For this month’s webinar, we discuss the tests from the February Dossier. We discuss Part 2 of the TFIDF test checking if a longer H1 would provide a boost. He also discusses the results on the Meta tile and Meta description tests and provides practical applications on how we can take advantage of the results.
The Intro Pop, so as the Font color test was discussed and lastly, he discusses the last test of the dossier - does the number of h2s on a page affect rank - he further discusses how other series of tests are being done to test zones further and how our group can have a massive edge on our competition from these kinds of tests.
Tests this month
Test 170: Will a Longer H1 Zone Positively Affect Rank?
Test 171: Does Google Read Long Meta Titles?
Test 172: Will Google Display Meta Descriptions that are too Long?
Test 173: Do Intro Pops Affect Rank? Part 1: Just the Intro Pop
Test 174: Do Intro Pops Affect Rank? Part 2: Keywords in the Intro Pop
Test 175: Does Font Color Affect Rank?
Test 176: Does the Number of H2s on a Page Affect Rank?
February 2018 Dossier
Transcription Below:
Will a Longer H1 Zone Positively Affect Rank?
This is a follow up to a test that we did in January where we did a test just looking at having a target keyword as the H1, so a shorter H1 zone. It is an old school concept from TFIDF where importance is designated by shorter zones. In this case, the H1 would just be target keyword and we wanted to see if that would provide a ranking boost and it did not. Then, I wanted to ch
eck the other way, it makes more sense that a longer H1 would provide more context. Wanted to see if a longer H1 zone would provide a ranking boost.
In this test, 5 pages were created in the normal way. The test page, the number 3 ranking page has a longer H1 – 5 or 6 word H1 containing the target keyword and the control pages 1, 2, 4, and 5, all just had one word H1s with the idea that if the longer H1 provides ranking boost, it would move up. It it is a negative, it would drop down. No movement is what we got. The test was run three times.
Test one and two, there was no movement.
In test three, it did drop to the bottom of the serps, I think it is just an anomaly. Looking at this test being run six times when you also count the other one, no movement here. I don’t think zone length, how long your H1 is, is going to make any ranking impact. Obviously, the words that you use inside your H1 certainly can impact rank but just the size of it alone isn’t going to be a benefit one way or another.
Does Google Read Long Meta Titles?
Basically what we are looking at is, when you do a SEO audit, one of the things that come out in the audit is too long met titles. It is a section in every SEO tool where it tells you if you have meta titles that are “too long”. What they are talking about are meta titles that are probably either more than 60 or 70 characters, depending on the tool you are using. The reason they call them too long is that Google will clip the results in the serps.
Your meta title will have an ellipse in the end (…) and it won’t display the rest of your title. So, with tools recommending that you need to change that, the question in my mind is, is that because Google ignores everything beyond the ellipses. In this test, we see if Google ignores it or reads it. We put a target keyword beyond the clipping, beyond the (…). In the serps, you cannot see the keyword. It does not exist anywhere else on the page other than in the meta title and beyond the area that is shown.
If the page indexes for that keyword then we know that Google is reading beyond that point. That is exactly what we got. Google is reading and factoring in things that are clipped in the meta title.
I think this has really cool implications. Let us say that there is a misspelling that you would want to rank for but you do not really want that seen on the page. That is something that you could put on the (…). Let’s say there is something that is off-brand. Sometimes, you will have a keyword that you would want to rank for where people in the industry call it one thing an that is the correct thing to call it but people who are your potential clients might call it something else because they are not educated in that industry.
You could put that term that you really do not want in your copy because it is not really the right term, you could put that in the (…) so that it doesn’t interfere with the copy or usage of the word but you could potentially rank for it. Something else that I can think of here as well is if you have a page where you must target multiple keyword locations. It is a one service page but you are targeting two places.
You could put those multiple places beyond the (…) and you actually get those cities, locales, neighborhoods, etc. Still in the meta title but you could have the title that is more service oriented or product oriented where you do not want people to be necessarily turned off because they are in one city and not the other but you have to use the same page for both, that would be a great way to do it as well.
The other thing where I want this concept is in those SEO audits, why do you need to change your too long meta titles, a much better way to go about things is to see if you are actually getting harmed in your click through rate. You can do that in Search Console. Pull up the page and look at the keywords you are ranking for, do a little click through rate analysis and see if it is harming you in the first place. If it is harming you, that is an opportunity to change your meta title whether it is too long or not.
I think that is a better way to look at whether or not to change your meta titles not simply because they are too long and some tool says that you should shorten them. You should actually do a click through analysis to see if you are being harmed by that. You should probably be doing click through analysis anyway and that is something that we talk about a lot in the Tuesday Chats. That is a better decision making tool rather than an SEO audit that says your meta titles are too long and you need to shorten them simply for the sake of that.
