Analyzing Results from Recent Testing
Nope. Ok. Great blog post everyone! Time for high-fives and a few cold ones!
Alright, I should probably go into more information on why we bring this up, what we did and what came out of it. Also, just a little peek into the future, be on the lookout for more posts like this as Thruline ramps up our testing. To continually better serve our clients we are running tests on different ideas for SEO and PPC to identify how things can impact performance allowing us to be even further on the cutting edge.
Adding Schema markup to a PPC lander or mini-site will improve quality scores and enhance the listings.
To understand why we came up with this hypothesis you need to know a little about what Schema markup is and what it does. If you’re unfamiliar with it, Schema markup is code added behind the scenes to a website that gives additional context to search engines about what a website, webpage and organization is about. This code was standardized by all of the major search engines several years ago. They use this information to help supplement the information they are able to understand about a site through their crawls. The way it works is that you add the code to the site with all of the information you want and the search engines can compare that to what they already know about the site to reinforce that understanding. For example, you could add some code that says “Hey, we’re a college/university and we’re located here”. The search engines would look at it, compare it to what’s on your site and what they know about it, and go “Yep, that’s you. You’re a college/university and that’s where you’re located”. If you do it right, the search engines can use this information to give you a little boost in authority and even enhance your listing in the SERPs such as adding breadcrumb navigation to your listing.
With this in mind, we decided to see if code could be something to improve paid listings. We knew it was a bit of a longshot that it could be used to enhance the listings themselves since a lot of that is determined by the setup of the ad, but we thought it was worth trying out. Where we thought there might be a better chance of Schema doing something was with quality scores. Since we knew that search engines looked at this code and used it to improve their understanding of a site, we wondered if the same could be true for paid since quality scores are partially driven by the relevance of the site to the ad being presented. Would reinforcing the relevance through Schema markup help at all?
It’s basically impossible to run a true A/B test on something like this since it requires the crawling of a page by search engines and you can’t have them crawl two versions of the same page. To overcome this issue we decided to select a couple of different pages from a client’s PPC site that was well established and also a few from another client’s site that was fairly new. We placed some Schema code on those pages. This allowed us to compare the quality scores against similar pages and also see if the quality scores were impacted more on a site that hadn’t yet built up much authority. The code that we placed on each site included information about the location and details of the organization. It also included information on the specific programs that the page was about.
We took baseline quality score measurements along with a few other metrics for the sites prior to placing the code. After putting the code in place, we ran business as usual for over a month. We wanted to be sure the pages had enough time to be crawled and indexed with the code on them and to ensure that if Google was going to recalculate the scores, we would leave time for that to happen.
Once we were sure we had given our test enough time to provide the data we wanted, we went back and compared the quality scores as well as the average CPC and average position for all of the pages we marked up against where they were prior to adding the code and also compared them to the rest of the site.
We also compared how the ads showed in the SERPs and saw no difference with the listing.
Although we didn’t see any improvements to the metrics that we measured, the test wasn’t a complete bust. We took a look at the organic traffic going to these pages and saw between a 10% and 20% increase in organic traffic. We can’t completely rule out seasonality shifts with these increases, but it does seem to indicate that even on a PPC site, adding Schema markup can have a positive impact on organic traffic. So, as mentioned at the beginning, adding schema to your PPC site does not appear to have any effect on quality scores or the listings. But, if you’re objective is to increase organic traffic, Schema markup is something to consider.
Allen Harkleroad, Manager of Inbound Marketing