent results can make a mid-range keyword winnable.
Does Google have a keyword difficulty score?
No. Google doesn't publish a keyword difficulty score. The Competition column in Google Keyword Planner shows how many advertisers showed ads on a keyword relative to all keywords, which is about paid search, not organic rankings. Every keyword difficulty score comes from a third-party SEO tool with its own formula.
Do keywords matter anymore with AI search?
Yes, as questions to answer rather than phrases to repeat. Google says its AI features may run several related searches across subtopics to build an answer, so pages covering those sub-questions can appear among the supporting links. Knowing which keywords and sub-questions people search still tells you which sections a page needs and which searches a new site can realistically win.
What are the four types of keywords?
Keywords are usually grouped by search intent into four types: informational (learn something), navigational (find a specific site), commercial (compare options before buying) and transactional (buy or act now). Intent decides which page type ranks, so it is part of judging difficulty: a keyword is only winnable if you can build the page type the results show.
Judge keyword difficulty with evidence, then track the result
Keyword difficulty scores are a useful first filter; the results page, scored on authority, depth, intent, freshness and SERP features, is the real test. RankTera doesn't give keywords a difficulty score or a search volume; its KW Research shows what the top pages contain, and the Rank Tracker follows your Google position for each keyword and whether ChatGPT and Gemini cite your domain. Start your 14-day free tier — no card needed; it includes 3 keyword researches, 3 AI articles and weekly tracking for 300 keywords. Plans are on the pricing page.