[Method] Complete KGR Strategy

AllyBarns

Regular Member
Joined
Apr 7, 2021
Messages
213
Reaction score
169
KGR - Keyword Golden Ratio

It is a technique of keyword research developed by Doug Connington. KGR technique claims to rank a blog in as little as 48 hours. He also has a YouTube channel. In one of his videos, as you can see below, many people agree that they got success within a short time span.


KGR Method:
In the KGR method, Doug suggests finding a keyword with a monthly search volume of less than or equal to 250
professional zoo.png
and no. of search results for the same keyword in Google with 'allintitle' command which should be less than 63.
allintitle_profe.png
Allintitle command shows you the blogs in search results that have the same keyword in their title. But this title may contain more additional keywords. You should not add quotes (" ") to your keyword when search with 'allintitle' because it will then show only those blogs that contain exact same keyword in the title and no additional words. Which will reduce the no. of search results that appear in Google, hence our calculation will be wrong.
Now divide the no. of search results that appear in the 'allintitle' search and divide with the monthly search volume of the keyword.

Here in this case: 5/90 = 0.05

The result comes is the KGR value of the keyword.

To rank higher in Google within a short time with a new website, Doug suggests choosing keyword with KGR value less than or equal to 0.25.
Range from 0.25 to 1.00 can work for older sites with some authority. KGR value greater than 1 means, a lot of competition.
If you choose a keyword of more than 250 MSV. Your post will go in Google sandbox and it will take time to rank. Also, the 'allintitle' results should not be more than 63 in any case. (ignore KW with more than 63 'allintitle' results)
This is it. According to Doug, the method ends here.

The method works for some people, with low competition niche sites. Choose a sub-niche, create 30-50 articles and apply for amazon affiliate.
However, many people don't find this guide super helpful. They tried it and it didn't work for them.
Below is the guest post by Doug itself:

https://mangools.com/blog/keyword-golden-ratio/
Even he stated that only 5% of your articles will rank on page 1.
Keyword Golden R.png

I am not in any way more knowledgeable than these industry experts but I want to add a few more factors to determine your winning KGR keyword.
The method is as follow:
  • Brainstorm some seed keywords and search it on Wikipedia, Google, Reddit, Quora, Answerthepublic.
  • These 5 tools by the way are the best tools for good keyword ideas.
Digital marketin.png
On Wikipedia, you can further click on links in the content section to get more subtopics.

AnswerThePublic_.png
(2) Search.png

After getting a few ideas, sign-up for Moz pro 30-day trial and search the monthly search volume and difficulty of these keywords.
Keyword Research.png
You can also look for more keyword suggestions. In the SERP analysis, you can see the PA and DA of the first three competitors.
Look for keywords with MSV less than 250 and difficulty less than 30 (25 in others like LongTailPro, KWfinder, etc.)
After finding the desired keyword, search the keyword in Google.
Look if there are any snippets or paid ads.
Look for authority websites ranking on the first page, if Wikipedia or any high authority site is there and no weak website ranking, ignore that KW.
If there is no such thing then search the keyword with the 'allintitle' command. (results should not be more than 63)
Calculate the KGR. It shouldn't be more than 0.25
Now open the first three websites in SERP and look at their post.
Check below things:
No. of words in the post (You can use 'word counter plus' chrome plugin)
No. of backlinks (any SEO tool can tell you that)
No. of internal and external links (any SEO tool can tell you that)
Copy the post, go to [https://monkeylearn.com/word-cloud]
paste the content and check which keywords are large and bold. On the right panel, you can see no. of times keyword appears in the content.
Free Word Cloud .png
While writing your post, you should also include the same or more times these keywords.
Now go to [https://www.seoreviewtools.com/content-analysis] and paste the same post. (Do not exclude images and links. Paste in the original format and not 'paste only text'). Add 'focus keyword in the text field'
This will tell you the SEO score and all on-page SEO practice done by the writer including the alt-text of the images.
Content Analysis.png

After doing all that, Create a skyscraper for your website. (skyscraper doesn't mean 4000-word article, it means 'more valuable and thorough than your competitors)
Enter the focus keyword on Quora, Reddit, Answerthepublic, Google's 'People also ask' section to add further related questions to the post.
Do the internal and external linking, create custom images in Canva. (one image after 300 words is recommended by experts)
Publish it.
Answer questions on Quora and link back.
If the post is 2000+ words. Write a summary post on Medium.
Do other practices like PBN or whatever you prefer.
Share on social media.


