Google Traffic Throttling - revisited

[tedster]

The topic of Google traffic throttling came up a good number of times over the past two years or more. Some doubt that it happens - others are convinced but have trouble breaking through the wall of other's doubt.

Last year I saw the analytics for one site that showed a drop to zero Google traffic every day at around the same time - it stayed at zero until the next day and then it cycled again. I also saw another major enterprise site that only had first page ranking for a specific 4-hour period every day. Both of the cases persisted for many weeks, but both eventually "returned to normal".

Just recently someone else contacted me with the graphs to show something very strange. Three months of nearly level Google traffic (unnaturally level in my experience) followed by a huge spike that lasted 4 hours - I'm talking up 1,000%.

So I think it's time to look at this topic again - seriously. It's not a fantasy. Something Google does every so often looks like traffic throttling or rationing. Here's a post from just today:

backdraft: I get these "windows" of sales only minutes apart, then hours of nothing. It's as if Google can turn us on and off while still showing our site in the SERPS. That is very strange and an experience shared by other sites over in the UK. I am in the US.

[webmasterworld.com...]
Here's a small sampling of past and recent discussion:

1. Google Traffic Throttling - where are we on this? [webmasterworld.com]
2. Time of day rankings changes [webmasterworld.com]
3. The Yo Yo Effect - is it now getting worse? [webmasterworld.com]
4. Huge drop in my blog traffic [webmasterworld.com]
5. Google Toggles our High Ranking On/Off Again and Again [webmasterworld.com]
6. One day spike in Google referrals [webmasterworld.com]
7. Is there a threshold for google traffic? [webmasterworld.com]

And there's more where that came from.

It seems like it's a lot harder to pin down now, and it was never easy. So what do you see? Is anyone wrestling with this on their own sites? Any ideas how it is being accomplished?

[Tallon]

There will always be some greedy, scammy sob to mess it up for the rest of us forcing Google to tighten the system up once again.

Well I'm not so sure that this method of generating buzz IS a scummy practise (other than it's probably illegal in many countries since choosing the "contest winners" (if real) aren't moderated by an independent accounting firm). A site should be able to generate buzz right?

My point is that I'm very surprised google would give any real weight to social media links, comments, tweets, twitter followers, facebook fans, etc., considering how incentivized much of that stuff is (at least in some circles). Even comments on blogs, twitter followers, facebook followers are many times paid for.

If I say: I'll give you $5 to link to me, or $5 is yours if you retweet my article or leave a comment...or I'll link to everyone who links to me, etc...why would that be frowned on by google and not: If you link to me, retweet my post, subscribe to my twitter feed, I might pick you as my contest winner! What's the diff?

[micklearn]

Here's hoping that the traffic throttling and/or demoting pages in SERP's isn't reliant on or giving much weight to social media sites. They seem like an easy way to game the system...just like the article sites out there...

[wingslevel]

summing up, i am not sure both theories can't be right. there is pressure not to "do evil" and thus "spread the traffic around" - so, in accomplishing this, daily quotas are set for sites - a big component of the quota may be the backlink profile.

[jimbeetle]

I can't wrap my brain around a bad backlink profile being the reason to throttle traffic. Since the phenomena appears to be domain related it's apparently a piece of the domain trust part of the algo.

I just can't see the answer to the question...

"Hmmm, looks like this domain's backlink profile is a bit sketchy, what do we do about it?"

...being...

"Simple, throttle its traffic."

[IanTurner]

Also throttling traffic to specific domains isn't exactly going to be much of a long term solution either - all that will happen is that people will spend less time on the throttled site and more time on other non-throttled projects. I just can't see the reasoning behind it at all.

[backdraft7]

It's the Flintstones 50% anniversary and Google has another cutesy logo to commemorate the event - how appropriate, since they've pushed many of our sites back into the stone age.

[wingslevel]

jim, i think you are right about the trust part of the algo being the prime mechanism for throttling, but i think the backlink profile is a very big piece of a domain's trust signature.

[julinho]

People seem to be focusing on links, but what about the content of the sites?

Some patents indicate that Google is giving more and more weight to the way that users engage with a site.

If the "usefulness of a site" (as determined by user behavior/engagement and measured by Google's algorithm) doesn't increase, then Google may create a ceiling to the site's rankings.
Such ceiling won't be broken by more links or more age, only by more site usefulness; links can take you so far, but beyond then only with more useful content.

Increasing the amount of pages won't necessarily increase the total usefulness of a site (e.g., suppose such new pages only answer questions which had been already answered elsewhere - so, they don't add usefulness to users).

