Yoyo effect traffic

[Globetrotter]

For the last couple of weeks I'm seeing some sort of Yoyo effect. Every time when reach some sort of traffic level my Google traffic suddenly drops. Usually by 10.000 visitors a month. Then after that the next couple of weeks I (slowly) get back the traffic levels again and the process starts all over again. Whatever I do I never get beyond this magical traffic number.

If I get extra traffic from other sources (then Google) it seems to be compensated. I've spoken to my SEO adviser about this and he is telling me this is a domain authority problem at Google's end. I also spoke other people about this and they all seem to recognize the problem but do not be able to know what to do about it.

Anyone has got a suggestion how to "fix" this?

[goodroi]

there have been many conversations about google and traffic throttling. i am a little confused with your post. you say this has been happening for a couple of weeks and then you say you lose 10k visitors a month. what is the specific timeframe?

[Globetrotter]

I've read some conversations about the google traffic throttling but my situation seems just a little bit different. As an example my traffic levels (number of visitors) look like this:

1500 1600 1700 1800 1900 2000 1500 1600 1700 1800 1900 2000 1500 etc

Each number represents about one week. But this “routine” is already there for months. I think it started when caffeine was introduced. But especially in the beginning of this period my traffic was much more unstable. For the last 2 or 3 months I can clearly see this pattern. (but again it sometimes changes just a little bit because of other influences)

[scooterdude]

do you have g analytics or adsense code on the site ?

[Globetrotter]

Yes both. And what's strange as well is that when my traffic goes down my earnings in Adsense go up or remain exactly the same. Which in my opinion is very strange.

[mhansen]

Could you possibly be in a market that has more search traffic towards the weekend, like football, or some other niche market that would see an increase in different times during the week.

I have a football site I am a part of... Tues-Weds, dead. Thurs, trickle. Friday... here they come! Sat-Monday, crazy-busy! rinse and repeat for football season, over several years.

Check Google insights for the trends in your primary keywords. I know I have been surprised by it before.

M

[jimbeetle]

The numbers don't look to far out of norm to me. Many sites exhibit different traffic levels by day of week. mhansen's example is tilted toward the weekend by its nature; other sites get more traffice during the week when people are at work and the boss thinks they're working.

[tedster]

1500 1600 1700 1800 1900 2000 1500 1600 1700 1800 1900 2000 1500 etc

Each number represents about one week.

So that's not tilted toward the weekend days, its tilted toward every sixth week? That is peculiar indeed. Some kind of statistical testing and traffic throttling does come to mind, but I'd dig into the stats pretty deeply before deciding on that explanation.

Are you seeing this behavior on any particular query term?

If I get extra traffic from other sources (then Google) it seems to be compensated.

So the numbers you are reporting here are for ALL traffic, not just Google search traffic?

[scottsonline]

Tedster you remember a few weeks ago we saw a spike on one term after a power user search? Every Friday since we have had orders for that product on Friday.

[Globetrotter]

I do have normal fluctuations in traffic during the week, as which I think is normal. I also know these traffic levels can change week by week. In Google Analytics I have an profile for each section of my website. Before these things happened I saw fluctuations per section due to algorithmic changes, some went up some went down, but now I see fluctuations at all sections mostly at the same time. (sometimes one section doesn’t lose its traffic straight away like all the others but really soon after that)

Or the traffic on all sections it going up again until I reach the “magic number” again and then I will lose my traffic almost instantly and it will repeat itself again.

To me the “magical number ” of visitors is more of a trigger then the time frame wherein it happens. I do have active and slow periods during this yoyo effect. So sometimes it may take 6 weeks sometimes only 4. This also depends on other traffic sources. But it will always happen when I reach that particular traffic level that I will start to lose (Google) traffic. And at the end of the month I almost have exactly the same number of visitors as the previous one.

Over the last couple of months I’ve gained a lot of new (external) traffic from Bing, Facebook and Twitter, also the activity on my website increased but I never got more visitors in total and if I seemed to get over this “magical number” I got less traffic from Google. The numbers I’m reporting are for all traffic not just Google, but Google is seems like acting like some sort of regulator for my overall traffic levels.

I also use Awatstats and when I check the statistics for 2010 I almost have exactly the same number of visitors a month (max 2000 apart from the average) . When I check previous years this never happened. It was always was low in the summertime lots of traffic during the end and of the year (which makes sense to me). 2010 has been way to steady.

My site has got mainly unique content + a forum and I do not aim (or track) any specific keyword. I’ve always been doing well just on long tail (and I still do, except from this problem).

[tedster]

Can you separate out just the non-Google traffic and see if that alone also shows a similar pattern, without the influence of Google in the stats? This would help you pin down if it really is a Google effect or possibly some strange technical thing that's affecting your site/server.

