If you use Google Analytics READ THIS
 

http://www.chokertraffic.com/sample/

Whenever a potential buyer says "I use google analytics" I cringe, and for good reason. It is not accurate. PERIOD. But because it's a google product most people assume it's 100% accurate. I mean google has the Midas touch right? WRONG WRONG WRONG

This test is ongoing so I will post more results as tests are completed. My tests show about 15% loss in traffic but I know of people who commonly lose 30% or more. I'm thinking some sort of Java script interferes with GA. Anyone out here had any exp with this?

maybe your counters are wrong instead?

My server side stats are usually close to double what analytics says.

It's known to start losing track on high volume traffic.

It's a known fact that Google Analytic isn't 100% accurate. They recommend you place the code high up on the page. The higher up the better it tracks but since most people add it to the bottom that accounts for most of the loss.

The key is to simply not take the numbers as cut in stone and more of an estimate.

Read the page before you post such stupid comments. I used 5 different counters.

You know, I hear over and over again that Google Analytics is not always accurate. I've done a lot of testing, using a lot of stats programs, and they all seem to confirm what Google is telling me. I have yet to find anything that tells me Google Analytics is wrong.

Last month I spend an entire getting my entire blog network into a new stats program, only to have it confirm what Google tells me. I don't need precise stats, so I'm just sticking with Google.

Google doesn't claim it's 100% accurate though. 5-20% off is about what I would expect from that type of tracking. Plus you have to factor in the people who click out from your script but for whatever reason don't load the incoming page fully enough to activate the tracker.

so a billion dollar company with billions of dollars for research is wrong and we adult webmasters are right?



not taking sides here just making a statement....

I have seen both kinds.. sometimes my google matches all my other trackers sometimes it don't

what counters you use?

Well that's odd, I posted a link of controlled tests using 5 different counters, in fact 4 of them are trade scripts, all from different coders, and they all say way more than GA says.

Free trade scripts. 4 different ones plus a simple php counter/redirect.

which free trade scripts?

Nope, 4 of the counters are php includes so the page loaded enough to load the includes.

Do you have GA placed high up or low on your blogs? And if it's low, do you have really long pages or do you set your blogs to only show a few posts per page?

If i don't have to use them, i go with awstats.

one of my hosts doesn't have awstats so i have to put analytics on there just to get an idea.

Kinda like people that take alexa ranks seriously. they might be a fair indicator of popularity but traffic wise they are miles out.

Stick to server stats if possible

4 of the most popular free trade scripts, I'm not gonna post which ones I used and which counted better than the other, I'm not gonna get into a pissing match with free script owners over how many hits their script counted or didn't count.

Same here.

Javascript confirmed enabled for all traffic? I see no mention of that on your page. The default GA config used to be dependent on javascript. I'm not sure how things are today. I use awstats.

Consider that if someone uses NoScript and Firefox with the default config and have yet to visit that site and whitelist it, scripting will be disabled. Some also actively block google analytics.

GA is not that acurate, statcounter is better tool for me

pointless thread because you may be using shit stat counters. many of them are flawed.

no way to judge the validity of your claims unless we know what they are.

you sure its not repeat visitors with same ips? or did you take that into consideration?

Same exact thing man. Ive seen 30% difference and think thats pretty normal for google stats. Makes you want to pull your hair out lol.

Who cares about few hits up or down, it's sales that count

they probably do a pretty good job of filtering out bots/crawlers/etc... it's not uncommon to see few percent of bots/crawlers with normal traffic... brokered/traded traffic is probably double or triple that...

Traffic trading scripts / your own stats don't do any filtering, so that's part of the reason why the numbers don't add up... :2 cents:

well, with purchased traffic I guess it's CPA that counts :winkwink:

And yeah, I'd say about 10% discrepancy with analytics is normal

Doesnt a test need a control? How do you know what stats are correct and which are not?

If 4 FREE traffic scripts all show nearly the same stats while a billion dollar companies script shows very different stats, I'm going to have to go with the billion dollar company and figure they know a few more tricks.

