How accurate is SEMrush's Keyword Difficulty?

elocin

Regular Member
Joined
Oct 26, 2010
Messages
451
Reaction score
183
How accurate is SEMrush's Keyword Difficulty? Also, what's your personal KD score you stay away from? Im just curious if I should even bother going for anything 60-70?
 
Its kinda good, but always remember before deciding on wether to go for a keyword or not, do analyze it manually.
 
I think it's a little accurate, compared to Ahrefs'.

A site I managed which doesn't have much authority has no problem ranking in the first page for keywords Semrush considers 'easy'.
 
Don't rely on a generated KD score. You have to google the keyword and see for yourself. If you see user generated content such as forums, reddit, quora or weaker sites then you know you can rank for it. If that is not the case then move on.

There is a place for metrics in SEO, but it is never a criteria for final selection. Metrics are great for filtering large lists, for example, if you generate a list o 5000 keywords, you want to run them through a tool to get the keyword volume and the difficulty score. Then you set a threshold for each of those and filter the list. Basically eliminating the obvious targets that a tool will tell you are too difficult (which they probably are), once that is done you manually google the remaining keywords to make your final selection.
 
I saw a report last year which tested the KD of some of the top tools and Semrush came out on top.
 
It is important to keep in mind that this is only an estimate
THIS!

And I say this for ANY SEO tool, ALWAYS take those KD scores with a grain of salt and make sure to dive deep into each keyword before making any decision.

KD scores are a great way to pre-qualify keywords but, in the end, you will need to run your own SERP analysis in order to determine YOUR true chances of ranking in the first page of Google or at the top.
My rule of thumb:
if a tool tells you that a KW is difficult, it is.
if a tool tells you that a KW is easy, it might.
This is a great way to look at it!

Don't rely on a generated KD score. You have to google the keyword and see for yourself. If you see user generated content such as forums, reddit, quora or weaker sites then you know you can rank for it. If that is not the case then move on.

There is a place for metrics in SEO, but it is never a criteria for final selection. Metrics are great for filtering large lists, for example, if you generate a list o 5000 keywords, you want to run them through a tool to get the keyword volume and the difficulty score. Then you set a threshold for each of those and filter the list. Basically eliminating the obvious targets that a tool will tell you are too difficult (which they probably are), once that is done you manually google the remaining keywords to make your final selection.
SO FREAKING MUCH THIS!

Maybe these resources will help you out...

https://www.semrush.com/blog/most-accurate-keyword-difficulty/https://www.blackhatworld.com/seo/guide-newbie-friendly-keyword-difficulty-checking.1240721/
Cheers!
 
Third-party tools show only approximate value, but comparatively, Semrush is better than other tools for keyword research.

Before you select any keyword, manual research is also to be done.
 
If KD>60 it's hard or has some competition, everything less has to be checked manually. That's at least how I do it.
 
When I use SEMrush for keyword research, I stick to maximum KD of 29. I usually go for KD lower than 10.

I found to be quite reliable because lower KD always has first page of google filled with low authority sites.

And yes, you should stay away for high KD keywords unless your site has huge authority, I mean really huge authority.
 
When I use SEMrush for keyword research, I stick to maximum KD of 29. I usually go for KD lower than 10.

I found to be quite reliable because lower KD always has first page of google filled with low authority sites.

And yes, you should stay away for high KD keywords unless your site has huge authority, I mean really huge authority.

Lower than 10 is impressive, but 29 is a decent benchmark.
 
Lower than 10 is impressive, but 29 is a decent benchmark.
True and that's the reason most of my posts get ranked within 24 hours considering my main money site has an authority score of only 28 (it's hit and miss though). But it's very hard to find such keywords with decent searches and that's the only downside.
 
Its not exact correct
I found kw with 250 vlume kd 55
No one targeted that keyword
I wrote small article and published
It became no 1 within few days
Always check manually
 
I've had decent luck with SEMrush, building content around SEMrush keyword difficulty.

It is not a magic bullet.

Depending on your niche, if it's competitive enough, you might still have a tough time even if you are targetting lower-difficulty keywords.

With that said, SEMrush's keyword difficulty ranking is not the only factor you need to consider.

You should also optimize your content by targetting your lower-difficulty keywords using content optimization tools like SurferSEO.

It also helps if you interlink strategically.

Regardless, SEMrush's keyword difficulty system is accurate enough for me as much as my blog network's performance goes.

But you also have to keep your eye on other factors.
 
It does give a general over view on the competition but I dont rely soley on it because I use my commonsense and experience too.
 
Back
Top