Difficulty Level: 2/5
What we are going to be looking for today are some hints of what articles are doing best in the SERPs. So we can focus on what stands a good chance at ranking in google, or if we need to change our keyword strategy. I’ve never been a huge fan of looking at google analytics in the last few years due to not trusting the hand that feeds. The relationship between webmasters and Google is pretty rocky at times. The problem Google need us, but they don’t care all that much about us sadly. They are in the business of making money; they only profit when someone clicks on ad spot in the SERP results or with AdSense ads on publishers account. Of course, with time, affiliates messed up everything with being overly aggressive and shady. Not that I have a problem with being aggressive and dubious, but when it starts to affect the community…. Even now I honestly am even scared to hit the search console report often in analytics… Just the idea of Google thinking I’m trying to manipulate the search engine rankings scares me. Piwik is what I normally use for my analytics, so it’s privately collecting this data outside of google eyes.
Regardless, we can still use analytics to our benefit. We can pull the data they are showing us, and use SERPLab to confirm what they are saying. I started writing this article because I saw it showing I was ranked for MOZ API, and buy EDU Backlinks. Keywords I knew I was hardly optimized for, and with this site being new. There is 0 chance that I can be that highly ranked for it, and that’s where everything gets a little fuzzy as you will see below.
If you have connected search console to google analytics, then you can pull a report of the quires your website is showing up for. It takes a few days for the data to show up, so if you justed connect the two. Then you will have to wait a bit.
This screenshot is the most recent export of the keyword impressions, pulled on the same day the SERPlab screenshots were taken. I let the test run for about seven days to let the rankings mellow out a bit. They are bouncing around, but with enough data, you can spot some trends happening.
We are going to be using SERPlab again for this case study. If you dont have one already, go ahead and create a free account with them. You can perform 150 searchs a day for free.
On first glance, the data does look like we have a large mix of ranking positions. It’s what I first thought when I saw it, but if we export some reports. The data will show something different.
If you click on the campaign to bring up the full report. Export Data CSV – Brief Export. Assuming you used the same keyword list for the campaigns you are split testing against. The brief export will organize the keywords in the same order as you inputted them. So you can just copy and paste the other campaigns rankings into one CSV to compare them.
Out of the 60 keywords, SERPlab found the same 12 ranking on desktop, mobile, and desktop.
If you look at the ranking positions of each keyword, they are all somewhat similar.
So the data that matches from googles search console reflects what we see in real life. I don’t believe search console or SERP lab is sending us bad data with the keywords that are not showing up in the SERPs. What it very well could be is geo-located impressions, variables that we can’t tweak to test. Lets work with what we have though and see if we can find any hints of what’s happening here.
Its hard to say right now, not a lot of information is being shared where all the data really gets pulled from. Someone suggested it may be city or state geo searches causing the missing keywords in the SERPLab. Other possibilities are img search for these keywords. It is possible SERPlab is giving is a false report, the local search is still in beta. But let us assume its tracking correctly.
This is the first initial testing of the INTL searches. We are going to leave this campaign running for a week to allow some more data to be collected. This is the first hint of where some of the other impressions are coming from, though.
If we start searching for positions in different countries, we start to see are a wider mix of positions and unique keywords only ranking in that geography. Such as Query deserves freshness, so far only is showing up in Canada. So what’s likely happening is Google is sending us a huge mix of data from multiple sources. INTL impressions, local Impressions, IMG Results, Ect Ect etc.
What we can focus on are the results we have found that are ranking.
Keyword: gscraper footprints
About 5,670 results
Keyword: expired Tumblr
About 764,000 results
Keyword: Tumblr expired domains
About 34,900 results
Keyword: scrapebox long tail
About 12,600 results
Keyword: WordPress failure notice
About 373,000 results
Keyword: gscraper tutorial
About 19,500 results
I have not optimized my articles for any keywords, so it is interesting to see what Google thinks with these first early rankings. What it’s looking like that of course, little results = a good chance of getting onto page one. It would not take much to push these pages onto page one; they may even bump up on their own because there are so few results for these queries. I would focus on looking for more super low result terms like this to write helpful article. Even without any more keyword research, it’s also pretty clear writing about scrapebox and gscraper should pull in some trickles of traffic pretty easily. Ranking on a brand new site, with 0 backlinks built to it, and already showing up for some keywords. Though; Try to remember, don’t write the entire article based on one long tail keyword. Topic research will get you much farther in life with SEO.