MayDay - What Are the Quality Signals?

[MisterGemini]

In another thread this discussion started, and I thought it would be best to bring this into its own.

It starts with this article:

http://searchengineland.com/google-confirms-mayday-update-impacts-long-tail-traffic-43054

The basics of it come down to this part:

My complete speculation is that perhaps the relevance algorithms have been tweaked a bit. Before, pages that didn?t have high quality signals might still rank well if they had high relevance signals. And perhaps now, those high relevance signals don?t have as much weight in ranking if the page doesn?t have the right quality signals.

The question then is, what are the quality signals that Google sees/looks for?

Are these the types of quality signals similar to what you may have experience in Adwords?

More specifically, what are the factors that they take as 'quality signals'?

This change has impacted a lot of mini-sites, and even some business sites as well, like ecom sites. So what what is it about these sites that google gets 'quality' signals from?

My initial thoughts:

1. CTR - Does the number of times your link is clicked on in Google now impact the quality signaling? Humm, that would seem more like relevance.
2. Bounce Rate - If you are using Ganalytics, then you got what you deserved. This information could easily be collected to determine the bounce rate on a page, and thus signal the 'quality' being low.
3. Content Size - Does haven't more content contribute to quality now? Were the pages that were affected because they were just to thin?
4. Content Methods - Could it be that quality is higher if your site has a multi-media mix of text, images, video, audio?
5. Incoming Links - This is unlikely because the link domain has never been much of a quality scoring as much as a relevance signal method.
6. Site Navigation/Structure - I believe this could impact quality. A well structured site has always been good with google.

These are just some quick off the top of my head possibilities. I may be way off on this, because to be frank, I have never really paid much attention to googles quality scoring accept when it came to Adwords.

Anybody who can contribute as to what factors need to be looked at now for 'quality signals' please join in. Perhaps we can figure this out faster.

Mister Gemini

 

[chockomonkey]

It's always good to try to figure this shit out... Bounce rate, CTR, Incoming links, yea those sound like relevancy rather than quality factors, but who knows? Perhaps with all of Google's data they're seeing correlations between these figures?

Although #2 i gotta disagree with... No facts, just logic: how could google NOT know your bounce rate? It only makes sense that everything you keep track of in Ganalytics, they are already keeping track of (probably even before they released Ganalytics)... they're just sharing now.

 

[MisterGemini]

There is one more thing that came to mine in regards to quality that I didn't list, and that is, code.

This has been discussed extensively by another forum member here based on his findings regarding specific code elements that appear to impact ranking factors.

Thus, could it be part of this quality ranking might be in the way pages are coded, and the tags that are used?

Would this be an indication of a 'quality signal'?

 

[SuperLinks]

Great post and I'm glad this was brought to a new thread. According to the article linked, and Cutt's Mayday video http://www.seroundtable.com/archives/022293.html

I believe some of your initial thoughts are spot on, although I feel that the greatest affect was on MicroNicheSites and sites that solely focus on longer tail keywords.

My initial estimations are that exact match domains, with exact search volume of 2k - 15k were the ones that experienced the largest hit.

I also think that some larger sites that have longer tail keywords with many clicks from the homepage (your site navigation theory) also have experienced a drop in rankings as well.

Bounce rate is something that I always have thought that Google has taken into their equation as there's no reason that they can't implement it. When someone does a search for something and ends up on a website, Google knows which link they click on (CTR) and then also knows when they bounce backs to the results (bounce rate). Therefore if there's a site that does receive clicks, but receives a pretty quick bounce back to the results then chances are that's a bad result for that search term.

Content methods - I think that this has something to do with it. I'll need to look into this as it's only a hunch, but I believe that this could have a play in rankings.

I'd love to hear what others have to say. As I begin to dive deeper into analysis this weekend I'll report back on what I find.

