Discussion
Would you recommend this product?
4 Reviews
Maker
🎉🎉🎉 Thanks for checking out A/B Rankings, a split testing tool for SEO!
I built this tool after years of making changes to my own sites without ever being sure of the true impact. I'd make a bunch of tweaks, see improvements but was never sure I was spending my time in the most efficient way.
Now, instead of blindly making 'best practice' improvements you can see which strategies really move the needle. By using your Search Console data we can accurately track the change in clicks, impressions and positions for a page and tell you if your test is statistically significant.
You can create a 'simple' title and description test, to try and optimize your site for higher click through rates and greater long tail keyword coverage, or create an advanced test using CSS Selectors to add, remove or change literally anything on your pages.
Coming soon is a point and click editor for creating tests on your own website with no knowlege of selectors needed!
I hope you're excited as I am by the possibilities and would love to hear your feedback on my little project!
Thanks
Jonny
Upvote (7)Share
@jonnyplatt, This looks great.
A random thought if you have started with B2B sales :
a) Target companies who are hiring for SEO, digital marketing. If a company is hiring for this role, it means they are investing money in SEO that is why hiring for that role. They would be happy to use product like yours.
b) Target companies who have recently raised funding. Since they have money in their accounts now, they would be willing to invest money on organic growth.
P.S. Congrats for your launch.
Don't want to be a downy dotty here. In theory, this is awesome but there is no way in the world this can give you accurate results.
@pbriod was super polite with his comment. I will be a little blunter.
This can't work. Anybody that understands how SEO works, knows that this can't work.
good luck jonny
Maker
@utam_sharma Thanks Utam :)
Hi Jonny, congrats on launching!
How does this actually work? Traditional A/B tests run two simultaneous versions to see which does best. How does your software handle this?
Also, it looks like the site changes are done via your Javascript widget. Could this be a problem for search engines?
Maker
@beedesignllc Hey Brandon, good questions :)
You're right - strictly speaking it is not pure A/B testing as you can only have one version of the page indexed at one time. Instead, for a single page test I run a 'paired t test' to calculate whether the test has a statistically significant impact. Alternatively you can test groups of pages to compare your candidates' results against a control group.
Google now renders Javascript and so the changes you make using A/B Rankings will be reflected in Google's index. However this can sometimes take an extra day or two after the page is crawled.
I've also created plugins for Wordpress and Drupal, with more integrations to come, so that your site conducts the changes for a test before the content is even sent to the user. This avoids any rendering delay as Google sees the test right away.
Does that clear things up? Cheers :)
Hiring
A/B testing SEO improvements is a nice value proposition.
However, analyzing the situation before and after the change of a meta-title won't lead accurate results, as other factors could impact the CTR and rankings.
I think the tool would be more useful if it allowed websites with large number of similar pages to make changes to half of them an monitor the impact between the two groups with information fetched from the Search Console. This is something we do manually at Monito and having a tool to streamline the process would be very valuable.
Maker
@pbriod Hey Pascal,
We do have the ability to monitor groups of pages and only subject a fixed percentage to a test, leaving you a control group to track changes against.
It's a little more involved than creating a single page test and I haven't written the documentation yet, but I'd be happy to walk you through if it helps?
Thanks :)
Hiring
@jonnyplatt would love to discuss this more yes, you can write me an email at pascal@monito.com