You type a keyword into Ahrefs. It says KD 24. You feel a small rush of hope. Then you type the same keyword into Semrush and it says 61. Same keyword. Same day. Two numbers that tell completely opposite stories.
I've watched beginners freeze at exactly this moment. They assume one tool is lying. Neither is. The keyword difficulty score is an estimate, and every platform builds its estimate from a different recipe. Understanding what's in that recipe is the whole game—and it's the part most beginner guides skip entirely.
Key Takeaways
- Keyword difficulty score estimates how hard it is to reach the first page of Google for a specific term, usually on a 1–100 scale.
- Each tool calculates it differently: backlink profiles, page authority, search volume, and even ad competition get weighted in varying proportions.
- A low score with decent search volume is almost always a better beginner target than a high-volume keyword with a scary KD.
- The score never accounts for content quality, search intent, or how fresh your page is—those are on you.
- KD is a filter for prioritising your work, not a verdict on whether you can rank.
- Cross-check at least two tools before committing hours to a single keyword.
What is a keyword difficulty score, really?
Strip away the branding and it's a prediction. The tool looks at who currently ranks for a term, weighs how strong those pages and domains are, and converts that into a number. The higher the number, the more authority you'll need to displace them.
That's the definition. Simple enough. The trouble starts when you ask how the tool arrived at that number, because a score of 40 in one platform is not the same animal as a 40 in another.
How to read the 1–100 scale without panicking
Most platforms use roughly the same bands, though the exact cutoffs shift.
| Score range | What it usually means | Realistic for a new site? |
|---|---|---|
| 0–15 | Thin competition, often long-tail or niche phrasing | Yes—start here |
| 16–35 | Some established pages, but beatable with solid content | Yes, with patience |
| 36–60 | Competitive; you'll need backlinks and topical depth | Only with a real content strategy |
| 61–80 | Strong domains dominate the page | Rarely, and not quickly |
| 81–100 | Major brands and heavy link profiles | Not your fight yet |
Notice I said "usually." A 35 in a niche with five weak blogs is easier than a 22 in a niche where four government sites and a Wikipedia page sit on page one. The number compresses a lot of context into two digits, and some of that context leaks out.
Why Ahrefs, Semrush and Mangools never agree with each other
Here's the part that would've saved me months if someone had explained it early on. I once spent a full afternoon comparing a single keyword across three tools, convinced I'd found a bug. I hadn't. I'd found three different philosophies of measurement.
Ahrefs Keyword Difficulty
Ahrefs leans heavily on backlinks. If you want to rank, you generally need a similar volume and quality of referring domains to the pages sitting on page one. Their score is essentially a link-based difficulty estimate. A keyword with weak-ranking pages but decent traffic can score lower here than elsewhere, because nobody's built links to those pages yet.
Semrush Keyword Difficulty
Semrush folds in more variables. Alongside the authority of ranking domains, the score reflects search volume trends and the level of competition around a term—including signals from paid search activity. That's why a commercial keyword with heavy advertiser interest often scores higher in Semrush than in a purely link-driven model.
Mangools and the rest
Mangools takes yet another route, blending domain authority of the ranking URLs with volume and SERP features. Moz's version, historically, was built on page authority and domain authority figures.
Four tools, four formulas, one keyword. If your score jumps between platforms, that's not a malfunction. It's a reminder that you're reading an estimate, not a measurement.
What the keyword difficulty score ignores completely
This is where beginners get burned. A low KD feels like permission. It isn't.
- Search intent. A keyword scoring 12 will still destroy you if your page answers a different question than the searcher asked.
- Content quality. No formula reads your writing. A mediocre page on a low-difficulty term stays on page four.
- Topical authority. A site with 40 articles on one subject outranks a scattered site with 400, at the same KD.
- Freshness. Google rewards newly relevant pages for time-sensitive queries, and no difficulty score reflects that.
- Your existing domain strength. KD 30 is a wall for a brand-new site and a mild inconvenience for one with a solid link profile.
So what does the keyword difficulty score indicate in SEMRush specifically? It indicates how hard the tool expects it to be to reach the top ten, based on the competitive signals it measures. Nothing more. It's a starting filter, not a promise.
I learned this the hard way with a cooking-adjacent site I ran years ago. I chased a batch of KD 8–15 keywords, wrote eleven articles, and watched almost all of them stall around position 30. The scores were accurate about links. They said nothing about the fact that my pages were thin, my intent match was off, and my domain had no topic history. Six months of work, maybe two of those articles eventually moved.
Crossing difficulty with volume and intent
A difficulty number alone tells you half a story. The other half is volume, and the third piece is intent. Miss any one and you're guessing.
Here's a quick way to triage. Sort your keyword list by difficulty ascending, then filter out anything with volume below a floor you're comfortable with, then confirm the intent matches what you can actually produce. What survives is your shortlist.
What is a good keyword difficulty?
There's no universal answer, and anyone who gives you one is oversimplifying. What I can tell you is what worked for me: when I was starting out, I treated anything under 30 as fair game, ignored everything above 55, and accepted that some 20s would still beat me. For a site with real authority, a 45 is reasonable. For a site with none, a 45 is a trap.
The honest rule is relative. Compare the score to the strength of sites you've already managed to outrank.
Is an SEO score of 92 good?
Careful here—this is a different metric entirely, and beginners mix them up constantly. An SEO score (sometimes called an SEO audit score or health score) rates how well a page or site is optimised on technical and on-page factors. A 92 generally means you've handled most of the basics: titles, meta descriptions, internal links, page speed warnings, and so on. That's a good sign, but it says nothing about whether you can rank for a competitive term. You can have a flawless 95 on-page score and still lose to a page that scores 70 but has ten times your backlinks.
How does this fit with the 80/20 rule?
The Pareto Principle in SEO holds that roughly 80% of your results come from about 20% of your effort. A small cluster of pages, keywords and optimisation actions carries most of your traffic, rankings and enquiries. Difficulty scores are how you find that 20%—the handful of terms where a modest push produces outsized returns. Treating every keyword on your list as equally winnable is precisely the mistake the 80/20 rule warns against.
Seed keywords and the questions I get asked most
Everything above assumes you have keywords to score in the first place. That starts with a seed.
A seed keyword is your starting phrase—the broad topic you type into a tool before it generates hundreds of related variations. "Coffee brewing" is a seed. "Best pour-over method for beginners" is what you find downstream. Your entire difficulty analysis hinges on getting the seed right, because a vague seed produces a noisy list, and a noisy list makes every score meaningless.
A few questions come up again and again when I walk beginners through this.
How do I know when a keyword is too difficult?
If two different tools both score it above 60 and you don't already have pages ranking in that topic, move on. There's no prize for stubbornness. Bookmark it for a year from now.
Should I trust a single tool?
No. Cross-check at least two. Where they agree, the signal is stronger. Where they diverge sharply, dig into the SERP manually and see for yourself who's actually ranking. That manual check has saved me from bad decisions more than any score has.
The limit worth internalising
Difficulty scores are directional. They point you toward easier ground and away from hopeless fights. They don't know whether your content is any good, whether you actually understand the topic, or whether you'll still be publishing in three months.
The beginners who succeed aren't the ones with the cleverest keyword lists. They're the ones who pick reasonable low-scoring terms, write something genuinely useful, publish consistently, and check their assumptions against the actual search results instead of trusting a number. The score is a starting point. The rest is work—and no algorithm can score that for you.
