Google Analytics Installation Followed by Drop In Rankings?

[BobMM]

My site has been steadily going downhill the last several years despite steady improvements to the quality of the site before and after the drop started. I assume a lot of this is due to panda/penguin issues.

However, the initial and perhaps largest drop seemed to coincide immediately after I *started* using Google Analytics. Is it possible that once Google obtained increased visibility to my site that is what caused them to knock me down? For example Google may not have known my bouce rate before I enrolled in GA, but they do now.

Thanks,

[tedster]

Hello BobMM and welcome to the forums.

I've heard such suspicions from many website owners, but many, many others have installed GA with no ranking change at all, even though their bounce rates and so on were very poor. I'm pretty sure this can only be a coincidence.

I doubt very much that Google would use GA numbers to affect ranking when they cannot use the same numbers consistently across all sites they need to rank, thus comparing "apples to apples". That would be a very poor approach, mathematically.

[BobMM]

Thanks for the speedy response.

[ZydoSEO]

Hello BobMM. Welcome!

I have to agree with Ted. Though many sites might use GA, many still do not. This would be a very poor choice for a ranking signal since they do not have access to the same data for a large percentage of sites on the web. Chances are slim to none that any data that might be collected by GA on your site would be used to influence your site's rankings.

What date did you experience the loss in traffic? How big was the drop? Did you loose traffic from a small percentage of URLs on your site or did you seem to lose traffic across most of your URLs? Did you lose traffic for a small number of key keyword phrases or did you lose traffic across most all of your keyword phrases? Were there any major changes made to the site (redesign? changes to URLs? Changes to content? etc.) in the weeks or even couple of months leading up to the noticeable drop in traffic? Was there any link building going on in the weeks or couple of months leading up to the drop in traffic?

These are some of the questions I'd ask yourself to try to pinpoint what might have lead to the drop in rankings/traffic.

[Simsi]

I doubt very much that Google would use GA numbers to affect ranking when they cannot use the same numbers consistently across all sites they need to rank, thus comparing "apples to apples"

Just to play Devil's advocate for a moment...

What if Google used GA as a verification tool rather than an outright ranking tool. For example, if it's algos already "suspected" that a site was good or bad and that site had GA installed, why not look at the traffic to confirm the algo's suspicions?

One might also apply the approach to social media too: rather than take a factor into account that not every site integrates with, use it as a confirmation tool where a site uses it.

If I ran a search engine and analytics service, the temptation to use analytics to refine the service would be way too much to ignore.

[BobMM]

RE: "What date did you experience the loss in traffic? How big was the drop? .....?"

First, I am a small operator selling my project management software from my site. I used to get perhaps 1000 visitors a day, largely due to one organic search term "project management software". About 10% of visitors would download a trial copy of my program.

Actually I just checked and it was early-2008 when I turned on Analytics. Visitors dropped maybe 20% therafter and didn't bounce back. Traffic has steadily gone down since, with major drops somewhere about 1-2 years ago (penguin? panda?)

I had made improvements to the site (added relevant content, fixed HTML, etc,) on and off since that time but it didn't help. I took the analytics html code off my website after a year of declining results. Things continued to decline.

Visitors have dropped to 100 per day and that's probably including the various bots that visit.

As of last summer I was still showing up on the 3rd page SERPs for that search term. Then somewhere around October I dropped to page 5, and then in December to anywhere from page 15-20.

Recently I tried re-installing the analytics html code on the site. Now, I don't show up at all that I can find.

I admit I'm a small time operator and do my own web design. In the past it was relatively successful but now its not. Once I did employ an "SEO Professional" to work on my site but the resultant changes were trivial and there was no improvement.

I am sure there are many other factors at work here. For one thing out of 300+ actual backlinks google was acknowledging about 40. Now the links are still there but google acknowledges less that 10. (I'm using the Google Toolbar to see that),

I've come to this site in search of knowlege and am crawling the various articles. But I've always wondered about the coincidence of analytics.

Sorry for all the details....
.

[edited by: Robert_Charlton at 11:55 pm (utc) on Feb 2, 2013]
[edit reason] removed specific domain per Forum Charter [/edit]

[tedster]

Google has changed so much in the five years since 2008 - check into all the linked information about updates: [webmasterworld.com...]

We're talking about a whole new infrastructure (Caffeine) and a whole new way of ranking websites that has evolved since then, based on machine learning. With what they are doing today, they certainly don't need to violate data privacy of GA users that they have publicly guaranteed - and it would drive them away, any way. They do use GA information in aggregate when users give explicit permission

[TheMadScientist]

Just to play Devil's advocate for a moment...

What if Google used GA as a verification tool rather than an outright ranking tool. For example, if it's algos already "suspected" that a site was good or bad and that site had GA installed, why not look at the traffic to confirm the algo's suspicions?

Why would you target one site when you could use the data across many? I would personally try to not make the mistake of thinking one of the sites I work on is so important Google would target it specifically, but, they could use the behavior patterns of visitors to improve the entire algo and if the behavior pattern of one (or more) of my sites happened to be hit by the 'rule' they wrote, then it would Look like they targeted the site itself.

IOW: When you write an algo to rank 150,000,000+ sites and 1,000,000,000+ pages for 1,000,000,000,000+ searches a year, you don't target any specific site, you look for patterns, because then those patterns apply to any site, page, search in the future too, not only to the site or page they're found on now ... There's a Huge Difference in approach and application ... Quite a few people seem to miss the difference, not necessarily, in my opinion, to their benefit.

[Sgt_Kickaxe]

Established content? I've noticed that a page can remain untouched for the longest time and stay steady in rankings but the moment you make a small change(like add tracking) the old rankings can be lost. I wonder if pages are sometimes exempt to new Google changes if they haven't been modified in some time. These pages might get re-evaluated under current guidelines only when they are modified.

Just a theory.

[Simsi]

Why would you target one site when you could use the data across many? I would personally try to not make the mistake of thinking one of the sites I work on is so important Google would target it specifically, but, they could use the behavior patterns of visitors to improve the entire algo and if the behavior pattern of one (or more) of my sites happened to be hit by the 'rule' they wrote, then it would Look like they targeted the site itself.

IOW: When you write an algo to rank 150,000,000+ sites and 1,000,000,000+ pages for 1,000,000,000,000+ searches a year, you don't target any specific site, you look for patterns, because then those patterns apply to any site, page, search in the future too, not only to the site or page they're found on now ... There's a Huge Difference in approach and application ... Quite a few people seem to miss the difference, not necessarily, in my opinion, to their benefit.

That makes perfect sense and ties in with theories I subscribe to on how Panda doesn't traget specific sites but patterns.

GA could still surely contribute to those working out those patterns though. Contributing to a pot rather than profiling a specific site.

Either way, I still can't quite get my head around the fact that Google never intended to use GA to improve results in some way, shape or form lol.