Google ZOMBIE Traffic Observations

[samwest]

Mods Note: Split out of the October 2015 Monthly Observation thread into a separate Zombie Traffic Thread
---------------------------------
Sunday morning, usually banging away, but nothing but slow moving Zombie traffic, one or two at a time and sitting on the same pages for several minutes. Switch is currently OFF.
Still in decline. No sign of any seasonal upswing, which is way overdue.

[edited by: aakk9999 at 7:08 pm (utc) on Oct 19, 2015]

[FishingDad]

I have just deleted our Google analytics account, never used it much in 8 years of giving them vast amounts of data. What's the point of feeding them information to use against you, the least they know the better.

UK ECOM
3000 uniques per day

[Simon_H]

Could this thread please return to planet Earth? It has been 'revealed' that when you activate remarketing on your account, Google tracks the user and attempts to determine buyer intent. Yes, that's the whole point of remarketing. And BTW Google does this using conversion tracking page tags you add, not by itself. If this is new/shocking to anyone, then do as @samwest suggests and learn about using paid. This is unlikely to be related in any way to zombie traffic.

There is plainly a frustrated contingent whose logic begins with "I'm making less money than I have for the past X years", ends with "Google is evil" and the bit in between is just the latest conspiracy theory. The fact is that there are a large number of websites - not just big brands - that are very successful and are not seeing this phenomenon. You can either divert every conversation to "Google is evil" or you can help determine why our websites are being hit by this when others aren't.

[goodoldweb]

@Simon_H

The "zombie' non converting traffic and the lack of sales so many ecom owners are experiencing since about mid September is a direct result of google's tinkering with shoppers traffic using the methods described in part in the KB article. Hence, It is very important to discuss these issues in this thread.

Maybe you should get back to earth.

[netmeg]

But where are the winners?

Here. I monitor around 230 sites between my own and my clients, plus I run about $50k a month in AdWords spend. Nothing I can attribute to "zombie" traffic except for actual bots. Ecommerce and information and ad-supported sites; conversions are actually way up on all the ecommerce sites. On my informaiton sites, traffic is up as much as 40% this year, but AdSense is down - it's not because of zombies, it's because of mobile. Believe me, I have looked for this zombie thing for several years now, and I just plain don't see it.

It's not just me either. But the others I know of don't hang out here.

[ecommerceprofit]

nutmeg....as goodoldweb said earlier regarding your same response a few days ago, "Count your blessings. It will catch up with you soon"

you too may soon be pulled down to live among us...it's easy to look down from your castle above but not so much fun when it happens to you too...

Mobile is changing the World yes...but this phenomenon is different.

[Simon_H]

google's tinkering with shoppers traffic using the methods described in part in the KB article

What 'methods'? What 'tinkering'? The article relates to if you've turned on Google remarketing and added tags to identify your product, checkout, payment complete, etc pages, then Google will then use this to collect data on your users in order to determine if/which remarketing ads are shown to them. This isn't something Google does secretly; it's the whole point of remarketing/retargeting. The article doesn't apply if you haven't added the tags or use remarketing. If I'm missing something here, please explain non-generically how this would cause the zombie phenomenon. The rather important bit in between "Google can track users if you set up remarketing to allow it to do so" and "I don't use remarketing but see zombie traffic" would be nice.

[Jez123]

GFY .....and that means Good for You! Not the other one...lol. I haven't seen an upswing for going on two years now...and this is usually my seasonal upswing time. Hope it holds for ya Jez! Are you running Adwords or Adsense?

Lol, cheers @Samwest. I'm currently not running either. I paused my Adwords ads earlier this year as I was so disgusted with the return I was getting. Organic has been good to me this year on the whole. I run the occasional Facebook ad if things look like they are a bit slow. I offer a discount on those ads so I can see what returns I get for £ spent. FB seems to work far better for me than adwords.

[goodoldweb]

@Simon_H

Obviuosly you missed the whole point of my argument.

In the said kb article google are clerly admitting that shoppers intent is being very closly tracked via thier multiple spying channels. It takes just half a brian to connect the dots and realize what so many of us who have been in this business for so long are already strongly suspecting...