 

[MisterGemini]

Great post and I'm glad this was brought to a new thread. According to the article linked, and Cutt's Mayday video http://www.seroundtable.com/archives/022293.html

I believe some of your initial thoughts are spot on, although I feel that the greatest affect was on MicroNicheSites and sites that solely focus on longer tail keywords.

My initial estimations are that exact match domains, with exact search volume of 2k - 15k were the ones that experienced the largest hit.

I also think that some larger sites that have longer tail keywords with many clicks from the homepage (your site navigation theory) also have experienced a drop in rankings as well.

Bounce rate is something that I always have thought that Google has taken into their equation as there's no reason that they can't implement it. When someone does a search for something and ends up on a website, Google knows which link they click on (CTR) and then also knows when they bounce backs to the results (bounce rate). Therefore if there's a site that does receive clicks, but receives a pretty quick bounce back to the results then chances are that's a bad result for that search term.

Content methods - I think that this has something to do with it. I'll need to look into this as it's only a hunch, but I believe that this could have a play in rankings.

I'd love to hear what others have to say. As I begin to dive deeper into analysis this weekend I'll report back on what I find.

That's some excellent observations.

So to summarize the demise to the micro-site is due to the specific long tail targeting. This would mean that the long tails go to the sites that are higher up on the scale of the subject. This would suggest that google might have effectively implemented a categorical method of ranking sites. Which doesn't surprise me, as if you note using googles keyword tools, all the keywords always fall into specified categories that they generate.

This suggests that maybe quality is deemed by the expanse of your category focus, rather than on the extent of your micro niche focus.

If my site was about black leather sofas and nothing else, this new quality score would toast me. If my site was about leather sofas, and my site went beyond product and also dove into history/info on 'leather' and on 'sofas'.. I would be sitting pretty right now perhaps.

This effectively means that if I want to gain a strong quality signal on black leather sofas, I need to start from the top 'leather sofas', and then work my way down both in structure, and in content.

As for bounce rates, has anybody considered the possibility of generating clicks through google? This could effectively manuipulate the bounce rate in your favor. Years ago, I actually had a tool developed to do this, but it never got finished the way I wanted it to be. The initial tests though even with it half working were surprising. I tested it on pages in the 100s, with just a few runs, I bounced up to the top 20. This was years ago though. I am certain today requires a more elaborate approach. This is of course, if this appears to be a quality signal that is impacting terms today. Whups, I guess that would belong in the BHSEO area.

 

[westee]

So far none of my micro sites have been effected, at least the ones that have been in stable 1st page spots.

The biggest thing I have noticed is in one of my oldest sites. It is not a micro site, but more of an authority type site. It has been sitting in #1-2 spots for most of it's keywords (roughly 10-20) for over 4-5 years now.

It dropped yesterday to page 9...and today I gave up looking for it, lol. It's still indexed just not anywhere in the top 20 pages for anything. This site never made too much money as it was focused on a celebrity.

Now the original site was just simple static html. I had decided to revamp the site with a new blog interface. (2 days before this update) The next day after the changes (one day before it dropped completely) All keywords where sitting at #1 spot, most of the pages that weren't already at #1 had risen from the 2 spot.

Now I realize that after being stable for so long and the many changes I made to it that it is probably dancing now. (at least I hope lol)

I can tell you that I also removed Adsense from the site and replaced it with a CPA Gateway to block certain images on the site.

So this is where "bounce rate" bothers me. Because after the removal of the adsense and adding the gateway, my bounce rate went up around 20% according to G-anal-ticks.(usually was 20-25% was showing 40-50ish%)

So maybe bounce rate is something to look at. I can't really tell with the site because I knew it was going to drop with the changes I made, at least for awhile. Cause G was probably looking at it like a completely new site with the changes made. Hoping it will pop back up.

Maybe I am just lucky with the micro sites for now, but it hasn't affected them in rankings or earnings yet.

 

[stung]

I'm lost amongst all this..some people seem to have been hit so hard while others have suffered almost nothing.