[Globetrotter]

I’ve done quite some research before I posted but I did have a quick look again, with your specific question in mind. Almost all of the traffic is showing a u shape line if I get statistics from January till yesterday. (High traffic in the beginning and end of the year, low traffic during summer time)

Except traffic from Google. This traffic is showing fluctuations per week with dips and spikes a big as 2500 visitors a week, but overall in a straight line as it seems at first. But when I look closer and I check my lower traffic months I see more Google traffic, and it’s starting to get less as soon as my normal traffic is picking up after the holiday season. So the opposite kind of graph.

From the 12th of June I see traffic from Bing triple (still low numbers compared to other sources) at the same week I see my traffic in Google declining. When you look at the overall traffic is remains pretty much the same. I didn’t checked the exact numbers just patterns.

My biggest traffic source is Google, so just a little bit more or less traffic from google makes more difference then all the other traffic sources together. I know this is also a problem and someday I will start a separate topic for that :).

Any idea what I can do about it?

[Globetrotter]

Guess what, this week i accidently made a mistake but putting the new Google Analytics code and the Google Website Optimizer on the same page. This caused my direct traffic to rise significant (which was incorrect but still). At the same time my traffic boosted my Google traffic dropped. Am I still the only one seeing this?

[Vamm]

would I be the first to suggest that you drop Google Analytics now?

[backdraft7]

@globe - I too have GA installed on my site and one adsense panel on the right side of every page. I believe this gives G more than enough feedback to regulate my site traffic. I also have been a $20/day buyer of Adwords since about 2001 with very few changes in that time frame. I just pay for a dozen of so major key phrases. Are you set up similarly?

The absolute strangest part is that when traffic drops off, my serps remain unchanged. According to my AWSTATS I am ranking for about 8,500 unique key phrases per month. They'd have to either drop me instantly from nearly half of those, then restore me instantly to account for this new, odd fluctuation. As tedster mentioned, it would be unlikely that search volume changes that much on their own.

Since about 1998, I have observed my traffic patterns which were previously a nice natural weekly sinusoidal pattern, with a seasonal summer dip yet growing each year. As of May 18th 2010, the pattern is disrupted regularly and has flattened. I suspected the economy is a factor, but my niche actually should do BETTER in a poor economy.

The new year usually starts with 20-30 sales per day, consistently with a slow decline towards April. It stated fine this year, but is now interrupted with days that have only 2 or 3 sales. It's not my dedicated server, (I use alert site to monitor it), it's not DNS poisoning, it's not a hacker, it's just "low T" (low traffic, not low testosterone - lol, coined a new phrase).

I'm not going to say G is being anti-competitive or socialist by "spreading the traffic around", but all evidence points in that direction.

[Globetrotter]

@backdraft7 I’ve got a banner on the right side as well and on some pages 2 extra banners. I’ve never spend money on adwords. So maybe we can rule out adwords?

Same here, de serps remain the same, even if you check with proxies. But when you check Google Analytics you see big fluctuations in total traffic and google seems to correct it. I’ve seen the pattern over and over again since my first post on this topic. But there seems nothing I can do to change it. May I ask the percentage of traffic you get from google? Is it more or less then 75%?

In awstats I’ve got 60.000 different search phrases in December same for October and November 2010. When I compare this to Februari March April (also 2010) I have the same traffic levels but 40.000 to 50.000 different search phrases. So it might be we get traffic of different keywords in the same total amount as we were used to.

I’ve never really optimized for any key phrase, but since last month I’m working on this one. So I cherry picked some keywords and start researching them. One thing I found strange was that for the key phrase I choose there was around 165.000 searches per month and low competition. For months I’m at the second position for this key phrase and I “only” get 200 visitors a month according to analytics. Some of the information but be wrong?

I was seeing the same traffic patterns as you describe until (I think it was) March this year. I do not think it’s because of the economy because in my sector the dip was already there because of other influences (think of global natural disasters etc). In March I did a major website update, where the internal linking part, design, html and navigational menu’s where radically changed (but I did not changed the urls) and all this time my (total) traffic levels never changed. This is again abnormal behavior if you ask me.