Its real man... they've taken the art of deception to a whole new leavel.

[ecommerceprofit]

Jez123 + - what you are seeing is different than the point of this thread. I for one am not debating on adword returns in an overall sense...this is not what is being discussed really...we are discussing what has been happening since Sept - the up and down, etc.

[Jez123]

Jez123 + - what you are seeing is different than the point of this thread.

Nor am / was I. It was merely background info and in reply to what Samest asked. FYI I did take a huge dip in conversions during about a 10 day period mid October that was very unusual but the following week the conversions returned and have been seemingly OK since. Given that it's a slowish time of year for my niche anyway (but the dip was way slower than normal).

[netmeg]

it's easy to look down from your castle above

And that is why you don't hear from the "winners"

Good luck; I'm out.

[Yukko]

Switching off your GA or quitting Google Services at all does not help. Google identifies customers (your visitors) and gives them a buyer score, not owners of the websites. If you quit GA, Goolge has lots of data learned about your website.

Would you like to cross the road with blind eyes or just see with your own eyes that the car is coming? Analytics is your eyes. It does not affect (cars) anything in terms of ranking.

Google identifies visitors by hashes. Again, you can find the proofs inside your Adwords account.

Another point about Adwords:
It is not related to organic search and is not a ranking factor, but it is a proof of the technology level they use inside the company. If you clearly see you can remarket to chosen individuals, be sure those individuals are well known for all internal algos.

Visitors are identified by hashes. You can make your own hashes valid for Google "scripts" based on email ID.

There are other methods to identify the customers apart of hashes.

[iamlost]

Much of the concerns discussed on WebmasterWorld appear to be from webdevs relying significantly to totally on Google referred traffic and AdSense revenue; however they affect (some of) the rest of us as well and observations from the less affected may be valuable as edge cases. Or not.

Note: when speaking of my own experiences my sites are all evergreen information with occasional urgent/critical news updates; Google traffic is 15-35% of total, overall average of 22%, has risen every year; revenue is direct ad sales, affiliate pre-sell, with AdSense as default/filler. Also, I have not been noticeably adversely affected by any algo changes nor received any penalties.

1. traffic quota/shaping/smoothing/throttling.
The definition appears to be that there are upper traffic referral thresholds applied to sites; the limit can be met early or late but when reached traffic suddenly stops or slows to a trickle for the rest of the day. Some suggest that it is not only held down but on occasion boosted up with the boost being a poor match (see zombie traffic).

I haven't seen such a sudden stopping, traffic varies day to day and through the day but without sudden on/offs. However, I have noticed what I call traffic plateaus: while varying it tends to stay within a range before taking a jump up to a higher range where it stays for some time before again jumping up. Further, I can usually predict a jump as there will be traffic spikes in the month or so preceding. Jumps are irregular, once in two years, thrice in one; have not been able to identify a trigger (my sites have been pretty well complete and in maintenance mode for years).

2. zombie traffic.
The definition appears to be repetitive dramatic sudden on/off shifts in traffic conversion while traffic volume remains constant and other revenue channels see no similar conversion change.

I don't follow AdSense closely, it is pretty much on set and forget unless a (daily) warning is triggered (eg: Active View Measurable < 99.9%, Active View Viewable < 80%, CTR/CPM is up/down > 10%). Further, it has retained a pretty static conversion rate over the years.

That said I also parse my log files regularly and when I first saw zombie traffic mentioned here some years back I ran Google traffic against conversion drops (and dramatic shifts in click tracks) and did (and continue to see) see a recurring result: the IPs of the traffic suddenly shifted from preferred geo-locations such as US, Canada, UK, to less desired such as Asian, African. Time was not a trigger, the shifts were sudden and massive at varying times of day, and then shifted back. Again, I've been at a loss for what the trigger(s) might be. Or, once triggered, what resets it.

I can see that if a site was subjected to such a shift from qualified to unqualified traffic that lasted longer than not it could be catastrophic; further that if traffic from targeted geo-locations shifted from mostly transactional to mostly informational for extended periods it could dampen eCom sites significantly.

