Google Updates and SERP Changes - April 2013

[sunny_kat]

< continued from [webmasterworld.com...] >

I am seeing a lot of ranking drop. Lot of keywords ranking between 10-50 have disappeared from Google top 100.

Is anyone else also noticing changes?

[edited by: tedster at 11:54 am (utc) on Apr 1, 2013]

[taberstruths]

I had that happen from March to August of last year. It wasn't until I broke all the links heading to certain pages that I started to see recovery over the last month. Went from 2500 visitors a day down to 600. Now I am back up to 1200 per day.

[Sand]

I work on two different sites (one that is mine and one that is a client's), and they both saw really big positive movements on Monday (the 15th). GWMT shows a nearly 100% jump for impressions on that day for both sites, and search referrals are up by about 60% since then (averaged out between the two).

It feels like something big rolled through...

[Martin Ice Web]

taberthruths,

i think this time it is not links. Its a Panda refresh/algo update, also they Claim to have it in main algo, i don´t believe them. This fells, smells and Looks like a Panda and it is a Panda. But we never saw such a big hit like from thursday to now. User interaction has never been this bad, bounce rate is nearly 100%, it went up from a bounce rate about 25% before this update. Our best time is now in the night, at day time there is no traffic that is converting.

ecom, germany

[rj87uk]

Two things sound interesting from this thread.

1) Ranking based on if a website will be easier to use on different device sizes which I could accept, needs tested.

2) The reference to Psycho Pass.

[linkbuildr]

All this talk of user interaction although Google states they DO NOT use data from Analytics in their algorithm. Belief or not?

[Str82u]

It feels like something big rolled through...

I've seen movement in SERPs the last few days, below us mostly. Page 2's seem to have the most movement on them but mostly relevant listings.

Everything is fairly steady including the keywords that "breath" as usual.

EDIT: @linkbuildr - all the data they feed us in Analytics they already collect themselves, so in a way you could say they do use Analytics data. I don't think they use our goals, annotations or anything we set to monitor ourselves.

[Sand]

Everything is fairly steady including the keywords that "breath" as usual

Funny thing is I'm seeing the same thing. The majority of traffic gains on both sites seem to be in the long tail.

[taberstruths]

@Martin

This latest gain started for me on Sunday the 14th in the States. This is the second Sunday in a row that I have had an increase of traffic that lasted through the week. First week was around 10% the second week seems to be around 30%

Biggest noticeable change is that I am getting long tail again but I am noticing some exact match as well. I don't track serp positions anymore. I have over 1000 pages and so it becomes impossible to track.

If it was panda, then google is penalizing widget sites for using quotations from authority references off-line as well as online when writing the article. Example; Using a reference to chilton's when explaining car repair. (I hope that doesn't break the rules, but believe me when I say I am not a mechanic and do not have a car repair site.)

[diberry]

All this talk of user interaction although Google states they DO NOT use data from Analytics in their algorithm. Belief or not?

They don't need it, actually. If they're just tracking bounces back to Google, and whether or not that user picks another result (indicating the first wasn't the be-all-end-all for that query), Analytics wouldn't come into it. Also, Analytics data doesn't scale because loads of sites don't use it. If they shape the algo around Analytics, it would be skewed. (Ditto with Chrome data.)

[n00b1]

Google's John Mueller has stated in numerous recent videos that it doesn't make sense for Google to use 'bounce backs' as a signal. It's too noisy and seen as normal user behaviour rather than a negative signal. He gives examples of people wanting to find further information on a topic, reading other opinions and seeing things from a different perspective. If you read a review for example it is quite normal to bounce back to the SERPs to see if other reviewers/sites bring up similar positives and negatives. It doesn't mean the first review you read is poor.

