How to hold a variable constant for an A/B test
-
For example, let's say you want to A/B test a title tag change. You are hoping to identify whether a title tag change increases CTR. But, position is always fluctuating a bit and that affects CTR, too. So, I'm interested in how you could hold position constant in order to isolate the change in CTR that is due to the title tag change. Does anyone know of resources/tools/tutorials for how to do this?
It's been... a very long time since I took statistics (-: I have access to Excel, MS Access, and R studio.
-
Thanks for the reply. I'll check out the links you included.
I thought it might be helpful to show an example of what got me thinking along these lines. This is a study about A/B testing of title tag changes. They don't say they accounted for position fluctuation; maybe they didn't, but it seems like you'd have to in order to have meaningful results.
-
I don't see a way of doing this accurately. With organic rank fluctuations, seasonality, competition it's near impossible to "hold" a position. However, if you have a position in serps for a specific keyword with little fluctuation and get decent traffic you can establish this as your baseline. Once the title is updated and appears in the serps, begin monitoring w/search console, analytics or etc. and see if you see any decent jumps. Some links for additional information: Finding the ROI of Title tag changes using Google's CausalImpact R package, SEO Split Testing & A Beginner’s Guide to A/B Testing: Effective SEO Landing Pages. Would love to hear if anyone else has a better way.
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
For a sitemap.html page, does the URL slug have to be /sitemap?
Also, do you have to have anchors in your sitemap.html? or are naked URLs that link okay?
Intermediate & Advanced SEO | | imjonny1230 -
How do we avoid duplicate/thin content on +150,000 product pages?
Hey guys! We got a rather large product range (books) on our eCommerce site (+150,000 titles). We get book descriptions as meta data from our publishers, which we display on the product pages. This obviously is not unique, as many other sites display the same piece of description of the book. It is important for us to rank on those book titles, so my question to You is: How would you go about it? I mean, it seems like a rather unrealistic task to paraphrase +150,000 (and growing) book descriptions. As I see it, there are these options: 1. Don't display the descriptions on the product pages (however then those pages will get even thinner!)
Intermediate & Advanced SEO | | Jacob_Holm
2. Display the (duplicate) descriptions, but put no-index on those product pages in order not to punish the rest of the site (not really an option, though).
3. Hire student workers to produce unique product descriptions for all 150,000 products (seems like a huge and expensive task) But how would You solve such a challenge?
Thanks a lot! Cheers, Tommy.0 -
Links / Metadata around Recent Posts etc in Wordpress / Blog - Good SEO Practice?
Hello In a Wordpress blog ( or part of an ecommerce site that runs under wordpress ) it is good to show recent posts in the sidebar on most pages. Obviously the posts aren't going to be relevant to every post , so my questions are: Is having these on the page hurting SEO for the page? Is there good metadata structure to put in there? ( like rel="nofollow" or similar ) Thoughts?
Intermediate & Advanced SEO | | s_EOgi_Bear
Thanks for your time
Marty0 -
Is it worth removing date from Blog Posts / Articles
Wondering, is it worth to remove date from articles from seo perspective. Am sure, Google search algorithm would like demote a post written a year back, as against an article on the same post (unless a year old post has very strong Authoritative links) May be it can turn out a bad user experience of removing dates, but if can hide date using Javascripts so as to show it as image to user and hide it from search engines, is it a good idea !!
Intermediate & Advanced SEO | | Modi0 -
Links from music/celebrity based fansites - sitewide images with no alt text
We're currently in the middle of a link audit on our website OneDirection.net and a large part of our incoming links come from fansites such as the following: ladygaganow.net nickjonline.com justinbieberhood.com joejonashq.com harrystylesfan.org brunodaily.org onedirectiondaily.com onedirectionfans.net Now, our previous way of thinking was that these are very relevant websites in the same niche as us, and therefore should be passing some value? However all of the links on these sites come from sitewide images with no alt-text. Some of the sites are passing 1000+ links to us. We've been wary to disavow or request removal of these links as we've usually gone with the thinking that Google applies "common-sense" based logic in its algorithms, and therefore these backlinks should be ok - in our opinion. However we think we are suffering from some kind of algorithmic penalty with our current rankings, and are now thinking these could be the cause. What are people's opinions on these links? Should we stay clear of sitewide links altogether? Should we contact the site owners and try to get them to mix up the alt-text? Or should we get rid of them altogether? Thanks, Chris.