One of my top sites is a fairly long tail search, but my site is just a 5 page piece of crap. It's ranking in position 2 still and hasn't moved at all...even though it couldn't really be more obviously an adsense site.

I think they may have done something with a strong factor based on how long the domain is. Like, above 12 letters...you're suspicious, above 18 letters you're clearly taking the piss and so on...?

 

[MisterGemini]

I'm lost amongst all this..some people seem to have been hit so hard while others have suffered almost nothing.

One of my top sites is a fairly long tail search, but my site is just a 5 page piece of crap. It's ranking in position 2 still and hasn't moved at all...even though it couldn't really be more obviously an adsense site.

I think they may have done something with a strong factor based on how long the domain is. Like, above 12 letters...you're suspicious, above 18 letters you're clearly taking the piss and so on...?

What we really want to focus on here are the quality signals that this seems to stem from.

You are voting up that the domain length could have an impact. Are you seeing this perhaps with domains with dashes being used? That I can understand. I wonder though if google should do devalue a site because of it's domain length which was purposefully lengthened some years back by ICANN.

I have been pouring over everything I can get my hands on related to this. Thus far, I have gotten nothing but confusion from most.

I spent some time just looking at the search results for some random product queries myself.

The quote from Matt Cutts about 'greater content' also got me to thinking about what I was seeing in terms of amazon and other big sites, and some autoblogs.

I am getting a sense that the quality signals are really based on the user response to the page.

 

[stung]

I think that Google may have put some kind of penalty on very long domains, because most of the time the motive for having those is just to rank longtail without much care for making a big, 'good' site. I'll see if I can get some stats from mine and report back.

 

[MisterGemini]

So far none of my micro sites have been effected, at least the ones that have been in stable 1st page spots.

The biggest thing I have noticed is in one of my oldest sites. It is not a micro site, but more of an authority type site. It has been sitting in #1-2 spots for most of it's keywords (roughly 10-20) for over 4-5 years now.

It dropped yesterday to page 9...and today I gave up looking for it, lol. It's still indexed just not anywhere in the top 20 pages for anything. This site never made too much money as it was focused on a celebrity.

Now the original site was just simple static html. I had decided to revamp the site with a new blog interface. (2 days before this update) The next day after the changes (one day before it dropped completely) All keywords where sitting at #1 spot, most of the pages that weren't already at #1 had risen from the 2 spot.

Now I realize that after being stable for so long and the many changes I made to it that it is probably dancing now. (at least I hope lol)

I can tell you that I also removed Adsense from the site and replaced it with a CPA Gateway to block certain images on the site.

So this is where "bounce rate" bothers me. Because after the removal of the adsense and adding the gateway, my bounce rate went up around 20% according to G-anal-ticks.(usually was 20-25% was showing 40-50ish%)

So maybe bounce rate is something to look at. I can't really tell with the site because I knew it was going to drop with the changes I made, at least for awhile. Cause G was probably looking at it like a completely new site with the changes made. Hoping it will pop back up.

Maybe I am just lucky with the micro sites for now, but it hasn't affected them in rankings or earnings yet.

You are not alone in those bounce rates. It looks as though a LOT of people are experiencing the same thing. Which is a clear indicator that this update has not improved Googles results, but diminished them, since people are getting to pages that they really were not looking for, and bounced.

So don't be too alarmed.

I think that as long as the long tails don't have a higher quality scored home to go to, you will keep your spot. At least until a bigger fish comes along and make a more quality signal optimized page.

Ohhh SNAP.. we are talking QSO now, not SEO! I call dibbs on the coin phrase

When you think content, and the page, and what people are trying to get, I am starting to move towards the idea that a mashup type setup is what google is viewing as 'greater content'. If your page provides lets say:

• product info
• product reviews
• what people are saying on twitter about it
• what people are saying in forums about it
• what people are saying in blogs about it.
• snippets from articles written
• wiki references

Now it is undeniable, that is one honk'en great big ball of great content all on a single page. Odds are better that the visitor is going to take a little more time going through it before they bounce, if at all.