[Martin Ice Web]

@taber,

it looks similar to my site(s) but i saw this movement weeks before. Long tail came back, very old pre panda terms came back ... it thought this is the break through but thursday, boom, all was reverted back to november panda traffic and terms. So my guess is, it is or was panda.

ecom, germany

edit: about the mention of links that you broke down.. i just found a very heavy error on my site that produced internal links 1-20 same links to internal widgets. Hope this was the failure what panda was/is not liking.

[thms]

Google uses Chrome user data, it's much better than analytics, comparing Analytics to Chrome data is wrong because Analytics doesn't include all sites but Chrome data includes all sites, just the user sample is not covered 100%

[diberry]

Chrome data includes all sites, just the user sample is not covered 100%

No, Chrome data can only include data about the type of sites visited by the sort of person who downloads Chrome and uses it. That tends to be a youngish person who knows a bit about computers, thus leaving out loads of Baby Boomers and the not-so-tech-savvy people that Google results also need to satisfy.

[thms]

diberry

But still chrome data is much better than analytics just because of the simple fact that it includes all sites. And your description of the typical chrome user just matches the description of internet usage as a whole.

And if you read the link posted on page 9 of this thread you can be sure Google is using chrome data, as ex-googlers said in an interview.

[diberry]

But still chrome data is much better than analytics just because of the simple fact that it includes all sites. And your description of the typical chrome user just matches the description of internet usage as a whole.

Not even close. Tons of internet shoppers who spend big money and surf a LOT (older, non-tech-savvy Boomers) don't even know what Chrome is, unless you mean the faucet. You can't leave them out and think you have holistic information about the web in its entirety.

I'm not saying Chrome data has NO use at Google, however. It does give them some user behavior and other metrics to play with, it's just a limited set that wouldn't scale up to indicate something about the entire algo. It probably does lead to models for testing, but it wouldn't directly shape the algo. That's all I'm saying.

[TheOptimizationIdiot]

I'm not saying Chrome data has NO use at Google, however. It does give them some user behavior and other metrics to play with, it's just a limited set that wouldn't scale up to indicate something about the entire algo.

I agree, but one thought I'm having about how they could use Chrome Data directly in the personalization of results.

For personalization, they don't need to know how everyone reacts to a site, only how a specific visitor reacts to a site, so the data doesn't really need to scale to the whole when it comes to deciding how to "tailor" the results for a specific user.

So, in my opinion, those using Chrome consistently could see more/better personalization due to the amount of data Google has wrt their onsite behavior, but beyond that, I think it's too limited in scope/use to "shape the algo as a whole" directly for the reason diberry indicated, which is scalability.

[graeme_p]

dilberry, have you got any numbers on typical Chrome users?

I can think of three Chrome users I know: one is 80 and not usually tech savvy (though he sometimes surprises me), the other is probably about 50 and not at all tech savvy, and the last is a teenage girl (not tech savvy).

Furthermore you can always adjust for an unbalanced sample if you know how it is unbalanced.

[Str82u]

CHROME is about 12% of our users. Been like that for months. Some of our sites it's slightly higher or lower but the average is 12% and our busiest site is 12%. On the whole it's #4 behind Safari, Android and IE, which is #1. FF is #5.

[HuskyPup]

Was there an update this last weekend and I've missed mention of it?

I have some sites where the traffic has more than doubled Monday to Wednesday and other sites where the traffic has halved but overall I'm down with Wednesday being one of my lowest days of 2013.

[petehall]

@HuskyPup I suspect there's been something Panda related going on because I notice one of my cleanest sites started to slip. This led me to some duplicate content URLs being caused by a scraper type site linking to us via an iframe. I've put yet another 301 rule in place to combat this.

This isn't the first time I've seen this happen now either. I will now need to go through every site and make sure nobody is linking to /folder/subfolder/?scraperjunkarparam and creating duplicate content URLs. SIGH!

[HuskyPup]

I'm seeing quite a few new 2013 registered domains all using WordPress/Blogger platforms immediately ranking on the first page of the SERPs using my text and at the top of images once again using MY images and blowing me out completely!

WTF is up with you Google, why can't you tell the original image creator?

