Every SEO knows that the meta description is not a ranking factor. It’s something you learn on day 1 of your SEO career. Some people might qualify it and say “the meta description is not a direct ranking factor but it does impact CTR so it might be an indirect factor.” Then comes along a powerful correlational software and things like the meta description begin to correlate, consistently, to rank.
This development caused us to look at how SVTs were being done. Was it possible that we were missing very faint signals in the way that our tests were set up? We test this out, now, taking away some of the stronger signals we have. Will it pop up? Let’s check that out.
Test Downloads
Transcription Below
This one is by Ted Kubaitis. I saw Ted speak in November in Changmai and he showed some pretty amazing correlational analysis that he has done where the meta description correlates to rank. He also showed that he was following the serps and there was this one particular site that was moving up three moving down three and nothing was changing on that page except for the metadescription.
He has talked about these things in length and there are places where Ted talks about these things and you can see him talk about it and its truly great. But this gets into what we were talking about earlier, are there factors that are too much noise? In a test environment, when we use an h1 and a meta title in the body and the URL, the keyword in all these places, are we creating too much noise so that very faint factors cannot be heard.
I really like this test, as well, because every SEO knows that the meta description isn’t a ranking factor. I think you learn it on day 1 of your SEO career. This is just one of those things. The metadescription isn’t a factor, everybody knows it, they might tell you that it is a secondary factor because it goes to people giving you the clicks, so it is clickthrough rate. I have said this to clients.
These are words that actually came out of my mouth. And then Ted comes along with his software and his research and he is – not so fast, I am seeing this in the serps and it seems to be a factor. At the same time that that is going on and I was thinking about the noise issues in our own test, let’s take that out.
Let’s strip that out, the URL, metatitle, h1, and we will just put in the body so that the page will index for that particular keyword and then, let us see if we can find these extremely faint signals but the things actually do correlate to rank. That is what we did here.
In this set up, we got the target keyword one time in the body but we don’t have it in those other strong places, those places like the group A place – the strongest places to put a keyword. We kind of removed those.
Then, we added a meta description to all of these. So none of these pages had a metadescription, now they all get a metadescription and in the 1, 2, 4, and 5, it’s just a regular lorem ipsum. In the number three, it is a lorem ipsum but it also contains the target keyword. So basically the target keyword exists twice on the page and one on all the others, is the idea.
Our number three page moved up to number two. We have done this multiple times and it has worked every time. It has always moved up. I haven’t done this test in the inverse, I haven’t done it the other way and we know we really need that for confirmation and that is why we will redo the metadescription on the other ones and this one does not have metadescription in the number three.
So, we haven’t done that yet, we can’t say for 100% certain that yes, it is a ranking factor, but if you look at the correlational data that is coming out from Ted and his research, and what is crazy, you can replicate that. You can use that tool or other correlational tools and you can replicate this and see that metadescription is correlating to rank.
And then, we can see this now in a single variable test replicated a few times, it is a very exciting thing that one of the standards of SEO, the metadescription isn’t a ranking factor, might not be true at all. We just want to look at it properly. This isn’t the be all and end all factor. This isn’t like, I put my keyword in the meta description and I moved to number one but, it is possible that it is calculated within the algorithm as something and that it is easy to do and something you can do within a handful of minutes.
Probably something that you should do anyway because you want to get that click but now, we can say that, no – other SEOs will tell you that it is not a ranking factor but that is just because they haven’t tested it. Other SEOs will tell you this because they read MOZ, all these other blogs, and they haven’t actually taken the time to do the research for themselves, and what is actually cool is it is actually a ranking factor and I know that because we tested it.
We are really getting close to saying that it is 100% sure. It is only a good thing to do, to get your keyword in the meta title. And if you have a script or plugin that is pulling it into your open graph or twitter cards which is usually how that works, the open graph description, they are just pulling from your metadescription, that is a ranking factor that we have found, so you should do that anyway, for that particular signal.
Andrew: I think most SEOs would probably be doing that anyway, putting the keyword in the meta description on their pages, but they weren’t doing it for a direct ranking benefit, they were doing it for the secondary ones and to get the bolds.
I also think there are those SEOs that follow the things that say don’t write these things for SEO so they also intentionally don’t include them – I’m not stuffing my keyword, I am writing for click, etc.