In this webinar, we talk further into detail on the tests for this month's dossier and how you can apply it to your site or your client's sites to help you move rank.
Test 60: Mythbusting: Will Adding A Conversion Goal Through Google Tag Manager Change Bounce Rates?
Test 61: Does Google ‘See’ (And Index) Content That Is Hidden In An Accordion/Drop Down Tab?
Test 62: Is EXIF Data On Its Own A Ranking Factor?
Test 63: Do Brand Mentions As Implied Links Affect Rank?
Test 64: PNG vs JPG: Is There A ‘Better’ Image File?
Test 65: Mythbusting: Adding .html To The End Of Your URLs Gives A Boost In Ranking?
Test 66: How Long Does It Take For A Page To Index Using Different Indexing Methods?
Test 67: Penguin 4.0 Follow Up Tests
October 2016 Dossier
Transcription Below:
Mythbusting: Will Adding a Conversion Goal Through Google Tag Manager Change Bounce Rates?
There is this myth going around that if you set a conversion goal in tag manager, that would lower your bounce rate. Bounce is when somebody goes to your site and they read your content and leave without doing anything else, that is a bounce. When you go into your Google Analytics, you will see that they record your bounce rate and it makes sense that a lower bounce rate would be better. We’ve tested that and we’ve moved the needle with bounce rates alone.
Now we are looking for ways on how we can fix our bounce rate. Then there is this myth going around that you can do that through setting a conversion goal so we set this up. This test was spearheaded by Ryan Hough and he did a fantastic job.
In the end, what we found is that when you set up tag manager, you can get your bounce rate to trigger. You could get an event to trigger which tells analytics that something is happening on the page. The idea when you are getting your bounce rate down is it’s a click. In this sense, when you fire the event and we were doing it in a certain amount of time, you can actually lower your bounce rate.
What is really hot is after this test, I have put this in place with another SIA member – Dwayne Kula – I worked with him to put this on his site and his bounce rate has dropped dramatically. It has been amazing to see it put into place. The nuts and bolts of this are in tag manger. You can set an event to happen – it can be a certain amount of time on the site, it can be when somebody scrolls, and that fires and event and what that does is then it triggers in Google Analytics as though somebody has clicked, they have done something on the site.
What that does for you is that it lowers your bounce rate. What is also interesting is that it records time on site. If somebody bounces from your site, the time on site isn’t recorded. For example, you have this great piece and someone reads it and loved it but they do not do anything else because they do not need to and they leave but maybe they were on your site for 3, 5, or 10 minutes, none of that time on site is recorded when they bounce.
If you have this set up, they will actually record that time on site which will help your other stats which could be good especially if you are doing client site sort of work. You can say “hey, our time on site is increasing and people are consuming this content”. You can get those metrics up as well.
This is a fantastic test and I am very happy with the results. I am putting it in play in my own client work now and the results have been excellent.
Does Google “See” {and Index) Content That is Hidden in an Accordion/Drop Down Tab?
One way you could get content on a page is in these hidden areas, you have an accordion drop down or a tab feature. If you read some different things, you would have different ideas on whether Google actually reads the content that is hidden – if they can read it, if they care about it. Some say they can’t, some say they read it but they discount it.
At a minimum, we just wanted to see if it is even readable, just to see if where we are with that. When we ran this test, we did it two times. Once was in a dropdown and in the screenshots you can see the idea of it – it’s a standard WordPress plugin dropdown so it is closed and you click on it and then it opens up information.
In this information, we put the keyword one time so the target keyword wasn’t in anywhere else on the page but one time in the hidden area. The page indexed for that keyword so Google was able to read and index the page for that particular target keyword. I then did it with tab content that you often see on websites and I put it in the middle section so that it would default close. The first tab is the one that is showing and then the next tab would be the one that you would have to click to actually see the data.
The keyword was in one time and the same result – the page indexed for that one keyword. Keep in mind that the keyword was only in the page one time and only in the hidden content and they very quickly indexed and showed within a day on both of these. What is great about this situation and the complete benefit is that when somebody tells you that this content, if you put it in a hidden area – a dropdown or tab – that Google does not read that and doesn’t look at it at all, that is completely untrue.
