In this webinar to end the year 2017, we discuss the results of the tests in the final dossier for the year and some insights on the results.
Can PPC Traffic move the needle?
Can a shortener help in passing more juice?
Can Google listen to sound and does it know when bad words are being used?
For the last test of the year, a quasi real test environment is built to test out if a link can stack up against highly optimized pages. Great tests and discussion to end the year!
Tests this month
Test 158: Can PPC Traffic Increase Rankings?
Test 159: Does a Second Internal Link Housed in a Shortener Result in More Juice Being Passed to the Page?
Test 160: Does Google Limit the SERPs to One Exact Match Domain?
Test 161: Will a YouTube Video with Sound Rank Better than a Video without Sound?
Test 162: Do Bad Words Affect Rank?
Test 163: Can a Single Exact Match Link from a PA, DA1 Site Beat a Highly Optimized Page?
December 2017 Dossier
Transcription Below:
Can PPC Traffic Increase Rankings?
This comes from Derek Booth who spoke at Rockstars, he’s been on several chats, he is a PPC expert and one thing that we wanted to do is to look at traffic from PPC, traffic from Adwords – in particular. We have done a lot of traffic tests and we know that traffic is a ranking factor and the thought process is, why don’t we just buy it from Google.
We could do Microworkers, we could do bots, why not just buy it from Google. It would actually be real traffic and it would be people that might actually convert. Kind of serves two purposes. This test turned into two things and Derek has a really nice write up when you go through this.
Initially, he wrote an ad that is just all lorem ipsum and it got declined because they actually manually review the ad. That is actually cool just to confirm that they claim to do that and they did it. Then what he did was he actually made an ad for a lorem ipsum generated content. What was pretty funny about this was then, was this keyword has no competition whatsoever, low volume, when he put it in, Google’s suggested bid was 67 cents. He just put in 10 cents in the bid and Google won’t show the ad.
No one is bidding for this term, people from the group is searching so it is being searched but Google refuses to show it. So he moved up to the 67 cents and sure enough, the click came through at 54 cents which is a complete ripoff. Then, he got it back to 10 and the ads stop running again. He ended going up just for the sake of doing the test and got a bunch of clicks.
The average amount per click was 44 cents per click and you can see he got 16 clicks, 33 impressions, clickthough rate at 44 cents a click. That was interesting right there, a keyword that Google does not know anything about, no one is searching for it, Google sticks to its guns on the suggested bid which was pretty interesting.
Before it wants to figure anything out, before maybe it even knows what the inventory is, like how many people might actually be searching for this thing, it will go for its suggested bid and that is that. If you are in the PPC space, in the Adwords space, keep that in the back of you mind when you are setting up campaigns for something that is not terribly competitive, you might actually have to go for that suggested bid just to get that ad to show in the first place.
I would bet overtime that you could take it down once Google understands how much search volume there is for this particular term that it doesn’t know yet, I bet you could scale down after that but maybe that initial thing, in order for it to justify its time, it is going to go with the suggested bid.
We finally had clicks, not a lot, as you can see – 16 clicks and I think a few more clicks came in after that. What was interesting is that the test page did not move on that number of clicks. I would think with nothing else going on and with traffic, that that should do it. It kind of goes in with what we are seeing with traffic. It definitely works but there is something more to it.
In this case, the traffic was good. People went in and stayed on the page and they left. It was decent traffic, everything should have been fine about it but it didn’t just seem to move the page. I am not ready to say that PPC is not going to be a good option.
I think there are other elements that probably needs to come into play first. This might kind of go with the recipe concept of SEO where you have to do things at a certain order or the algorithm isn’t crunching things until it gets to a certain point. That is maybe what we are talking about in this situation. In this situation, the PPC traffic did not just move the page at all.
Dori: How long ago was this checked? I would be curious to know if maybe it might have changed.
It did not move at all and that is a good point and I know it hasn’t been checked in a week or so. That is a good point. In the traffic that we have seen, it is kind of just like links.
If you make an onpage change, you can see results in test environments sometimes within minutes – around 20 minutes. But in live environments, you can see results within 24 hours sometimes, even less. But with links, even in test environments we see that it could take 2-3 weeks sometimes. Traffic is the same kind of factor. It can take weeks before traffic has any effect or before we can see anything so it could just be a long delay as well.
Does a Second Internal Link Housed in a Shortener Result in More Juice
Being Passed to the Page?
This test is from Ryan Hough. We have sent multiple links from one page to a test site and we have seen that multiple links doesn’t seem to add any extra benefit, from the same page. The idea where you have a pbn and for some reason, do the old school tactic of three links on that page going to your target page, you are not going to have extra benefit from that.
