DAN, WHY ARE YOU DOING THIS TO US?!?!?!?!?!?!?!?!?!
Sorry, friend. But it just has to be this way.
For those who haven’t been playing along at home, I did a comparison study on using 3rd party tools for search volume data against native Google data. If you haven’t read it, I highly suggest checking it out before continuing.
In the discussion that ensued, I got a great idea 💡on how to dive deeper into my comparison via looking at the terms that were EXCLUDED from each data source. Here is a link to Conrad’s thread.
This test uses the same data set as my previous post above.
Just as a quick reference:
- I looked at pages that converted in Q4 2019 for a B2B SaaS client (calendar year 2019)
- I gathered all the terms that drove at least one click to those converting pages from Google Search Console (2,187 terms)
- I looked to see how many of these terms WERE NOT in the data sets of AHREFs, SEMrush & Keywordtool.io (“no volume” search terms)
I was going to include Moz’s Keyword Explorer in this test, but the 2,187 keywords were more than I was allowed in a list for a demo account ($179/mo tier). Have I mentioned I wish this tool was easier to get access to?
Big Yikes and Fun Surprises
Let’s just dive right into the data.
Can you guess the big yikes and the fun surprises?
Let’s break down the numbers:
SEMrush was missing approximately 67% of keywords that drove traffic to a page that converted in Q4!!!
On the flip side, Keywordtool.io only missed 22 keywords, which is an astonishing 1%.
Meanwhile, AHREFs only missed 102 terms for an incredibly respectable ~5%.
But What Does This Mean?
Well, first of all, it means that research is good because sometimes it can change your mind. For example, after my last piece, I was considering cancelling our Keywordtool.io subscription as we use it almost exclusively for search volume and less than the other tools. Def won’t be doing that now.
Second, we use SEMrush a lot for keyword discovery right now. Next week we are incorporating AHREFs’ API into our discovery pipeline 🙃
Third, I have a lot of ideas about how to better approach 3rd party keyword data. Expect more posts around this topic from me in the future.