Weird thinking - Remove Google Analytics from sites with low traffic?

[AjiNIMC]

If Google comes to know that my site has low traffic or low user experience then I guess thats the worst signal to open it for Google. One should improve and eventually that happens too but meanwhile how about removing GA and going with some paid analytics to send a signal to Google that I am ready to pay to get better analytics so I am serious about my website rather than showing my weakness.

Once I have good user experience and traffic, Google you take it all and rank me better :), just a weird thought but I guess it has some points to think over.

Also I am aware that Google will get data from many other ways but still it will have to guess it.

Thanks,
Aji Issac

[aristotle]

Google still might be able to use the Google Toolbar and the Chrome browser to collect data on your site's traffic.

Btw, it has occurred to me that if you do a lot of artificial link building for a site that gets very little traffic, Google's algo might spot the inconsistency, since a site that gets few visitors isn't likely to pick up many natural backlinks.

[wheel]

Control as much as you can what info you give google that can be used to rank your site. More backlinks, that's hard to be used against you if you build them yourself.

Traffic, keywords, referrers, all that stuff that GA provides google, that's stuff they don't need to know about because it can be used against you as you've noted.

'Will it', who knows.

[tedster]

Although many webmasters have expressed this same concern, no one has yet shown any data that supports the worry. I've worked with sites that added GA during the time I was working with them and saw no effect, up or down.

In fact, spokespeople for the organic ranking team at Google have explicitly said several times that data from Google Analytics is not a factor in the rankings. The two areas of Google don't even communicate with each other - as we saw when the SERPs team completely broke Analytics capabilities with their first shift to AJAX a year ago.

There's been a lot of online discussion about the "bounce rate" metric, in particular. The discussions get confusing because different people mean different things:

1) the Google Analytics bounce rate -- "the entry page was also an exit page"
2) the simpler kind -- "the visitor returned to the search results and chose a different click".

Again, no one doing actual testing has found a clear case where the #1 type of bounce rate is affecting rankings at all. However, the #2 kind of bounce rate (which is NOT affected by Google Analytics} is a part of Google's Quality Assurance check for their algo and ranking changes.

[FranticFish]

I've seen a study that (to me at least) demonstrated a correlation between #1 bounce rate, activity on site, and ranking. The user group was tens of people, less than 100 in all (and pre-personalisation). I've not been in a position to observe it first hand.

#1 is trackable in Analytics and on the face of it should indicate that the site was useful - the query ended there. But suppose I open my clickthrough from Google in a new tab / browser window? Would that still count as #1 or would it be #2? My entry page was my exit page but my original query is still open. Perhaps it's reliable to count it as #1 in any case because I bet I'm in a minority that open a few windows and reads them all rather than go backwards and forwards.

#2 is trackable on Google, but suppose the page was bookmarked before hitting 'back' to see what else was on offer? What if the query is reviews of a product and the person reads the first, then goes back to Google to read the second and so on. Is the first page useless? There have to be many other examples where hitting the back button doesn't mean the page failed. Can this be disregarded without skewing the stats?

[tedster]

Right - all those questions, and many others, are why Matt Cutts once remarked that bounce rate is too noisy a signal for Google to use as a ranking parameter.

About opening new windows, that alone does not affect the metrics. Using a different proxy server for each window - now that could mess it up. I'm not sure about the effect of switching browsers. That's something I do quite often.

[Reno]

I put a lot of emphasis on the old Joe Friday admonition: "Anything you say can and will be used against you". In this case, "say" is providing data that is none of their concern, as wheel noted above. There are too many reasonably priced log analysis tools on the market to take a chance that this information might work to your disadvantage.

....................

[arizonadude]

Although many webmasters have expressed this same concern, no one has yet shown any data that supports the worry.

True, however nobody has shown any data that it does not either.

I proved it to myself with my own situation and will never give Google anything again. ALL sites in analytics got wacked, none of the sites not in analytics were touched and I have a lot of sites. That's all the proof I need for me.