BTW, couldn't the old sandbox of 4-5 years ago be considered a ceiling to new sites, which had yet to show any usefulness? If and as sites proved to be useful to users, then, PageRank permitting, the ceiling would move up.

[tedster]

Very interesting ideas from a theoretical point of view - thanks for that. However some of the most powerful observations in this thread don't seem to line up with it.

I do agree that something like the old sandbox may well be the mechanism, but it's probably not the same target (new sites).

[IanTurner]

I had also thought that this may be something to do with user engagement - definitely worth looking at in more detail.

The problem being is if you get a lot of sites in the same niche which are similar the user engagement on all of them may be low but similar and thus lead to throttling of an entire niche, leading to real crap being shown in the SERPs - which may be what some in this thread have been reporting.

[jimbeetle]

backlink profile is a very big piece of a domain's trust signature

Very true. So let's try that again.

"Hmmm, looks like this domain's trust signature is a bit sketchy, what do we do about it?"

"Simple, throttle its traffic."

It still doesn't work for me. Why send traffic to it in the first place?

The still very interesting bit for me are drall's numbers: 36K per day referrals really doesn't reflect a site with many problems, if any. And actually, a million referrals per month from Google alone, tells me it's quite a strong site.

At first I really leaned towards dayparting as being part of what's being observed. But again, drall's numbers, smack on 36.5K referrals each and every day kind of belie that for me. Even with dayparting there has to be some variation in the numbers.

And I guess that's why I have a problem wrapping my mind around any of the theories that have been proposed. There are peaks and troughs in all Internet traffic, total search activity, search activity for certain topics, etc. But it shouldn't be reflected as it's reflected in drall's numbers.

[tedster]

Agreed - drall's account is worth a second and third read. If this were dayparting, we would see time based variance. Instead, we have this account:

Lately as I have been pointing out in other threads pages with social networking strength have been rocking for us.

So much so that some of these pages can bring in 3-4000 visitors a day. What is incredible to watch is after our visit count spikes for some page like that, other pages that normally perform [four figures of] visitors per day will all suffer just enough the rest of the day to level us out to that 36,000 number.

[outland88]

The still very interesting bit for me are drall's numbers: 36K per day referrals really doesn't reflect a site with many problems, if any. And actually, a million referrals per month from Google alone, tells me it's quite a strong site.

Yes even Google would argue the site is already receiving a handsome amount of traffic to begin with to dispute the theory.

Yet from all the comments especially the later ones I am satisfied people know what they are seeing and are sufficiently suspicious.

It seems to be most provable on a case by case basis but as whole it might not be scientifically provable without a research grant.

I don’t buy into the link profile thingy though. Google does what it wants to and unless you have a high page count with them its better to get between Google and your traffic in the coming months.

[drall]

Hi Guys, im not sure about the link stuff with the site. We dont go after links with this site, they are all natural. Some months we get hundreds, some months we get thousands.

In the last couple of years we have gotten many natural links from tier 1 sites. Think Microsoft, Apple type sites. We also get tons of links from blogs, hobby sites, government sites, state sites, the whole range I guess. The only really spammy backlinks I guess we have are from all of the scrapers and mashup fake search sites but once you get popular you just cant avoid that I guess. Maybe its backlink data but I dont know what we could be doing wrong. Heck most our anchor stinks because it is natural linking.

[wingslevel]

yes, just re-read drall's account for at least the third time.

for a pr8 site with 1 million + pages, 36,000 uv's per day doesn't sound like much at all to me. on a given day, google may be organically referring visitors to only one or two percent of his pages.

backlink profiles are immensely complicated, so we can't say for sure, but it doesn't sound to me like drall's cap is coming from a backlink problem. he says his links have come naturally and he makes reference to social network driven traffic spikes, so he is participating in that medium.

[drall]

Hi wingslevel, we dont have 1 million + pages, I think you confused searchs with pages.

We have 120,000ish pages built over 12-13 years. Just wanted to clarify that is all. Our staff makes around 20 pages per day. Mostly tutorial how to type stuff.

We really dont participate much in social networking. Sure we have pages setup at fb, twitter and such but most of what is happening is word of mouth.

Example would be Bob tell his fans to go to our site and it spreads through the social graph. Some days we have a couple hundred Bobs and friends, some days we have 10,000 of them. It's all totally random and natural.

[wingslevel]

should have read it for a 4th time....

[backdraft7]

@drall - just curious and please don't take offense, but is your monetization solely from adsense ads? or do you sell a product or service? Is your content 100% original or is it aggregated from other sites and if the latter, do you link back to the source? It seems those sites selling branded products or services are getting hit hard, while adsense based sites are showing either increases or a generous ceiling. Kinda makes me wonder what happened to the May Day branding filter. It's all a mystery, but over the past year or so, I have seen our searches being drown out by more and more MFA sites.