On a side but possibly related note: for the most of September and October when I did searches on gdotca (only personalisation being location and en-ca) the first page was, more often than not, half or better non-English cctld European sites. So far this month that has stopped.

I will not get into the AdSense, retargeting, or even revenue generally as that is a conversation for another sub-forum except as a metric for quality of traffic. The kicker in this Google traffic topic is that we are NOT discussing changes due to a drop in traffic, rather a drop in the quality of traffic based on a corresponding drop in conversions as well as associated perceived and documented changes in the origin and behaviour of referred traffic while the traffic volume remains relatively unchanged.

I read a piece a decade or so ago that suggested that in Google's attempt to index the world's websites/knowledge it was actually becoming not the new librarian but the the new trash collector, that sooner or later Google would become GIGO. I have to wonder if what I had thought satiric was instead prescient.

There is a definite somethings going on. Identifying a cause or causes is about all that's left. :) Until then there is only test and try and test and try to see what works. In other words: business as usual.

[ecommerceprofit]

iamlost + - great analysis - thank you

[Yukko]

One more thing. I clearly see not only zombie traffic waves, but also the waves of toll fraud, the waves of charge backs or waves of different kind of finance fraud (not only credit cards).

Edit : also the waves without final conversion based on the Goal known only to me even if people paid their money to get to the restricted area where that conversion should happen.

[ChrisWilson]

That's an interesting observation Yukko. My site never gets chargebacks, but I received 4 in October. This was way out of the norm. Interesting you bring that up, something for me to look into more.

[mrengine]

the waves of charge backs or waves of different kind of finance fraud (not only credit cards).

I've seen an increase in refunds, but no chargebacks. I'm assuming these problems are a side effect of poor traffic quality.

[Yukko]

You're able to make experiments with your sites to check if Google manipulates organic traffic and if it divides people into categories by simple A/B tests.

Add a simple sign up or squeeze form at the bottom of your page, somewhere in the footer.
Mark signup completed page as a goal. Make no sales when people reach this Goal, just say thank you to them. The visitor should just read the text before signing up, fill the form and read some 'secret information' and thank you message.

Measure the conversions for your organic traffic. You will see a very low figure, because the form is not optimized for conversions. The form should be unoptimized for conversions, but still bring your some conversions a day, let's say 5 fake signups.

Then start marking your organic traffic for remarketing until you have 5000-10000-20000 people marked.
Start a remarketing campaign with the predefined conversion to complete that signup form, but make your home or any other common page as a landing page and do not promote signing up on your website through that form in the text of your ads. As a result of reading your text ad a person should read an article on the page and that article should not promote the form in your footer as well.

If you're lucky (and rich, you will need a lot of money for the experiment) you will see rising conversions for that form. You can make some tricks to silently filter out the organic traffic which is 100% not interested to fill anything and mark all other visitors as willing to fill. So you can save some budget by remarketing only to them.

What does it mean for us? It means that Google is definitely manipulating its organic traffic.

After you see the conversions rising for the fake signup form then make a trick. Increase the visitbility of that senseless signup form and on the "signup completed page" give some real bonus (coupon) to the people who completed the form and check if they use the bonus (give away something feasible a product, service or gift), so start being a marketeer and try to get the real conversions, but do not report bonus usage through offline conversions upload form. The results can be even more interesting ;-) The conversions by GA and Adwords will rise naturally, because you've increased the visibility of the form, started to mention that form in the ad text, but the final conversion (bonus usage) will never happen or happen rarely.

The chain of traffic acquisition will be like this:
Google increases the amount of organic traffic to increase the amount of people marked for remarketing which has as a final goal the fake sign up form. People completing this form will have no clear intension to use the results of their acitivty which is proved by giving these people a feasible real bonus which has noticeable value and can be used online right away after they get it.

It will be real people. No bots. Real emails. You can collect their phone numbers and even call them.

Sorry for my English. I'm not a native speaker, but I hope you are still able to read the post and understand me.