Google is not the friendly search engine it used to be and the less information they know about your sites that they don't need to know, in my opinion the better off you will be.

[tedster]

Google was never the friendly search engine that many thought it to be. And it is not the dark enemy that many now fear it to be. That rapid change in disposition is like a rubber band snapping back sharply after being stretched out way too far. But in my view, neither belief gives a very accurate model of the real world - neither "Google the cuddle bunny" or "Google the evil empire".

I know of scores of websites that added Google Analytics and then pushed no other changes live for several weeks. There were no ranking effects from that, not in either direction.

If anyone doesn't want to use Google's free Analytics product, that's certainly a very workable position. Many sites I own or help with do things that way. But the spread of paranoia bothers me almost as much as Google's reach for as much data as they can acquire.

I don't doubt that Google uses aggregated data from Analytics for various purposes. But it just doesn't make sense to use single site data for ranking purposes. Google measures many other ranking factors that are more universally discoverable and actionable.

[BillyS]

We were getting 500 visitors a day when we installed Analytics three years ago. Now we get 25,000 a day.

Don't be paranoid, be smart. Why in the world would Google use Analytics against us? Think this through:

Google sends most sites around 80% of their traffic.
Google knows the frequency of each search pattern.
Google knows the exact positions we have in their results.
Google knows the CTR for each position in their results.

Google doesn't need any help from Analytics. They can probably calculate your volume with a high degree of accuracy based on information they already own.

It's not very logical to suggest that Google would say:

Hey, that website doesn't get much traffic, I better penalize it.

Since they send most websites the majority of their traffic, the rich would get richer and the poor would get poorer. It's a never ending cycle. The folks at Google are too smart to make such a simple logic mistake.

[Reno]

Billy S, just to clarify, are you saying that you installed Analytics and increased your visitor traffic 50 fold? That implies that the rapid growth is connected to Analytics. It may very well be true that you would have reached that number with or without them. If I'm wrong about that, please tell us what it is about Analytics that contributed to such a healthy jump in daily visitors.

...........................

[tedster]

Reno, Billy is talking about a three year growth pattern - right?

[Reno]

tedster, that would be my assumption (and I congratulate Billy for such a successful run!), but with the 2 statements directly next to each other, it almost appears they are connected. I would imagine there is no connection whatsoever.

.........................

[AjiNIMC]

Btw, it has occurred to me that if you do a lot of artificial link building for a site that gets very little traffic, Google's algo might spot the inconsistency, since a site that gets few visitors isn't likely to pick up many natural backlinks.

aristotle, such a valid point, yes a website with low traffic may not get better backlinks.

Google doesn't need any help from Analytics. They can probably calculate your volume with a high degree of accuracy based on information they already own.

Hey, that website doesn't get much traffic, I better penalize it.

Billy, you are missing the whole point. For a site that is doing good, GA is good, for a site that is not doing good GA may be bad. Also Google sends you 80% traffic only after a certain period of time, not in the initial phase.

In the initial phase, it is more of a struggle period. I was asking as there are many type of webmasters, some who are really serious about their websites and some who are not. With GA wide open to Google they will come to know 100% who is serious and who is not. Without GA, they will have to guess using some 10 to 70% user captured data through various modes.

[McMohan]

Just my 2 cents -

1. Google is good at placing red herrings
2. Many times we over estimate what Google can do and what it actually does.

[tedster]

No argument with all that, McMohan. And here's another part of the picture:

1. Webmasters are very good at building myths from almost no evidence. We make assumptions that spread like a grass fire, with no one ever testing their accuracy. If only our websites went viral as fast as our myths do, you know? I really hope the community here can put a halt to at least some of that.

2. When Google gives us just a little transparency, that can fuel webmaster frustration rather than mitigate it. Listening to public information from any company and learning to separate the spin from the gems is a true art form. It takes practice to get it right.

Paranoia is not the same as healthy skepticism. The first is not conscious, and the second is.