It is overwhelming but it will definitely work.
Add something if I missed it. Correct me if I am wrong.
 
If this is what I have to do to rank a bloody KGR keyword that gets me 10 visitors per year.... I might as well quit SEO altogether. And don't worry, I am not attacking you (on the contrary, I gave you an upvote for all of your trouble), but if this is what google wants me to do to rank for shit keywords than I would rather game the system and spam backlinks everywhere I can. Just... fuck this!
 
To add to this, search intent is absolutely essential when using KGR as a strategy. You can find lots of KGR terms but if you don't check search intent you'll likely not rank or get very minimal traffic. Secondly, I've found KGR needs to be done at scale to see consistent results. As Doug mentions I've found I get some great ranks, some average and some that just don't rank that great. Practice makes perfect with this strategy.
 
To add to this, search intent is absolutely essential when using KGR as a strategy. You can find lots of KGR terms but if you don't check search intent you'll likely not rank or get very minimal traffic. Secondly, I've found KGR needs to be done at scale to see consistent results. As Doug mentions I've found I get some great ranks, some average and some that just don't rank that great. Practice makes perfect with this strategy.
Totally agree with you. I have seen people blaming the technique. But I want to say, just use the method to determine the low competition long-tail keyword and the rest is common sense.
 
If this is what I have to do to rank a bloody KGR keyword that gets me 10 visitors per year.... I might as well quit SEO altogether. And don't worry, I am not attacking you (on the contrary, I gave you an upvote for all of your trouble), but if this is what google wants me to do to rank for shit keywords than I would rather game the system and spam backlinks everywhere I can. Just... fuck this!
I get decent visits even from keywords that show 0 MS in keyword tools. KGR keywords shouldn't be that much different. Even if it says something like 50-100 MS you'll always rank for other similar keywords + long terms which will add up.

The thing with KGR keywords is that you need to pump out content like there's no tomorrow so that it adds up over time.

Write 2 articles per day for a keyword that will get u 5 visitors / day and within 3 months you'll be getting 500-1000 visitors / day. (And that's being conservative. I get a lot more than 5 visitors/day per keywords with 0-100 MS)

I personally don't use the KGR method but I do target shitty keywords like he talks about.
 
It definitely works and it definitely has its place, with the caveats mentioned above.
 
Last edited:
I get decent visits even from keywords that show 0 MS in keyword tools. KGR keywords shouldn't be that much different. Even if it says something like 50-100 MS you'll always rank for other similar keywords + long terms which will add up.

The thing with KGR keywords is that you need to pump out content like there's no tomorrow so that it adds up over time.

Write 2 articles per day for a keyword that will get u 5 visitors / day and within 3 months you'll be getting 500-1000 visitors / day. (And that's being conservative. I get a lot more than 5 visitors/day per keywords with 0-100 MS)

I personally don't use the KGR method but I do target shitty keywords like he talks about.

too much work for me and too big of a gamble. If I have to work for 100 KGR keywords that MIGHT get me 1000 visitors a month after 1 year or work for 1 keyword that WILL get me 10k visitors a month in 6 months I choose the later.
 
Does this still work? I thought is has been debunked?
 
too much work for me and too big of a gamble. If I have to work for 100 KGR keywords that MIGHT get me 1000 visitors a month after 1 year or work for 1 keyword that WILL get me 10k visitors a month in 6 months I choose the later.
It's completely different though.

Ranking for 1 keyword that will get you 10k visitors a month implies:
1- Putting all your eggs in 1 basket
2- Investing a decent amount on backlinks (which is not necessary for KGR keywords and you can easily write the articles yourself thus it can technically be free)
3- Can lose the backlinks in the long term and with it the rankings for that keyword.