[outland88]

From what I have seen it effects both Adsense and commerce sites but more the commerce sites.

I have seen our searches being drown out by more and more MFA sites.

It is becoming epidemic in my areas. In the past two weeks I have seen the top ten results for eight keyword phrases go almost entirely Adsense. In fact with one phrase 27 0f the top 30 results were Adsense sites. It goes beyond merely coincidental. The vast majority of the sites are basically just assimilated from larger article sites. If you’ve seen three you’ve seen them all. It seems Google ranks them well for just being similar.

[tedster]

outland - I'm missing connection between your comment and traffic throttling

[drall]

backdraft7, we monetize the site via a combination of 70% direct ad sales, 20% adsense ads and a handful of direct discrete affiliate agreements 10%.

The site is purely an informational site with no direct sales of products. Most of our sites are the same structure as this.

[mrez74]

So if I'm seeing like big drops in traffic (i.e. 30-50% less) from one day to the next (and my rankings are the same) is this what you guys are considering Google traffic throttling? I've noticed this the past week or so where certain sites seem to have a great day in traffic and then the next day lose like 50-60% of that traffic even when the sites having the same ranking?

So what exactly is happening to that traffic even though I occupy the same spot in the SERPs? Is the ranking that I see not the same for everyone on those days where traffic takes a hit or what? I'm not clearly understanding what is that Google is doing that seems to take the traffic away? I mean it can't be that I'm getting a natural swing in search volume with people not doing any searches for that keyword on those particular days is it?

[backdraft7]

I still don't know whether to buy the 'throttling theory' or the 'still digesting data set theory'.
It's like rolling blackouts on the web. Thanks drall - I was just trying to see if there was some connection in site types. One site in the UK is experiencing identical outages yet is completely different than ours, and they have no adsense ads. I have one ad on each page myself. I know of one popular ezine site that gets away with up to seven! and they thrive! So it seems nobody is really immune for any one reason. Today's traffic seemed a little better, so far...

BTW, it's great that everyone has been sharing their experiences, opinions and idears (Pauly Sr). Thanks to all!

[jimbeetle]

So if I'm seeing like big drops in traffic (i.e. 30-50% less) from one day to the next (and my rankings are the same) is this what you guys are considering Google traffic throttling?

Not really, no. If you look at drall's traffic numbers (somewhere near the head of this thread), they are basically 35.5K referrals for each day of the week -- no variation.

What you describe, variations in referrals while maintaining the same search positions, would appear to be differences in search volume for the terms. Or, since you mention the "past week or so," possibly the effects of Google Instant and it's more agressive search suggestions.

You might want to check your referral stats term by term for, say, a couple of weeks before Instant and now to see if there are differences in referrals by terms.

[backdraft7]

@tedster - can you start a new SERP change topic for October? I'm seeing some interesting changes today with the little blue arrow. I think it's another personalization focus enticement. Also seeing some unusually good Friday traffic.

[tristanperry]

@backdraft7 There already is one:

[webmasterworld.com...]

[jimbeetle]

The little blue arrow = Google Instant to Get Keyboard Navigation [webmasterworld.com]

[mrez74]

You might want to check your referral stats term by term for, say, a couple of weeks before Instant and now to see if there are differences in referrals by terms.

Jim,

But I hadn't seen so much change in the sites when Google Instant was implemented. Like for several sites traffic is 50% lower today compared to just yesterday. I don't get the big shift in traffic from day to day.

[roodle]

I don't get the big shift in traffic from day to day.

That's exactly what this is about. Level(ish) traffic from one day to the next.

[Robert Charlton]

Several current posts on the Google Updates and SERP Changes thread [webmasterworld.com...] that should be noted here...

Posted 3:27 pm on Oct 6, 2010 (PST -8) by freejung...

I suspect we're seeing both a GA glitch _and_ a significant update at the same time, that's why we're getting mixed results. I've seen some results sets flicker in for a few minutes that were distinctly different on normally very stable SERPs.

Posted 3:29 pm on Oct 6, 2010 (PST -8) by scottsonline...

It appears we are seeing several issues, one of which is analytics related, one possibly related to peak hour instant or datacenter issues that occur at the same time, and also traffic throttling. It's impossible for us to say whether the destruction of goolge traffic from lunch until the close of business eastern time is one or the the other by here's a hint:

The moment traffic came thundering back analytics unstuck itself, conversions and traffic went back to normal? Is it possible analytics is part of the throttling equation?