Could it be that the quality signal could now be based on how many other sources of information you utilize within your page? Is this the new QSO?

 

[MisterGemini]

I think that Google may have put some kind of penalty on very long domains, because most of the time the motive for having those is just to rank longtail without much care for making a big, 'good' site. I'll see if I can get some stats from mine and report back.

I'm going to test this out in the future. I'm going to create a domain that is the maximum I can register at godaddy, and see what happens.

 

[westee]

As far as domain length. My micro sites are all around 10-17 char's and they weren't affected...yet anyways lol.

Try searching for the "longest domain name in the world", I know it's not competitive or anything but they still rank the domains going after that phrase(except for #1 spot).

You are voting up that the domain length could have an impact. Are you seeing this perhaps with domains with dashes being used? That I can understand. I wonder though if google should do devalue a site because of it's domain length which was purposefully lengthened some years back by ICANN.

The domain that I was referring to that dropped was a dashed domain. (It is firstname-lastname.com with 6 characters in each name.) This is the only dashed domain I've ever had so...but I do know that over the last few years (when buying dashes use to be the "cool" thing to do) they seemed to get weeded out of the rankings pretty quick.

 

[hoot33]

anyone noticed all sites scraping content from amazon are totally gone for every keyword???

Every one of my sites using plugins like reviewazon are gone.

Every one of my sites regardless of what it is, it's totally delisted. Only showing up for a search for the exact url.

 

[Jekashi]

I've been "toying" with this theory of a Mayday update for the past few weeks now. I have a small, 15-page website set up on a Blogger account that has been sitting for roughly 2 years now, mostly unchanged. I've seen it in the past four days for it's targeted 4-word 2k search per month phrase dance from 3rd on page 2 to second on page one to 9th on page 13 (the last page at all for my results without omission) to page 2, rank 3 yet again.

My conclusion is that the big crisis being faced right now is simply a matter of resorting. The caffeine preview some months back showed all my specific keyword sites going UP in rank for keywords I targeted and DOWN in rank for keywords that were "unintentional." I have a feeling that a site which mentions "Cooking Recipes" isn't going to have a subpage rank as well for "Sloppy Joes" compared to a website with every hamburger you could imagine.

I don't think this is as simple as a matter of CTR or bounce rates. I will admit that Google already has a local version of bounce rates implemented, but to implement it server-side would open leave it open to being exploited. CTR and bounce rates can be artificially inflated, so can traffic (as Alexa taught us).

Rather than relevance, I say that specificity is the key here.

As far as dashes go, I doubt that matters at all. Can you think of a valid reasons to devalue dashed domains? They're spammy? Why would someone get a dashed domain in the first place, probably because the regular version of that domain was already taken. Ergo the regular domain existed first and was probably being backlinked to long before. See where I'm going here?

 

[MisterGemini]

"Mayday" is bullshit. It's made up over at webmasterworld.
It was there conclusion over data they didn't understand.
And google just picked up on that.

Fact is, that anybody here that has some knowledge on how mathematics work, know that a algorithmic change on a database as large as the WWW, will take month before you start seeing effect.
Mayday is just the workings of Caffeine.

So Matt Cutts official statement about it being another algo change separate from Caffeine is an outright lie? I can accept that because he has sent out many other poison pills, but he also said that this was coming from the quality department, not the spam department (his).

What then is the mathematical equation(s) being used for quality signals?

Or would you say that the changes taking place now are not what everyone thinks they are?

 

[MisterGemini]

anyone noticed all sites scraping content from amazon are totally gone for every keyword???

Every one of my sites using plugins like reviewazon are gone.

Every one of my sites regardless of what it is, it's totally delisted. Only showing up for a search for the exact url.