It is clearly read, the content is indexed, and even having the keyword just one time would allow a page to show for it.
What we are going to do is look at discounting the keyword – we are trying to play with the idea of optimization rates within these hidden areas to see if we can identify if Google actually discounts this content or not. I actually have a feeling that it doesn’t. I think it works just as well as any other content.
Dori: Shelby Larson actually did a presentation on doing some SEO work inside accordion and tab areas that it did rank and that it did help. This was a couple of years ago. I am glad to see it is still effective because that is a good place to add content on a page when you don’t want to mess up the article or design that you already have.
Is EXIF data on its own, a ranking factor?
EXIF data is content that you can put on into an image. That can be the location, the name of things, etc. If you are into camera, you can put things like shutter speed, aperture, light, etc. You can put a lot of information to a digital image and that is what EXIF data.
For our purposes, geolocation is very important. You can put a description, you can put keywords in the image and what we want to see if just by having EXIF data alone, would that move the needle?
Hilariously, the test page dropped from 3 to 4. I don’t think that EXIF data is a negative ranking factor. What we are looking at here is when something tells you to fill it out, waste your time, put as much data in, that is probably not going to move the needle at all. EXIF data, in of itself is not a ranking factor. Which is important because there are some programs out there that says fill in everything. I would recommend not doing that.
When it comes to putting data on things, what I do are the description, the image alt. I will geotag images when it is relevant to local SEO and that is good but I do not waste my time filling out as much information as possible because I do not think that will move the needle for you.
Do Brand Mentions as Implied Links Affect Rank?
When you look at the future of SEO and you theorize where things are going, brand mentions are a hot topic right now, specifically looking at the idea of you do not need backlinks coming to your site and you just need mentions of your site and Google will treat those as backlinks and understand what they are. It will give you a boost and build your brand. We hear a lot of those concepts.
We put 20,000 brand mentions out for one of our test pages. We basically had two fake keywords set up. The idea of all these pages have one fake keyword and then the other fake keyword that was different on each pages would be the brand. Each page had a different brand. Then without hyperlinking, we brand the target page which was the middle page of the 5 test pages that we had.
We placed the brand and the product across the web. It did not move the needle a bit. I checked this morning to recheck – still nothing. Just having the text out there across many pages did not influence the ranking at all. Brand mentions are important but they still need to be hyperlinked in order to get the benefit of that mention.
PNG VS JPG: Is There a “Better” Image File?
This test was put together by Derek Booth. If you are doing client side work, one of the things you get into in the technical side of SEO when you are optimizing a site is making suggestions on making the the site better. One thing we looked at is if Google has a preference over PNG and JPG files. The idea is that PNG is considered as a superior image – it looks better in comparison compared to the compression of JPG images.
The idea behind this is that maybe the PNG files are better and Google likes it more because what they talk about is that you want a prettier and more authentic web. You want better content. If you want a better content and more unique things, PNG provides a better image and maybe Google prefers that. That is the genesis of this test.
Derek added the 5 page test pages, 4 were JPG and one was PNG which was the middle ranking one. Hilariously, it dropped to 5 which was a very unexpected result. We need to retest it and it is being retested right now. We are going to switch it.
The idea is that it is all PNGs and 1 JPG to see if the JPG would rise to the top and if there is a preference. I would not necessarily freak out and get rid of my PNGs but it anybody ever tells you that there is a specific file extension, you could probably bust that immediately.
We are working on the next test to see if there is a ranking factor the other way. Probably not, it would be interesting if there is. At a minimum, this image idea didn’t pan out and it is not anything to focus on in terms of changing things to PNGs.
Mythbusting: Adding .HTML to the End of Your URLs Gives a Boost in Ranking?
This was brought to our attention by Meredith, she has found on a Bryan Dean forum where they are talking about ninja hacks that it is a fun thing you could do that people are swearing by – intentionally adding .html to the end of your URLs because if you put it, it helps search engines understand and it would give you a ranking boost.