[fifi]

Hey,
I've been noticing REALLY crazy yo-yoing on our sites (informifi.com) since Dec 6th. Google shut down a number of our largest sites entirely, and for no reason that I can figure out. Then, some of our other sites saw massive drops in traffic, then massive gains.
What on earth is going on out there?
Here's a sample for one of our smaller web site (unique visits). This is for a site that has done very steady 4k-5k a day for its entire life, and we've seen the same issue with other sites, though mostly for the worse:

Dec 12: 17,000
Dec 13: 20,000
Dec 14: 5,000
Dec 15: 5,000
Dec 16: 5,000
Dec 17: 5,000
Dec 18: 6,000
Dec 19: 17,000
Dec 20: 27,000
Dec 21: 28,000
Dec 22: 8,500
Dec 23: 24,000
Dec 24: 22,000
Dec 25 - 26: No Data
Dec 27: 15,000
Dec 28: 38,000
Dec 29: 40,000
Dec 30: 42,000
Dec 31: 43,000
Jan 1: 48,000
Jan 2: 51,000
Jan 3: 70,000
Jan 4: 71,000
Jan 5: 65,000
Jan 6: 12,000
Jan 7: 11,000

[tedster]

Hello fifi, and welcome to the forums.

This kind of phenomenon is a very puzzling thing, isn't it? A good number of reports have been made here for over two years, and we still don't have a definitive answer. See Google Traffic Throttling - revisited [webmasterworld.com] for a look at the ideas we've been floating.

Google shut down a number of our largest sites entirely, and for no reason that I can figure out.

Good luck with figuring that out. You may find food for thought in the Hot Topics and FAQs area [webmasterworld.com], which is always pinned to the top of this forum's index page.

[fifi]

Thanks for the link. Our situation seems different than the issue presented in the throttling thread, since we have had our sites for a range of time between three years and younger, and have never seen anything but steady growth in traffic from google, up to the point where we did about 8 million visits network-wide in november. of that, 80-90% is from google. Starting in Dec, all sorts of chaos broke out. Google de-listed some of our sites entirely, but left other, nearly identical, sites running.
Then, many of the sites exhibited the wild swings that I describe above. My theory is that google is playing with its algorithm, but doing limited rollouts. The theory is that sites we have that are in one cluster of google's world get to be guinea pigs for the algorithm changes, with wild and unpredictable results.
I was wondering if other webmasters had seen any radical changes in their level of search traffic starting in late nov or early dec. Some of the yo-yo ing described above seems to sort of match our pattern.

[tedster]

One thing you can do is watch or participate in our thread Google Updates and SERP Changes - January 2011 [webmasterworld.com] - every month we have a thread that continues the discussion with people reporting on the SERPs changes they see.

There are some reports of post-holiday jiggling right now, but nothing major.

[Globetrotter]

There are some reports of post-holiday jiggling right now, but nothing major.

On January second my Google traffic doubled overnight. For me that is major :) It’s too soon to say if the pattern I saw before this date is still the same now, but I do recognize the pattern fifi is describing from what I had (but still not as big fluctuations in visitor numbers as he is describing).
I’ve no idea what triggered Google to instantly double my traffic.

I do make a lot of changes to my website all the time, but mostly for my users not for google.

@backdraft what is your status at the moment?

[mhansen]

I was wondering if other webmasters had seen any radical changes in their level of search traffic starting in late nov or early dec. Some of the yo-yo ing described above seems to sort of match our pattern.

I just want to preempt my reply with a disclaimer that although I am watching, checking, and paying attention to these patterns, we are trying our best to step away from any site tuning for organic Google traffic. If they send visitors great... if not, oh well. We have simply grown tired of chasing our tails and trying to be Google friendly. With that out of the way... you know my position.

Our own throttle situation...

On two sites we manage, each with very different profiles (different servers, owners, visitor profiles, etc), these two are experiencing a similar pattern. I described the newer of the two sites in this thread [webmasterworld.com] in Dec. After monthly log reviews on other sites I am involved with, I discovered the second site with the exact trend, and decided to watch it as well. We also manage several others that do NOT experience this same trend... so its hard to peg down.

When Google is sending traffic, they send it in droves... at other times, they send none at all, (except image search, which has never been throttled on these sites) Its literally "ALL or None" on these two sites! The first throttle I noticed was October 30-Nov 2nd-ish. After that, it happened sporadically throughout Nov, then a clearly noticeable "Off" in early Dec.

The latest throttle for me was:

- Jan 4th, 6:10PM UTC, Google-Pipe-On! (first Google referral at 6:10PM)
- Jan 7th 2:17pm UTC, was the last visit from a Google referral, Traffic-OFF

Its almost like a completely different dataset was being used for serps, and during the throttle, BOTH of the sites ranked well, which leads us to believe they are "bucketed". It's most noticeable to me on the newer, 4 months old site, since it's in a state of steady growth and I just see it more often. When Google spigot is on, its 5-7x the normal traffic and more easily spotted. (lack of other traffic sources due to its age)

Likenesses between them... The only real similarity I found between the two sites is a fairly thin backlink profile, and high concentration of social (aka spam signal) backlinks in each. (social bm's, tweets, article-directory articles, press-release type links, etc) Oddly enough, other sites have similar profiles and do not have the same throttling effect.

