Never prioritize keywords based on Keyword Difficulty.
Instead, focus on:
First, let me explain why KD is a joke of a metric.
Take a look at “semrush review” with KD of 33:
SEO tools calculate KD based on the number of backlinks results have on the SERP:
For this keyword, novice SEOs will say “oh god, we’ll need 100+ backlinks to rank.”
But what percentage of those are high-quality backlinks? Unique, dofollow links with DR > 50? In my experience, SEO tools don’t distinguish between those – but Google does.
Let’s inspect result #3, Stylefactory with 205 backlinks.
Article ranking on third spot has 8 high-quality backlinks out of 205 total. And even that’s questionable.
If you were to focus on KD, you would severely overestimate the difficulty of this keyword.
Instead, focus on these three metrics.
1. Allintitle: Results
Allintitle: results tell you how many articles on Google have your keyword in the title.
In other words, how many articles were written on the topic. The smaller the number, the fewer articles you’ll need to beat to get on the first spot. It’s a terrific indicator of “content supply” on Google.
Because of that, it’s the most important SEO metric that no one knows about.
How do you get this data?
Write “allintitle:” modifier in Google along with your keyword.
This one tells me I’ll have to beat 475 articles to rank first spot for “semrush reviews”.
There are 9,670 “what is CRM” blog post.
Good luck ranking for that.
And there are only 4 articles written on the topic of “best cameras for photo booth”.
You’ll rank tomorrow if you write an article on this.
In fact, that’s exactly what I did — the result on the image is mine.
I spotted this keyword for a client, wrote an article, and we ranked second spot in 12 hours after publishing. It then took us another month to reach first spot. We capped at 600 traffic per month, and earned $3,985 revenue in 6 months.
(Affiliate sales not included; in the article we pitched the client’s SaaS.)
2. Domain Rating
Domain Rating (DR) tells you the ranking power of a website.
It’s a complex metric, calculated based on the number of high-quality links a website is getting, and the number of links the website is giving away. Based on studies and my experience, the same article will rank higher if it lives on a domain with higher DR.
It’s possible to outrank domains with higher DR than yours with:
- more engaging content,
- better content-market fit, and
- manual link building to the target article.
But the conclusion is the same.
It’s harder to rank on SERPs filled with higher DR domains.
The good news SEO tools have this data:
The bad news is they don’t crunch it for you.
If you have 100+ keywords to analyze, it’ll be time-consuming opening every SERP and calculating the:
- average DR of the results,
- minimum DR of the results, and
- number of results with smaller DR than yours.
To automate it, you’ll have to export including SERPs and burn lots of credits.
Then do some Google Sheets magic.
It took me a good part of the weekend to fully automate my sheet.
I have more outside the screenshot, but the main metrics I crunch are:
- Total traffic (will explain in the next chapter)
- allintitle: results (find this manually)
- Minimum DR
- Average DR
- DR histogram
- Sites lower than my DR
Prioritizing keywords with this is a piece of cake.
3. Total Traffic
Like KD, volume is another joke metric.
- SEO tools are bad at estimating volumes;
- Number of actual clicks varies a lot between SERPs;
- No one knows how many keywords Google lumps together on one SERP.
So what to do instead?
I sum all traffic that top 10 results are getting.
This gives me the most accurate total search volume for a keyword and its variants. For example, I can get up to 2,480 clicks if I go after “semrush review” keyword.
When allintitle is low, generic results by high DR sites may creep on the SERP.
It happens because Google doesn’t have any other content to serve, and it can’t serve an empty SERP. So it goes for the next best thing. This will skew the traffic metric. A good article should outrank it though, so I subtract their traffic and add the average traffic to compensate.
Here’s an example for “best dog treats made in usa”, where the last result is generic:
To keep the metric standardized, I would subtract 15,395 and add 324 – the average traffic of the nine other results.
Ideally, you want high-intent keywords that have:
- High traffic potential
- Low allintitle
- Low DR
But that is hardly ever the case.
Most of the time you’ll struggle to make a compromise. The most optimal keyword prioritization will elude you no matter how long you stare at the screen. When you find yourself in a place like that, stop. Remember, these are rough numbers prone to change.
Getting it 100% right is futile.
When you’re 80% there, start executing.