[FranticFish]

I proved it to myself with my own situation and will never give Google anything again

I've worked with sites that added GA during the time I was working with them and saw no effect, up or down

These two statements to my mind aren't necessarily contradictory.

Please note I'm not casting any aspersions on site quality or criticising business models, just thinking out loud.

(a) suppose your business model is based on time on site, brand awareness, and a conversion = enquiry / phone call etc. Analytics reveals a healthy average use of your site. No adjustment required in rankings.

(b) suppose your business model is to get people to click an ad and then go. Google would class this as poor user experience and your stats would certainly demonstrate that clearly. I'm not criticising: if that is your business model you're not doing it right unless people arrive and leave quickly. Now, even if the ads are AdSense, Google makes more money from AdWords, and they'll also argue that your site does not provide much value.

If that's identifiable from your stats, why wouldn't Google use that information to downgrade your site?

[FranticFish]

P.S. And even if it's not your business model, if that is the usage pattern on your site for whatever reason, you could still get treated the same way. Google has always, always thrown the baby out with the bathwater.

[BillyS]

Reno, Billy is talking about a three year growth pattern - right?

Correct. Over the course of three years traffic increased from 500 visitors a day to 25,000. My statement was that Analytics was installed three years ago - and it didn't seem to have any negative or positive affect on our growth.

[BillyS]

Billy, you are missing the whole point. For a site that is doing good, GA is good, for a site that is not doing good GA may be bad. Also Google sends you 80% traffic only after a certain period of time, not in the initial phase.

AjiNIMC - you're missing my point. Google doesn't need to factor traffic into their ranking system. Traffic is very likely an anticipated result of ranking, not an input into it.

[internetheaven]

removing GA and going with some paid analytics to send a signal to Google that I am ready to pay to get better analytics so I am serious about my website rather than showing my weakness

I doubt "paying for analytics" is a ranking factor. Sadly, GA is one of the best - and it's free. I pay for analytics, but only because I'm fighting against Google's ownership of the internet.

I agree with others who have posted that there is no evidence that GA itself affects ranking. However, multiple websites under one GA account is another story ... ;)

[arizonadude]

I use Piwik now and find it provides me all the information and metrics I need about my sites without giving it to Google for free.

[arizonadude]

Webmasters are very good at building myths from almost no evidence. We make assumptions that spread like a grass fire, with no one ever testing their accuracy. If only our websites went viral as fast as our myths do, you know? I really hope the community here can put a halt to at least some of that.

Agreed, however Google is even better at it with all the half truths and misdirections that talking heads like Matt Cutts distributes. Also, in my case there was no myth about it. Clearly the numerous sites I had under 1 GA got tagged.

Case in point, the no-follow tag. Who talked it up and even told us how to page rank sculpt. Mr Talking head himself.

However, multiple websites under one GA account is another story ... ;)

That is actually a very good point. I don't think having 1 site in GA is going to hurt you at all, however if your have say hundreds, you have given Google a clear picture of your portfolio and that can only lead to bad things.

[outland88]

I think a little separation needs to be made here between commerce sites and Adsense sites with regard to possible GA effects and Google's vested interests. Bottom line I pretty much follow suit with Reno and Wheel.

[tedster]

Traffic, keywords, referrers, all that stuff that GA provides google, that's stuff they don't need to know about...

I definitely do appreciate that sentiment - and I used to share it. I still don't use GA on many of my personal sites because I set them up without using GA and I still have everything I need.

But the level of competitive data that is available to anyone these days for just a few dollars, or even for free is rather outrageous. and if someone is willing to spend some big bucks? whew!

Unless a site has a very limited presence, I think Google already has a very good idea about all those factors today, even without installing Google Analytics.

[blend27]

I started playing with my own script/db(not JS) just before "Florida" simply cause the stats/numbers from program we've used were not matching with what the raw log files said. With time and learning more about what I needed, looking at some diff packages, my script got much better. I have all the info I need to tell from where individual arrived to each page, user time spent on the site per page, all other juicy stuff which is all encrypted....

