MayDay - What Are the Quality Signals?

[MisterGemini]

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??

That's an interesting way to describe it. It means that it's traffic you never really should have gotten in the first place. Perhaps means if you want that traffic back, start building pages specifically for them.

So does that suggest from what you are saying here is the quality signal really has to do with fully matching the keyword search term to the content of the page that has more information devoted to the term rather than just the random word found in content that might actually be about something else. You are right, this seems reasonable enough to make sense.

 

[MisterGemini]

I found a post that actually offered some interesting insight about this. It specifically talked about this update impacting the yellowpages sites. You can read the whole thing here:

http://www.localseoguide.com/its-mayday-for-yellow-pages-sites/

The main commentary out of it all was this:

While we are testing various hypotheses on some sites, based on what I have seen thus far, sites that have a lot of URLs (e.g. tens of millions) with thin and/or generic content and have most of their inbound links focused around a relatively small set of URLs or that are not generating regular inbound links have seen the biggest impact.


This shouldn?t surprise anybody in SEO land as unique content, good site architecture and strong backlinking have been the cornerstones of SEO for years. That said, it appears that this algo change has raised the game to a whole new level. It used to be that an authoritative domain could add pretty much any content, even duplicated content, it wanted to its site and with a little bit of effort get strong organic traffic. That does not appear to be such an open and shut case anymore.

Neat about the authority domain stuff, that's what I said earlier. Some of what he is saying about the link focus and smaller set of urls that are not regularly visited also falls in line with some of the other contributions here.

He goes on to make a list of suggestions to combat these effects, they are:

1. Get unique content on your business profile pages and other important pages (and I?m not talking about just putting Yelp and Twitter feeds on them). You need your own user generated reviews, editorial, q&a forums, etc.
2. Consolidate the number of pages you show Google and limit them to only those that are ?valuable? (i.e. have unique content and/or inbound links). This one?s tricky as it?s hard to know on a large scale which ones are the best.
3. Continue or start to drive links to deep URLs
4. Make sure all URLs are as close to inbound links as possible
5. Prioritize valuable content in the architecture so that they are as close to the home page and other valuable URLs as possible
6. Do something remarkable with your service that gets people talking about you and linking to you.

A fairly standard list, aside from the user generated part. I would imagine not everybody does that. 3 out of 6 are all about site structure, and the other 2 are about links, specifically do something to gain links naturally, and get links to all your pages no matter how deep they are (if you aren't doing that, well DUH!).

This raises another factor that none of us have mentioned, which is user contributed content. I think it's possible that this could indeed contribute significantly as a quality signal. Not so much in terms of the content itself, buy by having access to contribute all such content. From a user perspective, getting first person input/response certainly could be considered a quality experience. As long as the response you are getting isn't 'shutup you stupid asshole'.

So what does that mean? Throw up a forum, a pligg, and of course the blog to allow for more user interaction options? This would explain why amazon recaptured their traffic, because every one of their product pages now allows for forum interaction and discussion of the product.

He said nothing about CTR or bounce etc. Perhaps rightly so, as they are factors that are not very reliable in measuring I think.

Think we are getting closer. Thoughts?

 

[SebastianJu]

[*]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.

Not using Analytics wont help in my opinion because google could determine the same thing by simply make a statistics on data they got from google-toolbar. Or they would know when you make a new request at google or visit another site that has analytics on.
Without analytics they would have less data but they would have data and with that data they could count up the numbers for all users. At least I would do it this way.

And I dont think that the bounce rate is so important. Because most of the time when I search something I find it on the first page at a resulting page.

The formula used to calculate bounce rate is: Bounce Rate = Total Number of Visits Viewing Only One Page / Total Number of Visits

So if you have a onepage where you explain everything the user wants to know he in most of the cases wont click other links because he dont need. If you have a onepagesite you would even have an awful bouncerate but the user will have still all the experience he wants.

And at the end you could cheat bouncerate by simply splitting your articles in parts. So that the users have to click 2 or more pages to read the full article.

But at the end it wouldnt make a difference for the user because he can have the info on the first page or on three pages.

Because of that i dont think that bothering about bounce rate is worth the effort...

 

[westee]

The formula used to calculate bounce rate is: Bounce Rate = Total Number of Visits Viewing Only One Page / Total Number of Visits

So if you have a onepage where you explain everything the user wants to know he in most of the cases wont click other links because he dont need.

That makes sense with the site I was talking about earlier in this thread....because I actually changed the site...to entice users towards one page.... vs. the rest of the site... so if they hit that first...then I read the target viewers right..(this is on an authority site.)

if that's the case then it worked how I intended it too...+20-25%.... (the platform I switch to did make much more $ in those 2 days it was there...)compared to adsense, but for only one-two days... and then...down from (like I said earlier in this Perma: http://www.blackhatworld.com/blackh...-mayday-what-quality-signals.html#post1826396)

I think maybe it's just a "post" Cinco De Mayo "dance" just 1 month late...and we know the G loves to dance....(just kiddin')

Giving everyone the annual big google shake up...and OMG what's going on...just because they need a massive viral move to stay where they are....with so many leaving/getting banned/strict adwords(not adsense) alot of big spenders have left to other SE's such as bing/yahoo...

