Google Referrals Way Down Since the SERP LayOut Change?

this is not about the algo update - this is about the design update

[Brett_Tabke]

The last week was the biggest single change in our referral profile graph since florida 2004.

Anyone else seeing major changes in referral numbers since the new serp layout last week?

(this is not about the algo update, but specifically with the new serp layout)

[gford]

Yes this problem has been ongoing for me since March 15th and it has *sort of* leveled off in the last 4-5 days at an awful 50-60% loss in Google traffic.

This has, so far, only hit one of our sites but it was the most authoritative site with the most uniques which seems to be the trend of who gets clobbered as discussed in the mayday thread.

[steveb]

"Anyone else seeing major changes in referral numbers since the new serp layout?"


He didn't ask about webmasters losing traffic weeks or months ago. Can we please leave that unrelated stuff in that complaining update thread.

A completely new thing is the busier formatting of the new results pages.

I've seen a small change since the formatting was rolled out, about a 5% drop, but I would say that such a fluctuation falls in the normal range. It may turn out to be a permanent results of the busy layout though.

[aristotle]

I'm confused. Was there an extra-big rankings shakeup at the beginning of May, or was it just a continuation of the rolling shakeup that had already been underway for two or three months?

[Brett_Tabke]

nuked a bunch of side topics. This thread is about the new serp layout only. Not the algo update.

>March 15th

not what we are talking about in this thread.

>continuation of the rolling

Yes, this thread is specifically about the new serp layout.

aka, the new layout is about keeping people on Google longer.

[Freedom]

You're right. Sorry. Read too many of the other posts before posting.

[aristotle]

The last week was the biggest single change in our referral profile graph since florida 2004.

Anyone else seeing major changes in referral numbers since the new serp layout last week?

(this is not about the algo update, but specifically with the new serp layout)

I can understand how the new serp layout could cause a noticeable loss of traffic. For example, the 5% drop mentioned by steveb seems reasonable to me. But I don't see how it could cause "major changes".

[TheMadScientist]

Traffic on one crumbled today...
About 25% of usual for a Sunday.

But I don't see how it could cause "major changes".

How about when they put a messed up image of the site in the SERPs rather than an accurate image or screen shot? I could see this as being one way at least.

A site I run that looks great when you visit it, but the image they show is an enlarged, blurry portion of a navigation section... (It's the site I'm referring to above.) The image they show as 'the site' is a snip of an image, and not even the full image at the correct size / resolution.

Honestly, I probably wouldn't visit it if I saw the inaccurate representation of what the site actually looks like, because it looks totally unprofessional in the SERPs even though that's not at all the case when you visit...

[steveb]

the left side "Any time... latest... etc" list really, really, really benefits disposable blog-type sites over more mature ones. If something has not been updated in the past week (or day or whatever the range), it won't show up. On the other hand, even just changing a word on a page will get it listed.

So, if a page normally ranks #15 for a query, it can stay at #1 for "latest" just by making trivial pages changes and constantly getting the bot there via pinging or however.

And what the heck... "wonder wheel"?

[dickbaker]

Google has consistently been 44% of my referring sites. It's been rare for it to drop much below that, and just as rate for it to be much above that.

For the last several weeks it's usually been higher, going to 46, 47 or even 49%.

The odd thing is that my traffic has declined to 2007 levels, which I can only attribute to a combination of the nice weather and less spending/interest in my niche. I doubt that Bing, Yahoo and the others would drop as a percentage of referrals just because Google is making changes.

[Asia_Expat]

I'm not seeing any changes at all with my traffic from Google, that have been reported by so many in the last week or so, other than the very slow but very steady growth I've always seen. Maybe I'm doing something wrong.

[kidder]

Wonder wheel is just another way to keep users clicking on Google pages, more impressions mate..its actually pretty good for long tail keyword research. As far as traffic goes we are down on numbers here in AU where this change was rolled out earlier. I don't think it's a massive drop but I need to take into account seasonal factors and so before I can be sure. Its funny but my wife and kids started complaining instantly about the new layout and were very quick to bring it to my attention. As everyday users they will never go near the LHS column like most people IMO. Radio buttons are much better but of course they don't generate impressions, Google are just trying to do what I would try to do, keep the users attention for as long as possible.

[tedster]

And if you stay on Google long enough, playing with enough combinations, you may see the information you want in a snippet and not need to click through to any website in the results. That's one possible reason that some sites are down in referrals. That and the added visual clutter(?) which can't be good for a world with so many attention deficit challenges.

[walkman]

My traffic is very much down too compared to last week, especially today.

Brett how do we know Google didn't change something else in the algo too at the same time?

Maybe it isn't entirely the interface

[ken_b]

I hadn't noticed this before, but there doesn't seem to be a consistent number of organic results on the first page. A few I checked just now have only 8, others had 9.

Is that new?

But yes, referrals seem to be off a bit less than 10%. Not sure I can blame that only on the new layout though. Too soon to make any definite conclusions?

[edit typo]

[edited by: ken_b at 4:15 am (utc) on May 10, 2010]

[arizonadude]

My referrals are up about 10%.

I don't think the new layout is affecting my niches at all and my niches are shopping related.

I find it odd that it would affect webmaster world referrals since my thinking is it would affect someone looking to buy something instead of looking for information.

[walkman]

OK so I used it a bit and automatically started to ignore the left after a few tries, just as I ignore the ads. I can't be the only one :)

My guess is that the new template is not to blame for drastic drops, maybe 10-20%, which is nothing to sneeze at but as people get used to the interface, it will be ignored.