Will Google Display Meta Descriptions that are Too Long?
Adding into the too long meta titles, we went in to the too long meta descriptions. This is something that I have known a little bit but I wanted to see how the serps would react. You can write a meta description that is too long and it can target a different keyword or perhaps a variation of your keyword on the page and Google will actually pull from the “hidden” meta description area.
Say for example, in this situation you have two locations. You can write one meta description for location one and then you could write an additional meta description for location two and when somebody searches for the location with your service or product, Google will pull the proper meta description, the one that is most relevant to that keyword.
You could do that for say, you have different buyer personas on your page where people would be coming to that page and clicking or using that product or service for different reasons, you can write different meta descriptions for each one of those personas especially if they were going to search for that term a little bit differently. In that situation, you could then write a description for each of those personas in a way that they are searching for that target keyword. I wanted to see how that would respond what that too long meta title.
What I did was I took a too long meta description and again used Yoast and in the area that is too long, I put in the same keyword that was used for the meta title and ran the search again. You can see our keyword and it is still clipped in the meta title but the proper meta description shows. This is pretty cool. It is the same thing that I have discussed with what you could do with the meta title. You could then implement things in this meta description.
When you have a page that you are spending a lot of time, money, and effort to optimize, it only makes sense to do something like this in order to capture those clicks that you want, by writing a meta description what is most fitting for particular keywords on a page. You might have a top level keyword that you are optimizing for and then in your H2s or H3s or how you have broken up the page, you might have variations on the keyword or you might find something that is very similar or different ways that you are trying to capture the secondary terms.
For the most important ones, you can write multiple meta descriptions so that when somebody is looking for that secondary keyword, you can provide an on point meta description. When you are in a situation where there is something a little off brand, you can still hide that keyword in the meta title and get a description that you want.
Do Intro Pops Affect Rank? Part 1: Just the Intro Pop
Last year in January, Google did the Intrusive Interstitials Penalty that was a “penalty” aimed at popups that were taking up too much screen, particularly in mobile. That is when you come to a page and instantly there is a pop up and that pop up covers the vast majority of the screen. Google says they hate that because it is a poor user experience and they want those to go away.
Additionally, just a couple of weeks ago, Google announced that the Chrome Browser has an ad blocker and it specifically targets ads that are falling outside of their ad policy. One of the ads they talk about that they want to block are pop ups. So clearly, Google does not like pop ups especially intro pops and those that are taking up a good chunk of the page.
If they are going that far and they are saying that they have this particular mobile penalty and then they have this Chrome Ad Blocker, the question in my mind is, are they going to affect rank. If you have an intro pop, will your ranking drop?
So, adding an intro pop to our number three page, no intro pop added to the 1, 2, 4, and 5. The into pop is just very basic, it has some text to it and it does not contain the target keyword or links or anything else. Just the pop up itself. Hilariously, in all three tests, the test page moved up. In test one, it moved to number one, in test two and three, it moved to number two.
What I think we can say is that Google is not punishing you simply by having an intro pop on your page which might make a lot of you breathe easy as intro pops are an effective way to capture emails, get clicks, provide information, sales, that sort of thing. You are not going to be directly harmed by using it as my wheel spins towards ranking and getting edges, my thought was what if we put keywords on them? That was the next test.
Do Intro Pops Affect Rank? Part 2: Keywords in the Intro Pop
We put keywords in those intro pops. In this situation, the intro pop includes the keyword. Would that provide a ranking boost? In tests one and three, it rose to number one. In test two, it did drop to number five. I think that is an anomaly. We are looking at 6 tests running here, five out of six it moves to the top, I feel pretty good that an intro pop is not going to be a problem.
I do not necessarily think that you are going to get a real benefit from putting a keyword in there but it is probably not harmful and if you think about it, in two of those three, it moved to number one and whatnot. If it is something that makes sense for you to do. I wouldn’t hold myself back from doing it, I wouldn’t have any reservations from doing that – having a target phrase in that into pop.
Would I go out of my way to do it to get a ranking edge? No I won’t, I don’t think that is worthwhile. But would I be concerned of having a negative effect by having it? No. Just by having it or even if there is content that contains the target keyword, I think you are completely fine.
Does Font Color Affect Rank?
This concept comes from the patents and theories on link strength. A link is stronger or derives it strength from its propensity to be clicked. A link that is more likely to be clicked is a stronger link and passes more juice, is the concept. A link that is higher up on a page, for example, passes more juice. We have tested that and we know that. Something that we haven’t really been able to nail down necessarily is if a link has a larger font, if it is larger, can you make that link stronger? Can you do different colors instead of a standard blue?