This only hurts them, and hurts there publishers...

Why the change so drastic...right after massive national(us anyways) bing commercials everywhere...fuck em... I survived before them...and will without....

so Big G :bryce:

not you gemini.... the other big g ....just stating that lol.

 

[ipopbb]

I see things like the following everywhere now in online retail:

Absurdly High Keyword Densities

Search: amber earrings

#1 http://www.amazon.com/s?ie=UTF8&keywords=amber earrings&rh=n:3367581,k:amber earrings&page=1

Matches:
> 100 amber
> 100 earrings
62 jewelry

#2 http://www.amberdesire.com/

Matches:
99 amber
12 earrings
2 amber earrings
46 jewelry

Before march of this year you didn't have to stuff anywhere near this much. It looks like google is tuning out e-commerce from the "Everything" bucket. And trying to force retailers into the shopping feeds that favor Google Checkout or AdWords which favors Google.

 

[MisterGemini]

@SebastianJu - Yeah, as I started to show in my last post, I am starting to think the idea of bounce rate doesn't apply.

You are right about the google tool bar also providing that data, as well as google accounts etc. The idea behind not using GA is to minimize that data as much as possible. There is a big difference between google being able to make determinations on your pages based on the occational Gaybar or Gaccount user, vs. being able to see every single visitor from every single source you get. That kind of data is definitive to make calculations on. You remove data elements, then you can't use a calculation to apply. So I still have to stay firm on the believe that Ganalytics will do your site more harm than good, if not because of bounce rate, but because of potential calculations being used against my site in ranking.

@ westee - I totally know what you mean. The sad thing to consider though is that it is very possible that as we go forward in the future, other companies will follow suit with improving their serps with similar quality standards. What we need to figure out is if there are quality signals, what are they. You don't want to be left scrambling after the fact YBing decide to make similar changes, or the industry as a whole starts knocking out certain 'types' of sites.

@ ipopbb - I've seen in your other threads you talked about keyword stuffing pages via tags/codes as being accepted by Google. You think that these methods could help with overall quality scoring?

 

[NoSuchAgency]

I have a couple dozen niche sites with exact keyword domains. A chunk of them took a big hit yesterday (June 4). I watch my ranks every day. This was the first big change I've seen in months, and it happened all in one day.

I went from having about 10 in top 10 and 12 in top 20 to 4 and 11 respectively. A number of sites got penalized (not sandboxed) and dropped anywhere from 10 to 70 ranks for the primary keyword. Those that dropped seem to have a couple things in common:

• Non-product (informational) keywords
• Mix of unique and scraped content
• Adsense was the only monetization on these sites
• Thin sites (1-3 articles) that have been around for months with few updates
• No backlink building at all
• Simple, crappy looking themes

So I guess they were typical old and forgotten MFA-looking sites.

None of my recent product-keyword niche sites that also have amazon ads were affected.

 

[biks]

What I find interesting are the ones I have that DIDN'T take a hit. I used the exact same template for all of my keyword domains. ALL had the same (rewritten) amount of unique content. All had 50% Amazon related stuff. I even rewrote the Amazon descriptions AND renamed/scaled the photo.

The ones that were getting high click through rates and making sales either via Amazon or to a related adsense link haven't budged. I was getting some fairly high paid clicks from the adsense box. These were also my older domains that I had established some descent links to.

I did something right. Most only seemed to have dipped around 10-15 spots. I just gotta compare and contrast to what's different about each, but it's not the formatting and content of the page itself.

 

[SebastianJu]

@MisterGemini

Google Analytics would give google more data but why do you mean that this will lead to your harm? I mean more data can prove that your site is bad or that its a good site. So when you plan to build a good site I dont see why google shouldnt see that your site is good.

For spamsites of course it could be another thing.

No offence but till now this analytics-thing sounds to me like a myth. Born out of the fear of googles allmighty presence. But is it true? Probably only determinable with a test... ipopbb... you wanna set up a new test in your seo-lab?

 

[bhjmc]

Just a note. Mayday started back at the end of April I think it was or early March. Some guy renamed that already long going thread "Mayday" in May and it got the name. Just a note. Maybe others have been having this ranking problem since before May. May was when Google introduced the new design which started affecting even more users.

 

[MisterGemini]

I have a couple dozen niche sites with exact keyword domains. A chunk of them took a big hit yesterday (June 4). I watch my ranks every day. This was the first big change I've seen in months, and it happened all in one day.