I do know that the coding I use to track referrers, track SE traffic/queries, and to trade traffic nearly mirrors my GA stats. And the coding is nothing overly technical, actually pretty simple just a few checks and stuff.

Of course server side stats have to be the most accurate, but awstats is the most common one and its painfully basic and doesnt really give you much information.

Have a friend who works on Analytics and says around mid-2011, their accuracy will increase substantially, especially on high volume sites. They're close enough for me. If I want exacts, I'll fire up awstats(which is usually pretty accurate afaik).

Interesting test, though.

Where did you read this? Please post links.

GA does not track those surfers with javascript disabled.

I think a little clarification here would be a good idea before people start making some wild assumptions.

As someone mentioned earlier, the only way to test which stats program is the most accurate is with a control. For most of us, comparing AWStats to GA on our own sites is useless because the data is interpreted by each program in different ways. It's like comparing apples to oranges.

I would trust GA over AWStats any day. First, because its Google. Next, because Google has a very strong interest in make sure the results they provide to you are as accurate as possible. More accuracy for you means more accuracy for Google, means better returns on Google searches, Adwords and Adsense. Just because Google gives you the tools to track your stats for free doesn't mean Google doesn't profit from this, and the only way they profit is by accurate data.

The only way to properly compare the two for accuracy would be to setup a highly "controlled" (ie: Free from all unwanted Internet traffic) server, and then have it surfed from a couple of hundred different KNOWN locations, a fixed number of times. Only once you know all of the controls can you properly compared any stats program to any other.

Personally, I love GA because the data it presents to me is more useful and seems more accurate. Especially useful is the Goal program.

If you run GA and you don't like the results you see compared to AWStats, dig deeper before you trash it.

I thought that was pretty much common knowledge. Anyone that checks their server stats and analytics should know they never match up.

LOL.

Choker come on, you cant be that stupid.
Off course there are some leaks and nothing is perfect, but if you think a couple of "free counters" match up to google analytics you are more of a noob then i thought.

Why are you even trying to "solve" this? People got offended you didnt send as many hits as they purchased?

If so, tell them to fuck off, its the traffic conversions that count, if they made much more then what it cost them buying your traffic they wont cry about it.

Guess that says it all...

(dont try to be a webmaster if you dont have a clue mate)

I block google analytics. Got really tired of the slower page loads when it was enabled. So I don't think I'd show up as a visitor in ANY google analytics stats.

Of course awstats or something local on the server is going to be more accurate :2 cents:

Earlier this year I tested GA on 4 of my sites. Only organic traffic (no traffic trades, no traffic buys). GA missed 17% of my visitors.

Ok, I may not be running a multi billion dollar company, but those that know me, know I log/filter/analyze the shit out of my traffic and am constantly trying new ways to detect more and more information about my visitors.

How do you know that this was actually "missed" visitors? Since GA and AWStats add up traffic differently, maybe it was due to a simple interpretation difference? Maybe 17% of your traffic had the same IP address for example and were not counted as "new" users?

Exactly what I would think. It's common sense. Awstats typically uses the actual server access logs. There's really no comparison. That said, GA tends to track certain metrics which the typical awstats install does not. It isn't really a question of one being better than the other. They both have different potential uses. However for a raw hit metric, awstats has to be more accurate because it uses the actual raw server access log compared to GA which typically uses a javascript based tracking code.

Because I'm logging EVERYTHING I can on some of my sites.... IP, browser, JS, Java, Flash, software installed and versions, lang accept, proxy or not, time, cookies, links clicked, type of links clicked, type of galleries visited, time between clicks, mouse movement in some cases, geo>timezone>localtime, type of internet connection, resolution, adblocker installed,... and a lot more...

someone burned choker or complained.

adapt or die.

Yep, there's always the people who swear that google is 100% accurate no matter what, posts in this thread proves that point. Because they are a billion dollar company thier product is perfect. LOL