The Keyword Golden Ratio – A Step By Step Case Study
If you are not sure what KGR is, why it will work for your site and how to find them you can view Doug's video, which takes you through the logic and process.
As you can see, the process is as follows:
- Find keywords.
- Find the local monthly search for those keywords (LMS).
- Remove any keywords that have LMS greater than 250.
- Google for the "allintitle" results.
- Divide the allintitle by the LMS.
- If the result is lower than 0.25, write content on that subject!
Easy as pie 🙂
Here is a screenshot from Serpstat showing the ranking results for my 4 recent KGR articles I published.
As you can see, 3 out of the 4 are ranking on the first page of Google with the other moving around on page 2.
Plus, you can see that I sneaked in a KGR article where the LMS was actually 590! Naughty me! However, it had so little number of allintitle results, it was worth the risk!
The table below (extracted from Google Analytics) shows the traffic and stats to these 4 articles over the last 30 days!
The good thing about using long tall keywords is that each of these articles are not just ranking for the KGR keyword, but many other longer and shorter tails too.
So... step by step, how can you find KGR keywords?
It all starts with a good list of keywords that have been gathered from as many sources as possible.
1. Find your keywords
You can't really start this process without a list of keywords to research the KGR ratio for, so getting a list that is as extensive as possible is the best place to start.
There are many different ways that this can be achieved and there are many different free and paid tools you can use.
I will only be looking at free tools here, so there is no need to get your credit card out yet!
When ever you find some keywords using these tools, we will be exporting them into an Excel spreadsheet or Google Sheet. Alternatively, you can just dump them to a text editor such as notepad.
1.1 Answer The Public
Answer The Public is an excellent site to gather some good question based keywords. You simply enter your starting keyword and it will show you a large number of queries such as "How to xxx" and "Why does xxx".
Let's do a test.
During this process, I will be pretending to find keywords for a site about dog and puppy training. It's quite a competitive niche, so finding some good KGR keywords might be tricky, but lets do a live test!
I will start by putting "dog training" into Answer The Public as my seed keyword.
Click the "Get Questions" button and the baldy man (no.. not Doug Cunnington this time!) will work his magic.
You will be presented with a visual view of the queries it has found.
This is all very pretty and interesting to see how the data is linked together, but we just want the keywords!
Click the "Download CSV" button in the top right hand corner of the page to get the data in format that we can load into Excel (or your favorite spreadsheet app).
Open the CSV file and you will see something like this:
The keywords are in column C and scrolling down, there are 822 in total from that 1 search! A good starting place.
Some of the keywords will be garbage, but we are not going to do any sanity checking yet - we simply want to get as many as we can.
Highlight the keywords in column C and then copy and paste them to either a new blank spreadsheet, or into Notepad for safe keeping.
Then, save this blank sheet/notepad so you don't lose anything!
Repeating the process on Answer The Public for "Puppy Training" gives me another 730 keywords to copy and paste on to the end of the "Dog Training" list.
Ubersuggest is another well known tool for getting keywords from the Google predictive search (where you see search suggestions as you type in the Google search box).
Enter your keyword into the search box and click the "Suggest" button.
Ubersuggest will grab the results and show them in a list as follows:
They even provide the US monthly search volume here if you have Keywords Everywhere installed, which is useful, but we can get that data later on in this process, so no need to worry about it now.
Click the "View as text" button and then copy and paste the results on to the bottom of your master keyword list, saving it again!
Again, I repeated a search for "puppy training" which added 335 more keywords to my master list (which is now at 2235 keywords!)
1.3 Keyword Shitter
Yes - I am sorry for the name of this tool, but Keyword Shitter is very good and works in a similar way to Ubersuggest, scraping the Google predictive searches.
However, it takes it 1 step further by re-querying the keywords it gets, feeding the results back into itself.
Whilst this is a great idea and very useful, it can mean that it goes on forever, especially if it generates lots of keywords from your initial searches.
I will enter my 2 seed keywords into the tool:
Click the "Shit Keywords!" button to start and watch it go!
I let it run for a few minutes, which returned another 1583 keywords for my master list. I suspect it would have gone on all night if I had let it!
You can go ahead and copy and paste the list into your master keyword list.
2. Get the monthly search volumes for each keyword.
As part of the KGR process, we need to weed out keywords that have a local monthly search (LMS) of greater than 250.
To do this, we need to use another tool, Keywords Everywhere, to get this done.
Keywords Everywhere is a browser extension (for Chrome and Firefox only) that helps provide you with the LMS for keywords as you search in Google, but it can also be used to bulk check keywords too, which is what we will be doing.
To install Keywords Everywhere, follow the instructions here for the browser you are using. You will also need to get an API key from them to make it all work - but they will take you through this process.
2.1 Upload your keywords.
Click the Keywords Everywhere button on your browser toolbar and choose "Bulk Upload Keywords".
Also check that you have the right country selected too.
2.2 Paste in your keyword master list.
Copy the list of keywords from your spreadsheet or notepad and paste them into the box in the middle of the page.
Click the "Get Search Volume and CPC" button to start.
If it gets stuck (which it does for me sometimes) simply close the page down and start again.