I could tell which products/pages were popular this comming Monday 3 years ago and what external/internal links were poing to those pages. I know which product image to tweak and which image dim to sell the product faster, sometimes without discounting it at all. As a matter of fact it was that script that convinced me to stop using Adwords at some point all together.

And ALL of it is REAL TIME without giving my customers/visitors privacy away, and my biz. data away, for free. But it's just me and a custom built ecom site.

2 Years ago I've approached an SEO Company to help us with some on page optimisation and after speaking with them few times was taled that in order for them to do their job effectivly, they will have to insert GA JS on every page. I respectfuly declined their bid for the project.

Blend27

[tedster]

without giving my customers/visitors privacy away

I like that motivation very much. Why should you let the Google/Doubleclick cookie stalk your visitors after they leave your site and "remarket" to them.

[brinked]

oh boy. I would like to get into this convo but then I would never get any work done.

Lets just say that I do not use GA or adsense on any of my sites. I use webmaster tools but I have many different account which I login to through web proxies, I separate sites on to accounts that have websites that are in similar industries. I don't have more than 3 websites per webmasters account.

[TheMadScientist]

2 Years ago I've approached an SEO Company to help us with some on page optimisation and after speaking with them few times was taled that in order for them to do their job effectivly, they will have to insert GA JS on every page.

WOW! That's one of the most amazing things I've read here at WebmasterWorld... They needed you to install GA to be able to optimize your site? LOL, are you sure it wasn't because they had a custom bot and wanted to spoof your stats or something?

I like the reasoning Blend27 doesn't use GA, and even more on the topic of this thread, has anyone even stopped to think if this is the road they're going down (data from GA is used to assist in determining rankings) and GA gives them more insight to what's going on with a site, then sites using GA should get an automatic boost for making their job easier and all other sites, not using GA, should see a slight decrease in scoring for not providing them with as much information for ranking purposes?

This whole topic is a cluster f*** of complete speculation, so use what you think is best...

If G wanted to use GA data in rankings, to me the most logical place to start is by giving sites using GA a boost in the rankings just for using their system and tanking the sites that don't use GA to 'encourage' people to install GA on their sites.

Beyond there I'll leave it to all of you to debate about, because to me it makes no sense to tank a site giving you the data you want, because it would encourage people to not use your system or give you the information... The converse makes much more sense to me, so if we start to see the 'top 10' across most (if not all) queries all running GA, then I'll consider it as more of a possibility, but until then, as long as the top 10 are 'intermixed' for random queries with GA and non-GA sites, I think any impact GA data actually has on rankings is non-existent to negligible and not even concern myself with it.

NOTE: No offense intended to the OP or anyone involved in this thread... I can actually see where AjiNIMC's thought originated, but I just keep reading the title at the top of the forum list and thinking, 'That makes absolutely no sense to me, unless they give a boost to sites for using it... I can't take it any more... I've got to post or something... LOL' Anyway, personally the only thing that makes sense to me if they really want the info and want to use it for rankings is to give a boost to the sites using it, not using any of the data to lower rankings, except against other sites using GA already, and to really do it, IMO, they would have to 'really encourage' everyone to use it by lowering the rankings of those not using it.

NOTE 2: They also recently announced the launch (or soon to launch) of a plugin for browsers to block GA data from being sent... So, it really makes no sense to me for them to try to use data from a system they are authoring a block for to determine rankings for sites.

[outland88]

to me it makes no sense to tank a site giving you the data you want

It doesn't work quite that way. Once you voluntarily drop out of a program like GA you are know longer feeding your data to Google saying you exist. What many people assume though is things will return to the previous state before GA. I would expect that with a fee based analytics not attached to a search engine. I wouldn't be expecting that from a free system of analytics attached to a search engine. There a return to the previous state when the data is not available might be an impossibility. It’s not a system of what’s right or wrong but in the way the data is used. Google sees it as system of helping you while helping them but help is a broad word. Do you sink, swim, or stay the same once the supposed free help vanishes.