What Ryan’s idea is, what if that second link was a link shorterner, in the sense that Google doesn’t necessarily see that it is the same destination because it is hopping through some redirects, maybe that would game it so that the shortener would then provide the extra juice – kind of getting that 2 for 1, from the pbn page or a site that you control, you put one link which is a regular link going to your target page, the second would be a shorterner going to your target page.
We found that it doesn’t make a difference and Google seems to be able to figure it out. The same way that it is figuring out multiple links on one page. This was a cool ninja hack idea but what we saw again though is that you really can’t double the value from one-page linking to another page with multiple links.
Dori: That is good to know, these kind of things let us know what really works and what doesn’t work and what doesn’t is almost as good as what does work.
For sure. You see these stuff in forums and other Facebook groups and stuff like those – here is what you need to do, here is the thing, do it this way, and so we try it. It doesn’t work. At least in this situation.
Does Google Limit the SERPs to One Exact Match Domain?
This is the one that I liked the most. An EMD, in case you don’t know, is a domain that is the target keyword that you are going after – exactly. So if your target phrase is “best red bicycles”, the EMD would be bestredbicycles.com. Your domain matches your search exactly.
Google has proclaimed for quite some time that EMDs are dead. There are a lot of SEOs that proclaim that they are dead. We have found that they are not dead. They totally work. Is it the be all, end all? No, but it is a strong factor. If you are in that space, getting an EMD is a good idea.
Recently, John Muller came out and reported that Google now has a filter. They will allow EMDs, but only one per SERP. Fortunately, we have Josh Bachynsky that watches the hangouts and reported it. I was lilke, let us test that. In my mind, I was thinking that it could not be true. It simply can’t be but, you never know. This is a complete statement. This is them stating this as a fact that they have a filter and only one EMD per serp.
I looked around my desk and picked the three first things that I saw which was a microphone, cup, and mouse. Then, I did a couple of searches just to be sure. This came back with results – 160,000 results so it is not millions upon millions of page but there are results there. Then, I bought three EMDs, the .com, .co, and .info for flavor and I launched them.
Initially, within day one, the .com hit within the top 50 for a moment, maybe a couple of hours and then fell out and the .co and .info weren’t anywhere to be seen. I set them up in Search Console and submitted them via the Search Console which is how I got them indexed.
The next day, I was checking the H1s and this is a side note – if you ever done tests with domains where you are trying to put content on the domain, you actually have to have some different content. It is really difficult to get domains homepages with duplicate content to index and show.
What I did, because we were not looking at onpage factors here, we are just looking at the domains. I got different content on the page optimized to the same amount but then my htags are different. On the .co, for example, the H1 was number one microphone cup and mouse. I think I gave some example here. My H1s, the microphone cup mouse you need, number 1 microphone cup mouse, best microphone cup mouse.
Those were the different H1s, the H2s were different, that is where they differ but the overall optimization was the same.
They started ranking the next day for a lot of their htags. You can do number 1 microphone cup mouse and the .co is there, same for the other pages but they weren’t showing up for the actual EMD search until three days later when all three were on page 1.
There are some options here. Why did this happen? Option one was it is a timing thing. The algorithm will catch them soon. I think that is the most generous way to look at these results which is John essentially lied, that Google is just a liar.
Dori: I know that Google does. We kind of think that John doesn’t, most of the time. I have a feeling the algorithm may catch it sooner or later but there is a filter, but it is manual, not algorithmic. That could it be, too.
Right, that is the last one. This one is, if there is a filter, number five is the most plausible to me and it makes sense with some of the other things that I have seen from Google. For example on page speed, Google said on 2012 that page speed is a ranking factor and then they give you three different tools where you could check your page speed – all of those tools though for the page speed focuses on page load time. They let everybody assume that page load time is the ranking factor. We found that it isn’t.
Page Speed is a ranking factor but it is not page load time. They didn’t lie, they just weren’t terribly clear about what they are talking about. Same as this, there could totally be a filter but only comes into play in really high volume, really important keywords where they actually have manual review and then they manually filter them out. They didn’t lie, there is a filter, it is just not applied across the board.
I think at best, it is number 5 cause I don’t know that they are intentionally lying. I think they are intentionally hiding the ball. They want to give out some information but they don’t want you to know exactly where it is so they hide the ball.
In doing so, they are saying sure there is a filter but it is not algorithmic, it doesn’t apply to all keywords which is something because they normally do not like EMDs that much so they would like that information to go out and maybe people would stop doing them on their own for fear of that filter but really that filter is only for some certain very specific situations that is manual.
Will a YouTube Video with Sound Rank Better than a Video without Sound?
I like this test a lot because it falls with the idea of good content and there’s a lot of thought process about looking at images, understanding images, interpreting images. What if Google is listening? Is Google listening to videos? So what we are looking at here is the idea of a video with sound is better than a video without sound, all things being equal.