While if you take the KGR/Low MS keyword approach:
1- Your traffic will come from hundreds of different articles, if one loses ranking it doesn't matter.
2- You don't really have to invest in backlinks since you don't need them.
3- You can write the content yourself if you want which is perfect if you don't want to invest heavily into the site.


A good example that I suggest you to take a look at is:
@Shaunm and @Kikerinka Journeys

Both have extremely successful sites using low MS keywords. Kikerinka increased his sites visitors from 100k to nearly 500k within 1 year by targetting keywords that had less than 1K MS (not all from what I recall but a large majority of his keywords had less than 1K MS)
 
too much work for me and too big of a gamble. If I have to work for 100 KGR keywords that MIGHT get me 1000 visitors a month after 1 year or work for 1 keyword that WILL get me 10k visitors a month in 6 months I choose the later.

It can be a part of the 'global seo' you work on. The process can be outsourced to someone to avoid all the hard work.
 
Does this still work? I thought is has been debunked?
It is a manual process hence most people avoid this. Every new kid wants to search for some gold keywords with a bunch of free keyword tools.
 
too much work for me and too big of a gamble. If I have to work for 100 KGR keywords that MIGHT get me 1000 visitors a month after 1 year or work for 1 keyword that WILL get me 10k visitors a month in 6 months I choose the later.
Most people will agree with you, and you're right. If you have a website with good traffic and you know your game then you wouldn't like to spend time getting low traffic when you can get much more in the same time span. That's why these low SV keywords are untouched and very helpful for newbies with new websites. They can rank without doing too much link building or other complex stuff.
 
It's completely different though.

Ranking for 1 keyword that will get you 10k visitors a month implies:
1- Putting all your eggs in 1 basket
2- Investing a decent amount on backlinks (which is not necessary for KGR keywords and you can easily write the articles yourself thus it can technically be free)
3- Can lose the backlinks in the long term and with it the rankings for that keyword.

While if you take the KGR/Low MS keyword approach:
1- Your traffic will come from hundreds of different articles, if one loses ranking it doesn't matter.
2- You don't really have to invest in backlinks since you don't need them.
3- You can write the content yourself if you want which is perfect if you don't want to invest heavily into the site.


A good example that I suggest you to take a look at is:
@Shaunm and @Kikerinka Journeys

Both have extremely successful sites using low MS keywords. Kikerinka increased his sites visitors from 100k to nearly 500k within 1 year by targetting keywords that had less than 1K MS (not all from what I recall but a large majority of his keywords had less than 1K MS)

the numbers I gave are just examples to make a point, which is: if I have to work that much to rank 100s of keywords that may, or may not give traffic I'd rather work with 1-2 top keywords that I know for sure will give me traffic.

But anyway, I'm not dismissing or talking against long tails. I know that they are good and I target them too (most of the time actually). But I don't stress with this bogus metric called KGR (I have a keen aversion towards basing my SEO on 3rd party metrics, even though I sometimes do take them into account). But not for finding long tails. My process is simpler: I casually go through long lists of long tails, find a few of them that call out to me, put those in google and look at the domain names ranking in there. If I feel like I can rank on 1st page for any of the long tails that I've chosen just by skimming through the websites ranking there I go with it, if not, I search for other keywords in the long lists of 1000s of keywords that I get.

It can be a part of the 'global seo' you work on. The process can be outsourced to someone to avoid all the hard work.

true, but I rarely outsource stuff because I'm too poor (for now). And when I have money to outsource I aim for better targets than long tail keywords

Most people will agree with you, and you're right. If you have a website with good traffic and you know your game then you wouldn't like to spend time getting low traffic when you can get much more in the same time span. That's why these low SV keywords are untouched and very helpful for newbies with new websites. They can rank without doing too much link building or other complex stuff.

for newbies are great indeed. I actually target long tails, too for my sites, I just don't go through such complicated analyses and convoluted steps and headscratches. I just look who my competition is for that keyword and if the instinct tells me to go with it, I go. If not, I drop it and find another keyword. There's millions of such keywords in almost every niche under the Sun, so I don't need to waste so much time finding a few keywords that work
 
If this is what I have to do to rank a bloody KGR keyword that gets me 10 visitors per year.... I might as well quit SEO altogether. And don't worry, I am not attacking you (on the contrary, I gave you an upvote for all of your trouble), but if this is what google wants me to do to rank for shit keywords than I would rather game the system and spam backlinks everywhere I can. Just... fuck this!