[Martin Ice Web]

WTF is up with you Google, why can't you tell the original image creator?

Cause Google does not care!

[petehall]

Things are certainly starting to look quite messy. But then I find it's like this at this time of year, every year, without fail. Something big is no doubt on the cards.

The top sites in this niche all do blog spam comments or write articles which are unrelated (sometimes loosely related) to the topic with a targeted anchor to the site shoved in there at random. Meanwhile I sit here too scared to do anything like this to compete. I think if Penguin manages to detect these type of links the Internet as we know it is literally going to change overnight... wipeout!

[backdraft7]

Anyone unable to connect with GA this morning? Even my Adsense ads are taking forever to connect.

[netmeg]

I listen to Cutt video and no buy link but now lose. He liar in my eyes

Yes, so you've said. Over and over.

It's not enough to follow the rules.

[frankleeceo]

The rules recommended by Google are simply "best practice ((recommendations))". Most of the recommendations are great, but you will get screwed if you follow all of them to the letter. Reality simply does not match up well with theory.

Sometimes and often times, practice and theory differ! Totally innocent parameters and signals can have unforeseen detrimental effects for your business. What works in one niche / vertical can totally be devastating in another.

The best rules are the ones that work, the ones that make "your" visitors happier. Given that the thought and reasoning behind the actions can be justified. SE really comes in second.

Listen to no one other than your own experimentation in your own niche / sites and "your" own traffic stats. Never listen and follow blindly to another person on the net...especially a big company's spokesperson.

[diberry]

dilberry, have you got any numbers on typical Chrome users?

I'm never sure what I'm allowed to link here.

From 2010:

GfK Technology data shows that 36% of Google Chrome users are aged 16-24.

From 2012:

a survey I recently conducted shows that approximately half of Americans 45 years or older prefer Internet Explorer, with the remainder of senior citizens opting for Firefox, Chrome, Safari, or Opera, in that order.

As you can see on the chart, the percentages are reversed when it comes to their Gen Y grandkids — Americans between 18 and 24 years old — who overwhelmingly use Chrome. About half of them prefer Chrome, followed by Firefox (30%), IE (25%), and Safari and Opera (3.6% each).

I've read similar numbers from other sources over the years. On my own sites, I see that the sites which trend younger and geekier show more Chrome and Firefox use, and the ones that don't trend that way still use IE a lot because it comes installed and they probably never even think to try a different browser.

[TheOptimizationIdiot]

The rules recommended by Google are simply "best practice ((recommendations))".

Or "merely guidelines", which don't guarantee anything except a better chance of being included in the index. They don't guarantee traffic. They don't guarantee good rankings. They don't guarantee anything. They're merely guidelines that when followed increase the chance of inclusion in the index. Nothing else.

If a site doesn't rank where someone wants it to or how they want it to, then it's up to them to figure out why the algo is scoring other sites better and make adjustments to "fit" what the algo is scoring better at the moment.

Counting on Google to rank a site where we would like it to rank really isn't a very good plan. Much better to realize the rankings constantly change and to maintain rankings and traffic we need to be prepared to change what we're doing as it changes.

[petehall]

I've only noticed this kind of shifting through the day on somewhat low competition search terms - not big money terms. Does this line up with what others see?

I agree with this depending on your definition of a low competition term. This phenomenon was frequently occurring on a 3-word term returning 92 million results.

[Dymero]

I'm not entirely sure if this is just a test or something that will stay, but another SERP in my weekly check has gone to seven results. This SERP is brand-related but is not only the brand name, as is typical with the seven result SERPs. It's not even taken over by the brand's entities, like with the other ones.

I reported something similar a few weeks ago, and that too was brand-related. In this case, I'm a little irritated because our site was on the first page until now, but because of the changeover it's now at the top of the second.

This could be potentially problematic, particularly to e-commerce sites like ours. Will Google now change all SERPs that are related to a brand to seven results?