All posts
Keyword ResearchSEO

Keyword Difficulty: What the Score Measures and Misses

The Reachavo Team7 min read

Your keyword export has 600 rows and a difficulty column, so you do the obvious thing: filter to everything under 30 and start writing. Three months later half of those pages sit on page four. Meanwhile a competitor with fewer links than you has taken a keyword you skipped because it said 62.

Keyword difficulty is one of the most-used numbers in SEO and one of the least examined. It measures something real, but it measures one thing, and the thing it measures isn’t “how hard will this be for you”.

This post covers what a keyword difficulty score is built from, the four things it can’t see, why tools give the same keyword different scores, and a way to judge difficulty for your own site.

What a keyword difficulty score actually measures

Google doesn’t publish a difficulty score. Nothing in Search Console or Keyword Planner estimates how hard a keyword is to rank for organically. Every KD number you’ve seen was built by a tool vendor, from the outside.

Each vendor has its own formula, and some fold in extra inputs, but the backbone of the common ones is the same: look at the pages ranking in the top 10 right now, measure the links pointing at them, and compress the result onto a 0–100 scale. A high score means page one is held by pages with strong link profiles.

Links matter, and Google says so. Its link best practices put it this way: “Google uses links as a signal when determining the relevancy of pages and to find new pages to crawl.” Note the article: a signal. Google’s guide to its ranking systems describes them as looking at “many factors and signals”. Links are the one input a third party can observe at scale, which is why difficulty scores are built on them, not because links are the whole story.

Two more properties follow from how the score is made. It’s a snapshot: when page one changes, the score changes with it. And it’s local: the same keyword can carry a different score in each country, because each country has its own page one.

Keyword Planner’s “Competition” is not difficulty

One mix-up costs people a lot of time. Keyword Planner shows a Competition column with Low, Medium and High, and it looks like a free difficulty score. It isn’t one. Google’s Keyword Planner documentation defines it as “The number of advertisers that showed on each keyword relative to all keywords across Google.” The Google Ads API reference is blunter: it “indicates how competitive ad placement is for a keyword”.

That’s a measure of the ad auction. A keyword can be High because advertisers fight over it while the organic results underneath are thin, and a keyword nobody bids on can have a page one that’s impossible to crack. Use the column for paid search and nothing else.

Four things the score can’t see

Your site

A keyword’s difficulty score is the same for every site that looks it up. A new blog and a site that has covered the topic in depth for five years see an identical number, even though they are in very different positions.

Google’s ranking systems guide explains why that matters. “Our ranking systems are designed to work on the page level”, it says, and then: “Site-wide signals and classifiers are also used and contribute to our understanding of pages.” What your site already is goes into how your new page is understood. A score computed only from other people’s pages can’t include that.

Whether page one actually answers the query

Links build up over years, and a page keeps them long after it stops being the best answer. A results page held by an outdated roundup, a category page for an informational question, and a few forum threads can carry a high score and still be beatable, because the pages there hold their spot on accumulated links rather than on fit.

The reverse catches people too. A low score over ten tight, current guides from specialist sites can be harder than it looks, because links aren’t what’s holding those pages there.

How the strength is spread across page one

One number summarises ten pages. Two giants and eight thin pages can produce the same score as ten mid-strength pages, and those are very different opportunities. The first has room at positions three to ten; the second has room nowhere in particular. You only find out which by looking.

Whether a click exists

Difficulty tells you how hard ranking is, not whether ranking pays. If an answer box, an AI Overview or a row of ads absorbs most of the clicks, a winnable keyword can still be worth little. That check is filter four in the keyword gap analysis guide, and it applies here unchanged.

Why two tools give the same keyword different scores

Every tool counts links with its own crawler, and no two crawlers see the same web. One index finds a page’s 400 referring domains; another finds 250. That disagreement is covered in detail in the post on backlinks and referring domains, and every difficulty score inherits it.

On top of different link counts, each tool has its own formula for turning them into a score and its own idea of what 50 should mean. So a 40 in one tool isn’t a 40 in another, and a rule like “target anything under 30” doesn’t carry across tools. If you have a threshold that works, it works for the tool you built it in.

How to judge keyword difficulty for your own site

The score is a fine first pass. The method below turns it into a decision about your site, and only step one uses the number.

  1. Sort by difficulty; don’t cut by it.Group your list into rough bands (low, middle, high) and work through each. A hard cutoff at 30 throws away the 34 you’d win and keeps the 28 you won’t.
  2. Open the results page and read it. For each of the ten results, ask: does it match what the searcher wants, how old is it, and is it from a specialist or a generalist? Forum threads and stale pages on page one are the clearest sign the space is underserved.
  3. Compare the ranking pages with your closest page. Look at referring domains to the ranking URLs themselves, not their homepages, since those are the pages you’re up against. If they have a handful each and your typical guide earns a similar number, the keyword is in reach whatever the score says.
  4. Check what Google already shows you for. In the Search Console performance report, filter queries to the topic’s main words. Impressions at positions 20 to 50 for related terms mean Google already connects your site with the topic, which is the site-level context no difficulty score can include. Queries already sitting just off page one are a separate opportunity, covered in the striking distance post.

In Reachavo, Keyword Researchexpands a seed term into related keywords with difficulty shown next to volume, intent and a 12-month trend, so step one is a sort and a filter. Steps two to four are still yours, and they’re where the decision gets made.

Ours is the same kind of number

Our difficulty score is built only from links: the link profiles of the ten pages ranking now, weighted toward the pages themselves rather than their domains. It doesn’t look at content, intent or your site, and every user sees the same score for the same keyword in the same country.

It also leans toward the stronger end of page one, so a single very strong result can lift the score even when most of the page is weak. That’s a reasonable design for a warning light. It’s also exactly the case where step two above will find you an opening the number hides.

Our keyword tables also show a Comp column beside KD. That’s the advertiser competition level from the Keyword Planner section above, and the same advice applies: it describes the ad auction, not the organic results.

We think that’s fine, as long as nobody pretends otherwise. A difficulty score that claims to know how hard a keyword is for your site is claiming to see something it can’t.

What a difficulty score is good for

UseVerdictWhy
Sorting a long list into bandsGoodIt’s fast, consistent within one tool, and points your reading time at the right end of the list first.
Comparing two keywords in the same nicheReasonableSame tool, similar results pages. A big gap means something; a five-point gap doesn’t.
A hard cutoff (“only under 30”)PoorThe line is arbitrary, and the keywords either side of it are usually indistinguishable.
Deciding whether your site can rankPoorThe score is the same for every site. Your site is the variable it leaves out.
Estimating how many links you needPoorOne number blends ten pages. Look at the referring domains of the pages you’d actually displace.
Comparing scores across toolsMeaninglessDifferent link indexes, different formulas. The gap tells you about the tools, not the keyword.

The short version

A keyword difficulty score measures the link strength of the pages ranking in the top 10 today, compressed onto 0–100. It can’t see your site, whether page one fits the query, how strength is spread across those ten results, or whether a click is on offer. Keyword Planner’s Competition column measures the ad auction and isn’t difficulty at all. Use the score to sort, then read the results page, compare the ranking pages with your own, and check Search Console for topics Google already ties to you.

If you want the sorting step done in one place, Keyword Researchshows difficulty beside volume, intent and trend for every idea from a seed term. It’s a starting point for your judgement, not a replacement for it.

The tools behind this

Every one of these runs on the same 7-day free trial, from $19/mo.

Stop guessing. Start measuring.

Keyword gaps, backlinks, rank tracking and AI visibility — the whole research stack, from $19/mo.

7-day free trial Cancel anytime