I went from having about 10 in top 10 and 12 in top 20 to 4 and 11 respectively. A number of sites got penalized (not sandboxed) and dropped anywhere from 10 to 70 ranks for the primary keyword. Those that dropped seem to have a couple things in common:

• Non-product (informational) keywords
• Mix of unique and scraped content
• Adsense was the only monetization on these sites
• Thin sites (1-3 articles) that have been around for months with few updates
• No backlink building at all
• Simple, crappy looking themes

So I guess they were typical old and forgotten MFA-looking sites.

None of my recent product-keyword niche sites that also have amazon ads were affected.

That's pretty amazing, because I know a guy who lost all his sites in ranking yesterday as well. They were not like your sites though. It literally happened in a 24hr period just like you described. So don't feel alone You are lucky you got some other sites not affected. Like biks mentioned, you can compare.

 

[MisterGemini]

What I find interesting are the ones I have that DIDN'T take a hit. I used the exact same template for all of my keyword domains. ALL had the same (rewritten) amount of unique content. All had 50% Amazon related stuff. I even rewrote the Amazon descriptions AND renamed/scaled the photo.

The ones that were getting high click through rates and making sales either via Amazon or to a related adsense link haven't budged. I was getting some fairly high paid clicks from the adsense box. These were also my older domains that I had established some descent links to.

I did something right. Most only seemed to have dipped around 10-15 spots. I just gotta compare and contrast to what's different about each, but it's not the formatting and content of the page itself.

What you should do is look at the ones that did take the hit, and see who is ranking for those keywords now that you used to. Then see what 'quality signals' you think would send them the ranking vs. your sites.

The ones that are doing good, see if you seen any movement around or below your rankings. I am betting it hasn't budged. Though, judging from what I am hearing all around, it seems this new algo has not run it's course entirely and may reach your other sites eventually. I would give it maybe another month until we see this thing finish. I am sure the other sites not affected were in another niche with different keywords.. right? Possibly a different category of site?

 

[MisterGemini]

@MisterGemini

Google Analytics would give google more data but why do you mean that this will lead to your harm? I mean more data can prove that your site is bad or that its a good site. So when you plan to build a good site I dont see why google shouldnt see that your site is good.

For spamsites of course it could be another thing.

No offence but till now this analytics-thing sounds to me like a myth. Born out of the fear of googles allmighty presence. But is it true? Probably only determinable with a test... ipopbb... you wanna set up a new test in your seo-lab?

When did Google become an altruistic company in your eyes?

I don't know.. a lot of people here have experienced this impact. For a myth, it certainly appeared to have been duplicated in results over and over:

http://www.webmasterworld.com/google/3248965.htm

Look them over yourself.

I have no fear of Google. I know how they operate, and have a glimmering of where they want to go. It's not good, but I'm not afraid of it either.

 

[MisterGemini]

Just a note. Mayday started back at the end of April I think it was or early March. Some guy renamed that already long going thread "Mayday" in May and it got the name. Just a note. Maybe others have been having this ranking problem since before May. May was when Google introduced the new design which started affecting even more users.

C'mon man, that's all you got to say about this?!

I think the algo update is still going on even today.. June. Maybe Mayday wasn't such an appropriate term. More like Googles Spring Smell the Kaka Update.

 

[ipopbb]

@ ipopbb - I've seen in your other threads you talked about keyword stuffing pages via tags/codes as being accepted by Google. You think that these methods could help with overall quality scoring?

I'd say look at the top 3 for your main keywords. See what they are doing... figure out how your site differs in the factors you can influence.

If you are behind in the content tuning or off target in the content tuning then it doesn't matter how much supporting "oomph" you get from backlinks or other factors. Your foundation is your content. You ability to target keywords starts there. Backlinks can only magnify what the content supports. Make sure you target your keywords properly by reviewing the keyword winners in the SERPs... google changes so you just always have to keep up with the times.

 

[ipopbb]

@MisterGemini

Google Analytics would give google more data but why do you mean that this will lead to your harm? I mean more data can prove that your site is bad or that its a good site. So when you plan to build a good site I dont see why google shouldnt see that your site is good.

For spamsites of course it could be another thing.

No offence but till now this analytics-thing sounds to me like a myth. Born out of the fear of googles allmighty presence. But is it true? Probably only determinable with a test... ipopbb... you wanna set up a new test in your seo-lab?

Actually, I have experience in "reasons to fear google"... About five years ago an employee of a client of mine decided to click on some of his own personal adsense ads. Got my client banned from adsense permanently even though my client's account wasn't the one being clicked. The accounts were associated via permissions to Google Analytics. You have to be very careful who you let have access to your reputation with google. Even a $1M annual AdWords spend can't get his AdSense privileges back. I often remind people of those events and finish by asking if the $0.72 in fraud clicks was worth the $250,000 that would have been made over the years with access to the program.