We made this a two parter, one was just launching the videos into the Serps, sound and no sound, to see which one would rank above. But then we also sent links from them to test pages. The idea is maybe a link from a video with sound is more powerful than a video without sound and the idea is this is good content, this is preferred so it will pass more juice, it has more authority, and Google likes them more, is kind of the concept here.
I ran this test five times just because you need a lot of run in this sort of things and the results were really all over the place. This was pretty similar to what we found so in this result, the video without sound wins, the video without sound loses, but the link video with sound wins, the link getting from the page with sound. The results were just all over the place.
The bottom line on this one is that Google is not listening to the videos. There is no extra juice coming from sound. I don’t know that Google is going to get any extra relevance or understanding from sound, I do not thing Google is listening to anything at all. Also, when you think about the bandwidth, the computing power that it would take to listen, I think it would be absurd right now for even their capabilities.
Any kind of sound benefit just doesn’t exist at this moment in time. I think this also kind of plays into some other thoughts of what Google can and cannot do. Google is not as omniscient as people would want to believe. We know that they are not looking at the images all that well or being able to read what is on the image, and now we also know that they are not listening to videos.
Do Bad Words Affect Rank?
This test comes from Ted Kubaitis. As I mentioned in the write-up, Ted has a very foul mouth and comes up often when he writes. I think he might have Turrets. He is concerned that his bad words would affect his ability to rank.
I put in bad words, I put them in there. Bad words for the sake of science is a hilarious thing. That was intentional. Not only bad words but I also wanted offensive words. Words that are clearly not good, with the idea that there is no way around it. Really offensive words that you just shouldn’t use, with the idea of, if this is a problem, something in there would be considered bad, if this is an issue.
This test was run three times. In two of the tests, as soon as the bad words were put on the page, they shot right to the top. They fell out for a while and then they all came back. In the end, they pretty much all came back to their original ranking spot. Bad words are not being filtered.
But then, we took it one step farther though. What about safe search – when you put on safe search, is this going to filter these results or is this going to change these results and there was absolutely no change.
Additionally, none of the pages were filtered. So, if you got safe search going on, that should filter these out if safe search were reading and they have a set of words and it is reading them and seeing them as a chunk of bad words that need to be filtered, safe search did not filter them. I found that pretty interesting. The thought process there in my mind is that safe search might be a manual thing. Maybe sites get flagged and they have to be reviewed manually before they are removed from safe search, based on words.
Because if you think about it, an article having a bad word in it doesn’t make it a bad article or something that should be filtered from the general public or even filtered on a safe search. Some of the words in that cluster could be used in a way that is instructive or could be describing what somebody else said and in that sense, they would not necessarily need to be filtered by safe search. Maybe it would be bad for them to be filtered by safe search. So safe search might be a manual thing.
The next thing I am going to do and I want to make sure that we have this test in is I am going to put some nudie photos on the pages and see if they filter automatically. If Google would say it is a picture of a naked person and now let us filter it or maybe not. Maybe not at all and it is completely manual and not any kind of algorithmic thing, whatsoever.
Can a Single Exact Match Link from a PA, DA1 Site Beat a
Highly Optimized Page?
This comes from Ryan Hough. What he did was, he stacked on page factors within the serps. What Ryan has been doing for this whole past year, which has been a labor of love, has been building quasi real test environments and in doing so, he has been slowly but surely adding one element to large sets of pages in order to create a somewhat real set up and also seeing how then maybe, what he is calling a wall, could get through or over the wall, what factors could jump you to the top.
What he did was, this first url got all of these factors – the target keyword was in the title, its got keyword density of 2.5%, the H1, the url, bold, etc. This url has all but the title, this all but the title and keyword density in the body, etc. down to the last one that has nothing.
Then, he has them submitted and is adding them as they go. What is pretty sweet is in the results, with the exception of this one, all the others are ranked in order of how much optimization they had. One oddball out there but the rest stacked in order which is unbelievable.
The point of this test though is he then took a link of DA1, PA1, the lowest level link and fired it into the page that did not have any onpage factors with the idea of, do links trump all? Are links that much more important than onpage, that the lowest of low link will move this page and how far will it go, will it take it all the way to the top, will it beat one factor, two factors, half the factors, or what is going to happen.
In this situation, what was cool with this one link is it didn’t affect anything. It didn’t move the page that has nothing an one link stayed at the bottom and hasn’t moved in three weeks. So the onpage factors are more important than the lowest level link that you can get.
Dori: This is an incredible test. Is he going to gradually move up the power of links and maybe multiply those links?
That is the idea. The problem is that a lot of this serps become volatile quite quickly once you start throwing links at them so we will see how it goes but that is the idea, to gradually raise the links to see if we can move that bad boy up and see if three links can beat these weaker factors, we’ll see.