What are we doing to try and fix it? Nothing. We have already spent enough time and money on this with no measurable results. From slow servers might be a cause, (we spent $$$ for a dedicated box and hours of transfer and rebuild) to poor link quality, and 10 other gray area signals we've chased, we made a choice to simply stop. Add content, add value, let the cards unfold where they may.

I will revisit this post when the spigot turns on next time and mention it.

[walrus]

<quote>We have already spent enough time and money on this with no measurable results. From slow servers might be a cause, (we spent $$$ for a dedicated box and hours of transfer and rebuild) to poor link quality, and 10 other gray area signals we've chased, we made a choice to simply stop. Add content, add value, let the cards unfold where they may. </quote>

I am experiencing the same and I don't recall seeing so many posts like this over the last few years. Really seems like a whole new ballgame. A few sites i manage are getting trickles where there should be streams. Google has tweaked us out and their traffic is diminishing I would guess, people are googling less and less everyday.

2010 dec
2009 dec
= - 50% traffic
= - 90% earnings

[Globetrotter]

May i ask if all the sites we are talking about are very easy to "bucket" (I mean they are on one specific topic site wide?) ?

[tedster]

I think that's a good question to look at. The sites I've seen with yo-yo troubles are not so easy to bucket - often only the most generic taxonomy or category would fit, rather than having a tight focus.

[fifi]

can you clarify "bucket"? I think we have an interesting test case because we have a bunch of sites with content that ranges from very similar to very different but all in the same layout and design. What the interesting variable to me is what makes one site different from the other in google's eyes.

Why shut off technifi.com, but not sportifi.com? Why shut off celebrifi.com but start sending 70,000 visits a day to musicifi? If you look at them, you can see how similar all the sites are. I would expect all of them to be turned on or off at once, not random sites from within the network.
I don't know if some of my sites are "easy to bucket" or not, since I don't know what that means, but we make master sites that are broad (sportifi.com) and then use those to spin up sub sites that are more narrowly focusted (jerseyshorereality.com). Yet there's no rhyme or reason to which sites google decides to kill or send traffic to.

[mhansen]

I can only speak to the two sites I have access to. When I referred to bucketing of the sites, I meant that Google has put them into some bucket of its (Google bucket) own. The sites however are completely different in visitor profiles.

How Goog buckets them remains to be known.

[tedster]

This is, clearly, not an area where I can claim any solid knowledge. Here's the best understanding I have so far - but please take it as my opinion and not "fact".

Google most likely has several different sets of buckets (taxonomies). One important set would be for the query terms themselves. What is the user intention behind this query and that query? Is it informational, transactional, navigational? Is freshness important for this particular query? How about Geography?

Then there would be buckets for types of sites. Google's work would be to measure and match up the query to the right type of site. Or, in the case where intention is ambiguous or mixed, to find the right blend of results. Questions like "how many pages on this domain contain this query term" might come into play. Does this domain contain other pages that are topically related? In other words, search results today are not just a question of text matches.

The actual taxonomies Google uses are probably quite fine-grained, and not nearly as coarse as what I suggested above. The main point would be the importance of clear relevance signals on a site and each of its sections and pages. The muddier things get, the harder a job Google has sorting everything out. For instance, a home page with a ton of internal links that are all over the place semantically might create a challenge that requires a lot of testing on Google's side of things, resulting in a traffic roller coaster in both quantity and quality.

Here's a related discussion from a while back - stimulated by the QDF (query deserves freshness) idea: Blended Results, QDF and User Intention at Google [webmasterworld.com]

And here's a Google patent: Automatic taxonomy generation in search results using phrases [patft1.uspto.gov]

[Jonny6]

I dont think its any form of traffic throttling. The dates you provided are the most significant factor.

"The theory is that sites we have that are in one cluster of google's world get to be guinea pigs for the algorithm changes, with wild and unpredictable results"

"Its almost like a completely different dataset was being used for serps"

Agree with both suggestions, Google is using an old dataset from early December with some data coming from a new dataset. The new index is not being introduced gradually, its sporadic and evident of still being under quality testing. Once this testing is complete, it will be implemented permanently, just like mayday.

[mhansen]

The Google Yoyo traffic pipe turned back on for both sites I previously mentioned in this post.

My throttle history -

- Jan 4th, 6:10PM UTC, Google-Pipe-On!
- Jan 7th 2:17pm UTC, Last visitor from a Google referral, Traffic-OFF

Latest Activity - Throttle back ON

- Jan 12th, 5:40PM UTC (First Google referral in 4 days)

I will revisit when the pipe turns off to both sites.