[aristotle]

One of my sites may be a victim of this new layout. Google traffic has dropped about 12% even though rankings for the major keywords are essentially unchanged.

I think the main culprit is the "Latest" option in the menu. My pages that are in the top 5 for normal search disappear when "Latest" is clicked, replaced by blog posts and newspaper articles. I wrote my articles 2-4 years ago and have never altered most of them, so they are not "Latest"

[JoeSinkwitz]

Estimated CTR not measurably down post 3-column layout push; we're still pouring through lots of data since we run lots of internal tests for conversion purposes, but the initial results point to the change not mattering much just yet for fat head terms...more data needed to see how it affects longer tail traffic.

Perhaps the lack of effect has more to do with the amount of garbage injected when looking for recently updated material vs "all time" OR that the commercial intent calculated by Google being less accurate than Bing -- the new layout feels like a Bing knockoff without a much substance to it...a bit noisy.

Last bit of info I'm going to look at closer: the organic section looks to be pushing results down .5-1 position below the fold (haven't "measured" what the actual screen squeeze is yet). I'm expecting longterm that the CTR will shift a bit more to position 1-3 since #4 is really getting buried (and not just on maps/universal integrated serps).

[randle]

The last week was the biggest single change in our referral profile graph since florida 2004.

Well its going to take a bit of time to study impression and click through rate numbers to prove guilt. But, if your experiencing the biggest single change in referrals right on the heels of the biggest one time change in the Google results layout functionality, you have a pretty good suspect.

This injection of search functionality on the left hand side is a monumental change, a bit surprised theirs not more discussion on it.

[bwnbwn]

What was the date the roll out started May-5-6. When this did come out all I thought was another way to keep a user on Google this does seem to be the case with our sites as well. Really not enough data being only 3 days with a weekend there so I will be looking at posting numbers soon, but from the looks of it yes we are seeing a decrease even with a top search terms at #1 spot after caffeen.

[ddogg]

Keep in mind it was Mother's day yesterday, could affect traffic. So far I am not seeing any difference (including traffic to my forums), but it's pretty early to say for sure.

[Chrispcritters]

Have you compared positional click through rates across differing date ranges in WMT?

When comparing CTR for several high volume search phrases (to my site) I do not see any change in the CTR for positions 1, 2, or 3 when looking at date ranges a month apart. Impression #'s appear to be too generalized to see anything there.

[Tallon]

How about when they put a messed up image of the site in the SERPs rather than an accurate image or screen shot? I could see this as being one way at least.

A site I run that looks great when you visit it, but the image they show is an enlarged, blurry portion of a navigation section... (It's the site I'm referring to above.) The image they show as 'the site' is a snip of an image, and not even the full image at the correct size / resolution.

I see this as the biggest problem...that and sometimes having a blank box instead of a picture showing (which could be my ISP I guess).

Seems all my results have a big "blob" of the article image showing in the preview option instead of a full page view like most of the results.

What I'm worried about: clickthru rates will be affected which could affect rank positioning. What causes google to show the "enlarged blob mess" rather than a nice page view like other results? Should I block google from crawling images?

[TheMadScientist]

Hmmmm... Thanks for posting Tallon.

After your comment about it being your host, I thought about it for a minute and just removed the .css block from my robots.txt... Couldn't imagine why they would need the layout of my site or the .js files so I've been blocking them for at least 5 years without an issue, just because I don't like giving them any more than I have to and it saves a bit on bandwidth. Until now they haven't really had any good reason I can think of to need to use either, but maybe they couldn't get a decent screen shot or figure out where things went because it was blocked? IDK, but it makes a bit of sense, so I'll let 'em in and see if they can do a better job with the image in the SERPs.

[Tallon]

oho! Interesting TheMadScientist, I too have blocked my css in the robots.txt...or used to ;). Now let's hope that fixes things! Thanks for mentioning it, I likely never would have picked up on that :)

[pontifex]

Basically I disagree to split this Layout effect from the Algo updates...

There is soooo much going on on Google that it is very hard to carve out the reasons for traffic changes:

1. I see the search parameters now sometimes added as "...#hl=en&q=phrase" which totally annoys me and makes the traffic impact for single keywords unmeasurable unless you use G-Analytics - bad move - really bad!
   
   
2. I started to query via proxies in D.C., CA, and Chicago in the US and 2 locations in Germany. Every time the whole SERP mutates like a techno video.
   
   
3. The new layout moves the results around which must have an impact on AdWords clicks, too
   
   
4. The MAYDAY/long tail thing just adds to this confusion!
   

Now that said: even as an experienced SEO it became damn hard to root the reasons for traffic changes to one single cause, if you did not note EACH AND EVERY position for the keywords which brought you traffic over the last weeks!

That means: has anyone noted down EVERY referer keyword to their site over a few weeks and the position of it?

P!

[physics]

I think google is really pushing the copyright infringement envelope with the "Page previews" feature.

[Brett_Tabke]

>traffic changes to one single cause

When they tie to a specific event at a specific time - they are quite easy to locate. No change in referrals during the so called 'may update' (we have seen zero change - as far as I can tell - there wasn't an update), and then stuff happens the morning they roll out the new layout. 1 to 1 correlation = causation.

> "Page previews"

yep. But how is that any different than 'cached' page.

[BillyS]

Brett how do we know Google didn't change something else in the algo too at the same time?

Have you compared positional click through rates across differing date ranges in WMT?

BINGO

If you have a high volume keyword that hasn't shifted position across this timeline, you can see a CTR pattern using the above approach.