[aristotle]

Here is a theory:

-- For most sites Google is now able to use its data collection power (Chrome, Analytics, etc) to determine the site's approximate conversion rate for
a "normal" mix of traffic.

-- After determining a site's approximate conversion rate for normal traffic, Google can compare that with the conversion rates for other sites that sell the same or similar products.

-- A relatively low conversion rate is taken as a signal that something about the site causes people to mistrust it.

-- So Google begins to send poorly-matched (Zombie) traffic to these apparently distrusted sites.

-- Periodicly Google sends some "good" trraffic to these sites to retest their conversion rates. If a site fails the test, then Google returns to sending Zombie traffic until the next test period.

So in summary, Google has started using conversion rates as a measure of a site's trustworthiness, and sends it good or bad traffic accordingly, with periodic retesting.

[Simon_H]

@Yukko Just so I understand... You're alleging that Google intentionally increases the organic traffic to a site that uses remarketing, in order that those users will get the remarketing cookie dropped on their machine and then get ads served to them on partner sites that Google can charge for when they click? Is that what you mean?

If you've done this yourself and have the evidence, could you please publish the (anonymised) results?

[goodoldweb]

@aristotle

Nonsense mate.

Your theory does not explain the sharp sudden drop in sales starting around mid September over 5 different personal ecom websites. Plus an Ebay store that suddenly went dead quiet ever since (yes eBay are feeling the pinch too).

[samwest]

@aristotle:

It's a nice theory, and I agree with the first part about data collection and throttling, but the problem is this: When they send me "good" traffic, it converts like mad, so following your hypothesis, I should be improving. I think there's much more to it, but you do get a cookie for creative thinking. ;)

In 10,000 sales I haven't had a single return or complaint. I'm also a 11 year BBB member with an A+ rating. I WISH trust was more of a factor.

[Yukko]

@Simon_H you have your own websites. You can make your own experiments. If I'm wrong or my cases are too much unique I can live with it. It is not too hard to follow the steps described above. I've told too much already.

I understand that it is nonsense to burn the budget just to prove something, but you can be smart to mark your audience which is very likely to convert (assuming you know you business and website better than Google) so you will not burn too much. When you remarket your organic traffic and this traffic converts without any common human sense, but definitely having an arteficial pattern you feel a little bit amused... On the moment you give these people a real bonus, but they do not use it, you're feel even more amused.

@goodoldweb my best guess is that automated A/B tests are a part of Panda algo. eBay should be checked for bad quality thin content as well.

After determining a site's approximate conversion rate for normal traffic, Google can compare that with the conversion rates for other sites that sell the same or similar products

Websites with engagement less than everage in the industry... and all other things compared to your industry.
Pay attention to the Benchamarking tab in the GA. BTW all or the majority of ranking factors are official GA and GWT reports.

And one more thing here. Your converting traffic can be equal your real market share.

Just forget about your sites for a moment and think of the market share. In case there is no organic traffic, but still there is the Internet Amazon is #1 in their niche, eBay is #1 as well. It will be a big surprise if some amateur website outranks them by some reason and has more organic traffic than those monsters who are still #1 even if they have no organic traffic at all. I cannot imagine this, because such a situation is not logic.

This is how your ceiling is defined. If your marketing team is strong you multiply organic traffic by engaging visitors and have them returned from other sources. In this case your market share is growing and you have more organic traffic at the end. If you waste your new customers and just let them go, then your market share is low, you cannot serve people, nobody will send you more customers.

How to check if something has changed on your side? Send waves and benchmark.

[Yukko]

That's an interesting observation Yukko. My site never gets chargebacks, but I received 4 in October.

Check your customer support for non buying idiots asking hundreds of questions. You will see waves there as well. I'm serious. Just start to mark new customers' requests with some marks:
good question, no intension to buy
idiot's question, intension to buy
idiot's question, no intension to buy
and chart everything. You can see patterns.

[FishingDad]

Here. I monitor around 230 sites between my own and my clients, plus I run about $50k a month in AdWords spend. Nothing I can attribute to "zombie" traffic except for actual bots. Ecommerce and information and ad-supported sites; conversions are actually way up on all the ecommerce sites. On my informaiton sites, traffic is up as much as 40% this year, but AdSense is down - it's not because of zombies, it's because of mobile. Believe me, I have looked for this zombie thing for several years now, and I just plain don't see it.

