It has long been thought of that click-through-rate plays a part in Google’s organic ranking algorithm. Rich snippets were great for increasing SERP CTR, but with most rich snippets now gone (thanks to sneaky, manipulative SEOs misusing them) SEOs have been exploring other options.
Rand Fishkin from Moz has run a few tests to see how SERP CTR influences organic rankings. In these tests, Rand tweeted his followers and asked them to perform a specific Google search and told them which organic result to click on. Within a couple of hours the site he asked followers to click on had been boosted to position 1!
While its not possible to say with 100% certainty that the cause of the ranking increases were a result of the increased CTR for the specific search, it does give us a very strong indication that CTR played a significant part in these results.
So, it’s all well and good if you can achieve these results legitimately with real people, but can you achieve the same results by using bots? Or is Google smarter than that?
This post looks at a 6 week test I did on whether click bots can improve your organic performance. I won’t disclose the site I was testing the CTR bot on (yet) because I don’t need any external influences on it and I will be using this site (along with a few others) in a more in-depth follow-up test. The site used and exact processes will be revealed after the next test so you can investigate.
How Do Click Bots Work?
There are many click bots out there that are ready made and ready to use, or you can get someone on Freelancer to make you one with all the features you want. If you want to do a small test there are even a few Fiverr gigs where you can get someone to use their bot and perform searches for you.
The bot I used for this test does its best to mimic a human user – which is what you want, if you are going to use one!
Due to so many people asking for it, the name of the click bot I’m using is Serp-bot. I am not affiliated with Serp-bot in any way.
The main features of a good click bot (if there is such a thing) are:
- It runs through a VPN and changes IP after each search. You can select which countries you want the searches to come from.
- It changes user agents between searches to make it look like the searches are using different browsers, devices and screen resolutions.
- It clears cache files between searches to leave no trace of the previous search.
- When it performs a search you can set it to click on competitor sites and immediately bounce before clicking on your site.
- You can set it to spend a certain amount of time on your site and browse through certain pages. It can also scroll up and down pages – just as a real person would.
- When it’s finished browsing your site it closes the browser – ie. it doesn’t go back to the SERPs.
- It waits a specified amount of time before beginning the next search.
So, as you can see, it is very good at imitating human search behaviour and covering its tracks! But does it actually work?
My Results from using a Click Bot
The site I was using for this test had been sitting idle for about 6 months. Main keyword rankings were floating around between positions 30-50+ in Google. It was fairly stagnant. I’m confident in saying that the results of this test are largely a result of the clickbot activity, although, there was a Google “Quality Update” two days into the test, as seen on Algoroo:
However, I don’t believe this update had any noticeable impact on this particular site.
How much traffic did I send through the click bot?
**Updated August 15**
Two days after I published this post, this article by Bartosz Góralewicz came out claiming that CTR is not a ranking factor.
Our tests were very different and I have no doubt that if I tried the same method that Bartosz used I would get the same results he had. Bartosz sent a massive, unnatural volume of organic clicks through to the test site and saw negative results. I think the huge volume of clicks is very easy for Google to identify as someone trying to manipulate rankings, somehow.
When setting up my clickbot, I checked average monthly search volumes for each keyword and sent only a small percentage of clicks per day for each keyword. To me, this is a more sensible and realistic approach to take when testing the impact of CTR, especially when the target pages are ranking on pages 3-5 of the SERPs.
Here is how I calculated exactly how much bot traffic I would send per day:
(Average monthly search volume /10) / 30
This works out to be 0.34% of average monthly search volume per day.
Sending 10% of average search volume to an organic result on page 3-5 of the SERPs is still very high, but it is a lot more natural than if you were to increase average monthly search volume for a keyword by 3781%, as was the case in the Bartosz Góralewicz experiment!
I’m not saying my results are definitive, because they definitely are not! But Bartosz’s results certainly aren’t definitive either and there are flaws with both tests. The biggest flaw being we are using bot traffic instead of real people, like Rand used.
Okay, okay, just show me the results!
I was tracking 30 keywords for click bot test. The ranking distributions graph for the 6 week period can be seen below:
Changes began almost immediately for specific keywords that were targeted early on and it seemed to have a flow-on effect to other pages on the site at about the 2 week mark.
Overall keyword ranking improvements can be seen below:
How did organic traffic perform?
I know exactly how many organic visits per day came from the bot, so I can subtract these visits from the total organic sessions being reported in Analytics to see the real organic traffic increase.
Comparing the first week of the test to the last week of the test, organic traffic increased just over 150%.
With no other activity going on in this time, I’ve seen enough proof in this test for me to perform another test. The follow-up test will have much stricter processes in place and I’ll be able to improve my reporting on the results.
Can Google Penalize This Type of Activity???
Being fairly confident that this works, it also raises a lot of questions.
Will SERP CTR manipulation one day be viewed on the same level as spammy backlinking and spun content?
It’s certainly a lot harder to detect and even harder to prove that this is happening, and how would Google penalize such activity? A timed manual penalty?
What if it was a competitor that artificially manipulated your SERP CTR? Would this become the new standard way to negative SEO?
Would Google discount CTR in their algorithm as a way of fighting the spammers?
If SERP CTR remains a part of Google’s algorithm you can be guaranteed that black hat SEOs will continue to look to manipulate it and give Google no option but to respond. How they respond will be interesting to see.
The Main Takeaway from this Post?
Your main takeaway from this post should not be that you should all go out and buy click bots and forget about good SEO! (although they are fun to play around with…)
Your main takeaway should be that CTR is important and you should do everything you can to improve it, naturally.
How can you increase CTR without a click bot, you ask? Here are 3 ways:
- Make sure your title tags are well written to entice clicks
- Make sure your meta descriptions are well written to entice clicks
- Take advantage of any possible rich snippets markup available to you. Images or star ratings in SERPs are great for increasing CTR.
I’m interested to hear other people’s opinions on this subject. Leave a comment below or reach out on social media!