No need to test... Google can query other accounts you are associated with through AdWords, AdSense, WebMaster Tools, Gmail, and Analytics... and by extension of that the websites that all those accounts are associated with.

I don't even scan the SERPs on google from my computer. Logged in, logged out, cookied, IP logs, session ids, ... bulk scanning the SERPs is an SEO practice. I don't want to get flagged for scrutiny so it all goes through anonymous proxies. It would suck if Google came down on me, but it would be career ending if Google came down on my clients because I wasn't careful.

Mistakes aren't allowed in SEO at my level. Traffic from Google is a multi-million dollar channel for my clients. That's why I have a lab and I test for months outside of production.

I know when I make a tweak on a production website what is going to happen and how long it will take to see the result.

The paranoia and attention to detail and due diligence is why I'm successful and so valuable. Its the suckiest and hardest part of the job. But thats where you create value... "Only Upwards" modifications and minimizing risks.

On sites that use GoogleAnalytics I have to be extra careful about what I do on and away from those sites. Fubar-ing an experiment on the side can impact them via google's very real and nefarious account and website associations.

I see no value in being on Google's radar ever. It is only risk.

 

[ipopbb]

Now imagine you have to do the entire web with 200 plus factors.
Thats why MayDay is bullshit. The actual algorithm change isn't actually done in may.
I would say it's most likely the effect of the caffeine update from January.

I concur with M.A.D. ... the damaging changes were most likely deployed in January and first became measurable mid February. Most online businesses started noticing in March when their monthly's were down and the shit hit the fan in April and then the sky fell in May. Everything I've read and experienced supports M.A.D.'s iterative cycles description.

(Starting fights again. I need help.) I, of course, don't put a lot of stock in google patents reflecting reality. I do concede that there is a kernal of truth in them, but for the most part they are obfuscated to protect trade-secrets and descriptive enough to protect inventions. That's all. They are merely barriers for Google's would be competitors to engineer around. Not the Holy Grails many people believe they are.

(Damage Control ) But without them we wouldn't have M.A.D.'s awesome insights and summary on this matter. So I certainly couldn't say they don't have value in the right people's hands.

( FYI: I have a search engine patent of my own! Got it issued in record time too! Only took 7 years! )

 

[MisterGemini]

Ok, so we can agree it started in January.

Then are you saying M.A.D. that this update has nothing to do with quality signals?

What we are seeing now is just the application of algo 190 out of the 200 that were launched back in January?

Do you think we have seen the end of it? Or there is still more to come?

Any thoughts on what factor then is contributing to this shift? Is the Matt Cutts vid bull in regards to how it affects long tail terms, or him saying it's something separate from the update?

 

[hoot33]

Ok, so we can agree it started in January.

Then are you saying M.A.D. that this update has nothing to do with quality signals?

What we are seeing now is just the application of algo 190 out of the 200 that were launched back in January?

Do you think we have seen the end of it? Or there is still more to come?

Any thoughts on what factor then is contributing to this shift? Is the Matt Cutts vid bull in regards to how it affects long tail terms, or him saying it's something separate from the update?

Another thought that just crossed my mind, perhaps the proverbial "sandbox" is true, perhaps google is taking all the sites that do not have a proven track record and putting them in serp purgatory, any webmaster whos just in it for the money will give up. Anyone who is commited will continue to build the site showing google it is not simply a spam site.

Any thoughts.

 

[MisterGemini]

Another thought that just crossed my mind, perhaps the proverbial "sandbox" is true, perhaps google is taking all the sites that do not have a proven track record and putting them in serp purgatory, any webmaster whos just in it for the money will give up. Anyone who is commited will continue to build the site showing google it is not simply a spam site.

Any thoughts.

This recent change is affecting sites that have been around for a long time as well. Most of the people who say they were affected had sites that were 1-3 years old. So I doubt this to be the case. What you are suggesting is a little to overly focused on us too. Google has a world to manage, and affiliate/bh marketers online are a small portion of it all.

I can agree with MAD that this is an algo update effect that started a while back, but what of the factors that have creating this impact is what I think we need to come right down to.

It also remains to be seen if we have seen the final act of this rollout. What if we are at algo factor 190 and the results we are seeing are the computational results of that, and then factor 191 kicks in and changes it to something else? If I am understanding what MAD is suggesting right, then we may see another shift take place.

Overall though, I am becoming more convinced that this is a matter of:

• exact match on search required
• more elaborate content for that exact match
• excellent code elements incorporating again that match
• an element that offers some form of user contributed content/interaction
• the same old continued linking with matching ancho

What seems to be happening with a lot of sites is the loss of the relatively unrelated search traffic. The long tails are no longer kept by authority, but were tranferred over to what appeared to be more quality.