It's not just me either. But the others I know of don't hang out here.

All that work looking after 230 websites and manage a $50,000 per month advertising spend, yet you find time to post an average of 24 posts per week for the last 10 years into this forum? :-/

Sorry, but this is the reason we do not out source.

Companies usually employ teams of people to look after only that one companies interest.

[Simon_H]

@aristotle We're a Google Certified Shop/Trusted Store and we see massive variations in conversion rates on Shopping. So, as @samwest says, really good to get the ideas on the table, but I don't see how trust can play a part in this.

[Simon_H]

@Yukko People often appear on this and other forums, claim to know Google's secrets, provide no evidence, say they've 'said too much already' and then disappear. What you're saying is interesting, but you can't expect us to try your expensive experiment because you won't provide any detail.

I like evidence. I'm very happy to be proved wrong, but there are 2 reasons why I believe what you're saying can't be correct:
1) I can't believe Google would make it that easy to prove/observe that they manipulate organic traffic in order to generate 'fake' paid clicks. If it were that easy to prove, we'd have read about it. I haven't and as far as I'm aware, it's not mentioned in the EU or CCI investigations so the complainant investigation teams haven't seen this either.
2) It would make no sense for Google to do this. The argument that Google does this to make money out of fake/non-converting clicks is flawed. People with low CRs/negative ROI reduce their bids or drop their campaigns, resulting in Google losing money. Suggesting that Google intentionally sends everyone fake/non-converting paid traffic for financial gain is like suggesting Apple would sell iPads that don't work to make money.

If you believe what you're saying is true, could you please provide some figures from your own site?

[Shaddows]

When they send me "good" traffic, it converts like mad, so following your hypothesis, I should be improving.

Depends- does it convert better with you than elsewhere?

I mean, if these are the qualified verified buyers, just looking for the best place to put their cash, G really really wants to serve the best, most trusted, most converting site they can find.

Any non-useless site will convert these people in droves. Every site will see these as the real meat, and (relative to the normal fare) believe they are doing a stellar job converting them.

Now, to be clear, I'm not saying your site is not the best. Nor do I fully buy Aristotle's theory that Samwest was critiquing. I'm just pointing out that split testing requires knowledge of the OTHER sites for comparison, not just knowledge of other traffic cohorts from your own site.

BTW, sceptics are essential for any informed discussion- it reduces herd behaviour. So it is unfortunate that the "your time will come" comments aimed at sceptical participants have detracted from the general improvement in quality of this thread

[Nutterum]

@Simon_H - actually one of the properties I monitor is just like Yukko. The difference there is that the traffic is highly seasonal. However I`ve seen Google Trends changing because of the remarketing efforts created by that website. So, Yes there is a pattern. I can't tell whether Google sends more organic traffic or the website created additional search intent that fell back on to the website, with visitors that are already more engaged than their previous visit, but I have seen what he/she described first hand.

[glakes]

The argument that Google does this to make money out of fake/non-converting clicks is flawed. People with low CRs/negative ROI reduce their bids or drop their campaigns, resulting in Google losing money. Suggesting that Google intentionally sends everyone fake/non-converting paid traffic for financial gain is like suggesting Apple would sell iPads that don't work to make money.

It's not a flawed theory and could be a short term strategy. It's right before the busiest holiday shopping season of the year, and for many businesses they do 50% of their yearly income or more in November and December. Comparing the current zombie/bad traffic to Apple selling iPads that do not work does not make sense. At least with an iPad you have something physically in your hands that can be proven does not work. On the other hand, how is one going to prove that Google is using bad traffic to earn more money?

Ask yourself why Google is shaping traffic at a time when the holidays are right around the corner. There used to be a time when Google publicly stated they avoided pushing algorithm updates around the holidays to minimize the disruption to ecommerce. Those days are long behind us, and this current trend of zombie traffic feels more like a money grab by Google to me.