I did a little research on it to check if it works and saw a bunch of people saying the same thing that is why we want to get into it. Should we put a .html on it or a .php, etc. – does it give an advantage?
We run the test and in the end, it didn’t provide any ranking factor. In fact, it dropped the page at the bottom. What was very interesting is that the SERPs would not adjust. We launched 5 pages that were regular. The middle page, we changed into .html and what interesting was that even though we had the page crawled multiple times, the serps would not put the new page in. It was hanging on to the old page.
If you were to click on the new page that was there that was in the serps, you would run into a 404 page because it wasn’t seeing the new .html. This is very significant because when we ran the http vs the https test, we ran the same thing and all we did was run 5 pages that were http and then on the middle page, we changed it to https but we didn’t put any forwarding on, we just let it go to see what would happen and immediately, within a day or two, the serps changed the URL.
It saw that this is the new page and accepted it and ranked it to the top. In this situation, by changing the end to .html, the serps wouldn’t bring the new page in. It was continuing to show the old page which was showing a 404 and then dropped it to the bottom.
This could indicate something more nefarious about 404s, at least it was seeing it wasn’t a good link compared to the other links. Not only did it not provide a ranking factor, but the search engines weren’t taking the time to get the new page up and running. Don’t do it.
How Long Does It Take for a Page to Index Using Different Indexing Methods?
We are looking at different methods on getting links indexed. What we were looking at was using the Google URL submit tool, Search Console Fetch and Render, Linklicious as a standard link indexing tool, publishing Google Doc that was hyperlinked and one that was not. Google URL submit and the Search Console was within 15minutes. They happened almost immediately.
Linklicious as a standard link indexing tool took 8 days and then publishing in a Google doc – what we were doing there was putting the link on the Google Doc page and making it public – that took 10 days. The Google doc that was not hyperlinked never indexed and it still hasn’t. I do not think it will ever index. The Google doc index does work. I have actually used this on a site where I have launched 2600 pages or so and they weren’t indexing so we put them all in a Google doc and published them and within 2 ½ to 3 weeks, they all indexed. Google gets there and it crawls its own stuff and you can get there as a nice indexing tool.
Do you think there is any difference between a Google Doc and a Google Sheet?
No, I think it will do the same thing. For this particular test I did not do that. Google sheets gets indexed and show in the SERPs. I imagine it is probably the same thing.
Penguin 4.0 Follow Up Tests
We’ve run several link tests and some nasty link tests, if you will, trying to look at where the loop of the envelope, where can we push it to. As we know, Penguin 4.0 came out and what we did was we fired a whole bunch of new links. The same style links at the original tests to see if there would be any kind of movement.
1. Unnatural Link Penalty Follow Up: What Does it Take to Get an Unnatural Link Penalty?
In this test, the page is still in the number one spot. We could not sink the page and we have sent some nasty links on it and it would not go away. I think it would be there forever.
2. Paid Links Follow UP
This was a Fiverr gig that we did and what ended up happening in this particular gig is that it was all blog comments – 20,000 blog comments – we purchased the exact same thing and what was interesting in this one is that it moved one spot then got back at the exact same spot. It didn’t get deindexed.
3. Follow up: Backlinks with 25% Exact Match Anchor Text Test
We fired another round at this as well and it is still at the number 1 spot. What was used in the unnatural links is GSA. For the exact match links, we used DripFeed links.
GSA and Dripfeed links use a variety of links so some are blog comments, others are a wide variety of links. What we are seeing between these three tests is do not use blog comments. If you are going to do some, make sure it is a part of a spread of links that you are doing.
In Dripfeed, they actually let you choose a percentage, you can say 10% this, 25% that, etc. I kept things at the standard of the system. Make it diverse, even at post penguin, it is fine. What is interesting with the middle one where we used the blog comments, the worst situation there would be they would be discredited, they were discounted, and they were moved back to their original spot.