Intermediate & Advanced SEO | | PixelKicks0 -
Regular Expression / Wildcard Redirect Situation
I am dealing with an interesting situation. Here's what's going on: Current URLs Example1:
Intermediate & Advanced SEO | | NakulGoyal
www.domain.com/red-widgets-cid-1234.html
www.domain.com/red-widgets-cid-1234-1.html
www.domain.com/red-widgets-cid-1234-1-1.html Canonical on All Above URLs:
www.domain.com/red-widgets-cid-1234.html New URL:
www.domain.com/red-widgets-cid-4567.html Current URLs Example2:
www.domain.com/red-widgets-cid-1234+10.html
www.domain.com/red-widgets-cid-1234+10-1.html
www.domain.com/red-widgets-cid-1234+10-1-1.html Canonical on All Above URLs:
www.domain.com/red-widgets-cid-1234+10.html New URL:
www.domain.com/red-widgets-cid-6789.html I want to make sure all variations of the above URL redirect to the new url. What wildcard 301 redirect / regular expression can I use to tackle these ?0 -
URL Structure - Keywords vs. Information Architecture/Navigation
I'm creating the URL structure for an ecommerce site and was wondering if it's better to structure my URLs according to the most popular way people word their key phrases or by what makes most sense from a navigation perspective. Let's say I'm selling clothing (I'm not, just an example). I want the site to be open enough so a user can navigate by Person Type (Men's, Women's, Children's), Clothing Type (Shoes, Shirts, Hats), and Brands (Nike, Reebok, adidas). My gut and past experience say to structure the URLs from the least specific to the most specific: mysite.com/mens/shoes/nike But I know "men's Nike shoes" is searched for more than "men's shoes Nike", which would render this URL: mysite.com/mens/nike/shoes I know mysite.com/mens-nike-shoes would be best, but the folders setup is what I have to work with. So which is best for SEO? URLs that play to the structure of the most searched for key phrases? Or URLs that follow the information architecture/navigation of a site? Nate
Intermediate & Advanced SEO | | rball10 -
Question about 301 redirect for trailing / ?
I am cleaning up a fairly large site. Some pages have a trailing slash on the end some don't. Some of the existing backlinks built used a trailing slash in the url and some didn't. We aren't concerned with picking a particular one but just want to get one set and stick to it from now on. I am wondering, would I clean this up within the same redirect in the htaccess file that takes care of the www and non www? example RewriteEngine On
Intermediate & Advanced SEO | | PEnterprises
RewriteBase /
RewriteCond %{HTTP_HOST} ^www.domain.com/ [NC]
RewriteRule ^(.*)$ http://domain.com$1 [L,R=301] I currently use that to redirect the www. to the non www as you can see. However here is what I was confused about. Would this code be enough to redirect ALL pages with a / to the ones without? or would I also need to add another code (so there is 2) to my htaccess like below? RewriteEngine On
RewriteBase /
RewriteCond %{HTTP_HOST} ^domain.com/ [NC]
RewriteRule ^(.*)$ http://domain.com$1 [L,R=301] RewriteEngine On
RewriteBase /
RewriteCond %{HTTP_HOST} ^www.domain.com/ [NC]
RewriteRule ^(.*)$ http://domain.com$1 [L,R=301] That way, now, even the non www pages with a trailing slash will redirect to the non www without the trailing slash. Hopefully you understand what I am getting at. I just want to redirect EVERYTHING to the non www WITHOUT a / Thank you Jake0