You will get some results and any duplicate keywords will be removed at the same time.
Click the "Excel" or "CSV" button to download the data into a file we can work with to filter it further.
The raw data in Excel looks like this:
I always do a little tidy up and remove most of the columns I don't need.
I deleted columns A, C, D, E, G and H.
I then added 2 new columns for "allintitle" and "KGR"and the sheet looks like this:
Unfortunately, on my PC the search volume is imported into Excel as text, rather than as a number, so sorting it from highest to lowest is not that easy.
Instead, I add a filter to the data and then remove the values that are above 250 and any that are 0.
Now, we have a list of 1127 keywords that we can search the KGR "allintitle" results for.
But hey! This is going to take ages doing 1 at a time, so is there an easier way to do this?
Insert a column in between columns A (Keyword) and B (Search Volume).
In the first cell, enter this formula:
This will take the value (keyword) in cell A2 and add the Google "allintitle" search link around it. You can then copy this formula down to the other rows in your new column to automatically create the other links.
Adding "&pws=0" to the end of the URL should prevent Google from showing you "personalised results" where it pushes sites you might have already visited to the top of the search results, or shows local businesses near to where you live.
Now, to do the "allintitle" search, simple click the cell and it will open the Google search in your browser for you. If Excel starts to generate errors when you click the link, you may just need to copy and paste the data from the cell into your browser.
Make a note of the number of search results and enter it into the Allintitle column on your spreadsheet.
I have done the first few as an example for you:
Ok - we are seeing some quite high "allintitle" results here - but we have over 1k keywords to check, so there will be some good ones in there somewhere.
Unfortunately it is a slow process that you will have to go through by hand.
I tend to get blocked by Google every 6-10 requests, asking me to confirm I am not a robot to proceed.
2.4 Work out the KGR
This is the fun part!
Once you have some "allintitle" values, you can enter a formula in your KGR column to work out the ratio.
Allintitle divided by search volume:
Remember, we are looking for a KGR value that is 0.25 or less to make it a viable keyword to write an article for - there are none that meet that criteria so far.
I guess it is important to also apply some logic and reasoning to the keywords you check the allintitle results for.
If you look at my example, I searched for "how dog training" which (if the ratio was ok) will be a hard keyword to write a KGR article for (especially if you are matching the title exactly).
I did some more searches and came up with 2 results that meet the criteria and that I could write articles about quite easily:
"Best collar for large dog that pulls" and "best dog leash for pullers" could be good review type articles that list the top 5 recommended products, with links to Amazon where I could earn an affiliate payment if people click through and buy.
Wrap up - and where did I go wrong?
Is this an easy process? No.
Does it take a lot of time? Yes.
Are there tools out there that will automate it for me? Not really...
My personal view is that you should be looking to create your KGR content around keywords that have a commercial intent, such as "dog lead reviews" or "best dog lead for small dogs".
These kinds of articles will be easy to write but will also help you get traffic and earnings more quickly.
It does also seem to follow that the higher the number of words in your keyword, the more chance there will be of it being a good KGR keyword.
One thing I have realised too, is that using Ubersuggest and Keyword Shitter probably duplicates the same work, so just use Keyword Shitter instead.
Finding some KGR keywords during this step by step guide was actually harder than I thought and perhaps I would have been better to use the seed keywords of "best dog" and "best puppy" and feed those through Keyword Shitter, which would have generated lots of keywords that were more commercial in their intent.
Update: 18th October 2017.
I decided to do a little more testing after not getting many good KGR keywords from my "Dog training" and "puppy training" seed keywords.
I entered "Best Dog" into Answer The Public and followed the same process of feeding the results back into Keywords Everywhere, then out into Excel, filtering out the ones with >250 and 0 search volume and removing ones that were not relevant like "best dog kennel nyc".
I then added a column to show the length of the keyword, using the =LEN(A2) function in Excel.
I then sorted the data to show the longer keywords first and set about getting the allintitle results for each one.
Once I had grabbed a few to use as an example, I sorted by KGR and these were my results:
It is clear that using more commercial keywords makes sense and checking longer keywords might also return better results. The "best dog teeth cleaning treats" and "best dog house for hot weather" are good ones!
I also experimented with using a private browsing window in Chrome (incognito) to see if that made any difference to the process. It did seem to allow me more allintitle searches before Google wanted me to confirm I was not a robot, so this might be a good way to speed things up.
Update: 6th November 2017.
Another tip I have started to use is where you find keywords with an LMS of 0.
Normally, you would think that these should be filtered out and that you shouldn't be writing articles where noone is searching for that keyword- right?
No.. no always!
Don't reject these immediately.
I have started to find the keywords where the LMS is 0 and change it to 10. This gives me more keywords to research the allintitle for and more options to bring traffic.
Look at this example where I have used the Google Chrome Keywords Everywhere addon to show the search volumes for these "related" results that appear at the bottom of the page when I search Google for "best dog collars":
Why would Google show "Perri's Padded Leather Dog Collar" as a possible search when there are 0 searches per month for it?
To me, that makes no sense - surely Google is showing us very relevant searches that people are using - so don't ignore the 0 LMS keywords!