Har har! I feel the same. Also my thought is: millions of f-ing articles, all saying the same shit about the same thing
 
I've read his post on KGR
He uses kw finder to assess monthly SV
But SV suggested by extensions vastly differ in some niches
That's where you could go wrong.
 
Hmmmmm...

If this is what I have to do to rank a bloody KGR keyword that gets me 10 visitors per year.... I might as well quit SEO altogether. And don't worry, I am not attacking you (on the contrary, I gave you an upvote for all of your trouble), but if this is what google wants me to do to rank for shit keywords than I would rather game the system and spam backlinks everywhere I can. Just... fuck this!

I concur with tazarbm

Countless people jumped on this (shiny new SEO toy) bandwagon when Doug first released it. What I noticed, is how many Internet Marketers have been indoctrinated and proselytized into NOT TARGETING competitive and fiercely competitive keyord phrases. To this day, those SEOers, are psychologically conditioned to ONLY target,"Low hanging fruit."

Granted, some do succeed, but many drag on - continuously suckered into purchasing the next great SEO hype tool or formula, while only experiencing a frustrating mixed bag of results.

Case in point, in the sample keyword, "professional zoom background for interview" smart SEOers examine ALL aspects of a targeted keyword. In this case, that keyword phrase was trending because ZOOM was trending . The highlight is TRENDING, it IS NOT evergreen by any stretch of the imagination.

So what happened a few months later? Take a look!

KGR Wait.png

YES! Keyord phrases like this are super easy to rank, but the PAYOFF is triple ZEROs.

This is what's happening to many newbies, they end up using KGR to target easy to rank keywords that look great, (Love at first sight) but are not MONEY KEYWORDS.

The moral of the story is; always do ALL of your SEO homework!

Does this still work? I thought is has been debunked?

It has its merits, but in order to win in today's SEO WAR, your arsenal must include way more than KGR firepower!
 
I've read his post on KGR
He uses kw finder to assess monthly SV
But SV suggested by extensions vastly differ in some niches
That's where you could go wrong.
I do not rely on SV from extensions and no one should. Always use another tool to confirm. I use extensions to filter out very high SV. It gives me idea, what keyword should I keep in my further research list.
 
Hmmmmm...



I concur with tazarbm

Countless people jumped on this (shiny new SEO toy) bandwagon when Doug first released it. What I noticed, is how many Internet Marketers have been indoctrinated and proselytized into NOT TARGETING competitive and fiercely competitive keyord phrases. To this day, those SEOers, are psychologically conditioned to ONLY target,"Low hanging fruit."

Granted, some do succeed, but many drag on - continuously suckered into purchasing the next great SEO hype tool or formula, while only experiencing a frustrating mixed bag of results.

Case in point, in the sample keyword, "professional zoom background for interview" smart SEOers examine ALL aspects of a targeted keyword. In this case, that keyword phrase was trending because ZOOM was trending . The highlight is TRENDING, it IS NOT evergreen by any stretch of the imagination.

So what happened a few months later? Take a look!

View attachment 171529

YES! Keyord phrases like this are super easy to rank, but the PAYOFF is triple ZEROs.

This is what's happening to many newbies, they end up using KGR to target easy to rank keywords that look great, (Love at first sight) but are not MONEY KEYWORDS.

The moral of the story is; always do ALL of your SEO homework!



It has its merits, but in order to win in today's SEO WAR, your arsenal must include way more than KGR firepower!
I agree with your statement "always do ALL of your SEO homework! "
I took this keyword just as an example. Although I didn't choose it after any research, Google suggested me as it is trending.
As I always say, KGR is just a technique to find competition for a particular keyword, the rest is common sense.
Searching for buyer keywords for your affiliate website is totally your responsibility and KGR will not help with that but it surely will tell you, how many people are looking for that particular product or service and how many people are offering those in the market. Easy supply and demand analysis.
 
Back
Top