Yikes that sounds brutal. I have heard that amazon seems to be retaking all the spots that affiliates used to have with their stuff.

Though just by doing a few random checks, I am also seeing some still holding their own.

I am going to guess it's not necessarily just about amazon, but perhaps once again the data on your pages is too thin in comparison to what amazon itself offers.

I'd certainly like to hear from others also who have similar sites.

 

[MisterGemini]

My conclusion is that the big crisis being faced right now is simply a matter of resorting. The caffeine preview some months back showed all my specific keyword sites going UP in rank for keywords I targeted and DOWN in rank for keywords that were "unintentional." I have a feeling that a site which mentions "Cooking Recipes" isn't going to have a subpage rank as well for "Sloppy Joes" compared to a website with every hamburger you could imagine.

Rather than relevance, I say that specificity is the key here.

As far as dashes go, I doubt that matters at all. Can you think of a valid reasons to devalue dashed domains? They're spammy? Why would someone get a dashed domain in the first place, probably because the regular version of that domain was already taken. Ergo the regular domain existed first and was probably being backlinked to long before. See where I'm going here?

My aversion to the dashed domain was based on tests done a few years ago. Since then I have just had an aversion for domains-like-this-one.com. Maybe the age/linking history of the existing one is the case as you say.

As for the sloppy joe, I think that's a good example. However, I'm getting the impression now that this is more about page quality and less about domain authority. If the cooking site has a totally kickass sloppy joe page, it could beat the hamburger site that didn't do as good a job with it.

Are we getting anywhere here?

 

[alex1]

There is one more thing that came to mine in regards to quality that I didn't list, and that is, code.

This has been discussed extensively by another forum member here based on his findings regarding specific code elements that appear to impact ranking factors.

Thus, could it be part of this quality ranking might be in the way pages are coded, and the tags that are used?

Hell no... I think they will start to use more and more off-page factors, specifically those that are harder to manipulate with. They dont need your site running GA to get the following stats just by measuring users interactions with SERPS:


- how much time user has spent on your site befere coming back to SEPRS and continue their search activity

- how often users click on your site in SERP (I would give it LESS priority if I were Google, becuase you could write a catchy title and people will click but leave soon and continue their search)

These are even harder to manipulate with than building backlinks.

The only problem is if it is a long long tail KW with 5 searches per day, then there is no data for Google to analyze. But if there is at least 20 searches per day, then Google could measure the stats for top SERPs real fast and decide who are doing well (quality site) and who are doing bad, and push the bad sites down in SERPS

For high traffic KW, Google dont need to ask anybody - its users vote for the good sites by thousands clicks and finished searches that terminate on quality sites, and Google keeping its notes ...

 

[biks]

Well this explains why my traffic dropped in half starting today (Friday, June 4th). I've got about 50 niche sites that are keyword specific with exact domains. Most were on page 1. ALL text was unique and had about 600 words per page. Plenty of Amazon products, photos, links, videos, etc. (nothing rehashed from Amazon). Most of the sites have dropped about 10-15 spots. I thought it was maybe the recent very short Scrapebox runs I've done. Guess not.

I have one authority site that I've built over the years that was sitting at #3. It's still sitting at #3. Tons of content on there. Gee...do you think Google wants to see content heavy sites?

 

[hoot33]

Im going out on a limb here, but Im thinking then "accuracy for longtails" that they're looking for is more geared toward targeted and non-targeted longtails.

eg. My site is on "pitbull calendars", I rank number one for pitbull calendars, all meta details are geared toward "pitbull calendars" and because in a comment someone said "pug calendars" are cuter, I rank 2 for "pug calendars". Being a dog calendar website and having the term "pug calendars" on my site this makes sense.

But, the rest of my site is targeted to "pitbull calendars" so with this update my site has fallen from the "pug calendars" serps, after all, my page was not created for that term and it had just been mentioned in passing.

Again I have no evidence to back this up, just makes sense to me.

Anyone have any thoughts??