For example, for a link as then that would make it stand out and thus make it more likely that somebody would click on it. That is the theory and if you listen to some of the old school guys, a lot of them subscribe to those theories that they would increase link strength. Something that we haven’t been able to nail down, I have tinkered with it from time to time but something that I wanted to check was just doing that to your keywords.
Could you affect your keywords by adding color so that they stand out? In a sense, it is kind of a way of playing with bold or strong or emphasis or italics if you will, but just looking at color. If I make the keywords stand out, does that give the keyword more importance. Will it provide a ranking edge?
In this situation, we couldn’t get any edge. Four new keywords were added to all pages. The number three page received the keyword in four different colors – orange, green, purple, blue. We were not able to get a ranking edge out of that. One of them dropped to the bottom but I do not think color is a negative. There is no movement in the other two.
These are the kind of things where I think once you have a base knowledge of how these things generally work – going into this I didn’t think it would provide any ranking benefit but you never know. It is the kind of thing where you are in a client meeting and these odd ball questions come up, I think these types of tests can give you a foundation so that you can confidently answer that no that is not how it works and this is why – we’ve run this test, etc.
In my mind, this is the kind of thing that this falls into. It would have been really cool though if it actually did provide a ranking benefit. Unfortunately, it didn’t but it was good to check and those were the results.
Does the Number of H2s on a Page Affect Rank?
I was talking with Daniel Cook, one of our members and we were talking about some on page theories and application and in that, he was asking about the quantity of signals or the quantity of elements on a page. What I mean by signal or what I call a zone, would be an H1 tag would be a signal or zone, an H2, H3, H4, image alt, bold, italic, those are all areas that Google is looking and crunching when it comes to the algorithm on onpage. Those are the different signals, zones.
The question came up, what about the quantity of zones, can we look at or should we look at how many H2s, for example, our competitors have on our pages? Should we be matching that when we are looking at comparative parity and just the quantity of the zones. We are not talking about using your keyword exactly in the H2, we are not talking about variations of your keyword, or LSI if you will on any of those zones, we are just talking about simply the number of zones that exists.
My thought was no, you do not need to do that because what you need to worry about is how many times you are using your keywords or variations of your keywords within that zone. Let’s say you have done your analysis and you need to use your target keyword three times in an H2.
Theoretically, you could just 3 in 1 H2, that is not a good writing and I would not recommend that but my thought process was that would satisfy what Google is looking for, that is what would make on page work. You probably want 2 or 3 H2s and space them out a bit but theoretically, you can put all three in one and you would be fine, was my thought.
Because guessing is not what we try to do here, I ran a test on this. The concept is we are putting unoptimized H2s on the page. Five unoptimized H2s go on page three and on the control pages, we are just putting one. So, there are no keywords in here, there is no LSI, variations, it is simply just the number of zones. Quite surprisingly, in all the tests, they all dropped. The page that has 5 H2s vs the pages that have one, the test page with the 5 H2s dropped every time.
This was very surprising and quite shocking to me. This is the results that we have at the moment, what I am doing currently is I am now being more mindful of this. I do not think my pages have really struggled because I try to write decent pages as it is and if you are outlining properly, you are going to get a reasonable amount of sub headers to match the amount of words you are trying to get on the page. I am more mindful and I am kind of counting those zones.
We are doing a lot more testing on this this month, we are looking at other zones, we are retesting the H2s because this is the kind of thing that I think might actually be a huge edge. If this bears out that if you have too many or too few of zones and that is an actual ranking indicator, this is a spot where our group can have a massive edge on our competition so that when it comes to doing our on page, we can look at, count, and measure how many elements we should have on that page.
How many H2s, H3s, H4s, how many image alt, images, how many should go on a page and actually count that as a ranking factor. We are doing a lot more testing this month because I am really excited about this to see if we could nail this down and see if we could find the limited specific zones. Perhaps this is just an anomaly, I do not think so. But we are going to see if we could nail this down because I think this is going to be huge.
In the case of H2s, should you have an H3 as opposed to just H2s?
That is a good question because it could be kind of semantics. I do not think so though because it is not unreasonable to have a page that would just have those sub headers and you should not necessarily have any points. I do not think that is completely outside the realm but that is something that we could definitely look into. Perhaps it is just more of a structure that these pages could have been fine if they have corresponding htags. We can look at that. I do not think that is where it is but quite possibly.
H2s being duplicate titles, that could drop it
That is possible though we have looked a lot at duplicate content and you really need a lot of content as duplicate in order to ping on the duplicate content filter. I do not think that is an issue but that is something that I could be mindful of in these tests to make sure that we do not have any duplicate H2s.