I thought I knew how many people visited this website. It turns out I had no idea. And guess what, you have no idea how many visitors your website is getting either.
Recent events gave me an education on just how poor the analytics we use to judge the performance of our websites are.
This started when my hosting provider complained about this website making too many CGI requests on the server.
My response was this was impossible, as my little website is tailored to a very niche audience and only gets about 2,500 page views per month. That’s straight from my Google Analytics account (clearly the authority, right?).
The webhost’s response floored me. They said I was getting half a million page views per month.
Wha whaaa what? Could those numbers be any different?
This baffled me, so I did some research. And what I found made me realize we really have no definitive way to determine how many people visit our websites.
The Problem With Google Analytics
Google Analytics relies on web browsers to detect visitors. It’s done on what the IT nerds call the “client side.” It uses JavaScript to detect visitors. So, anyone looking at your website without JavaScript turned on does not get counted. If you have certain web browser plugins, this may also make you “invisible” to Google Analytics.
It also doesn’t count bots or web crawlers (which is a good thing). It also doesn’t count when someone reads your RSS feed. And it has special rules it applies to determine what a visit or page view is. Lastly, some say that if your Analytics code appears at the bottom of your page code, that will affect the number of visitors Google Analytics detects.
So, here’s the Google Analytics data for August 1st. On that day, it says this website got 50 page views from 30 visitors.
The Problem With Webalizer
The other side of the coin is Webalizer, which some web hosting servers use to track analytics. Unlike Google Analytics, it uses the webserver to get its numbers. It counts everything and anything, whether it’s a bot or crawler or anything that “visits” your website.
Here is the Webalizer account of what happened on August 1st: 3,494 page views ( don’t get me started on page views vs. hits ).
Oh Yeah, What About RSS?
If someone is reading your website through Google Reader or another RSS reader ( I use REEDER ), then Google Analytics won’t track that as a page view. I have two RSS feeds on my site, the ones generated by wordpress and Feedburner.
Let’s look at the article about which font to use for proposals that I posted earlier this week (since I don’t have August 1st rss info). Feedburner says the RSS of that article got 164 views within a 24 hour period. Now, I don’t know which of those views were people and which were bots.
I can tell you who two of them are. One was Mailchimp, which I use to send articles to those who subscribe to this site. Mailchimp takes the RSS feed, captures it, and then sends it out. Mailchimp says 109 subscribers read the article they sent out within 24 hours.
I’ll also mention that Feedburner gets its RSS from my WordPress installation. I have no way of tracking who gets the RSS strictly from WordPress.
Oh, but there is more. That article got picked up by Zite, which is an iPhone/iPad app that provides users with “personalized news,” and landed as one of their “top stories.”
Zite has a sh*t ton of users, probably numbering in the millions. I’m sure not all of them saw my article, but I have to believe some users read it. I also have to believe that Zite caches the articles on its servers, which means the number of readers wouldn’t be reflected in the Feedburner stats.
I think a lot of people read that article. I can account for at least 200, but the real number is probably in the thousands.I really have no way of knowing.
During that same time period, Google Analytics reported 17 page views for that article. Therefore, if I went by Google Analytics, I would believe that only 17 people read the article.
Do you see how this could be problematic? Google is not just off, it’s way off.
Here’s what I learned. You can’t trust Google Analytics because it doesn’t tell the whole story. You also can’t trust Webalizer because it doesn’t distinguish between computers and humans. Plus, there are other factors to consider like RSS.
What’s The Right Answer?
I think the right answer is somewhere between Google Analytics and Webalizer (if your article does not get picked up by some third party service). But those numbers, at least for me, were so far apart you couldn’t even venture a guess.
Keep this in mind when people ask how much traffic your site gets. The real answer is, “there is no way to know the real number. But I can tell you that it’s at least ____.” Or you can provide client side numbers and server side numbers and state that the truth is somewhere in-between.
If you liked this article, please subscribe below. If you want to give us your thoughts on this issue, please leave a comment below.
A very informative article,i have ever read on that issue.I am worried too much about difference of site statistics recorded by google and my site hits recorder. You pointed out a very logical way to sort it out.
Thanks for the informative article. The stats are either overwhelming in terms of visits (you know that something aint right about those numbers) or downright depressing (huh? only three real views a day?).
Thanks for enlightening us on this. What traffic is the alexa ranking based on, Google analytics or webalizer. Any experience on the Awstats? I thought it was different from webalizer.
I am also using another wordpress pluggin called WP statistics which gives more than 10 times the numer in Google, I wish I could trust it. It also gives a breakdown of where visitors are coming from together with other important details.
Regards Amos
I’m not an expert on Alexa. I think they have their own data.
we have registered our domain name 9 years back. we have not tracked so far the number of visitors in our website.
Is there any way to know the number of visitors my domain received from the date it is hosted.
Sorry, not that I know of.
Good afternoon, i launched my website 3 weeks ago, the difference between the server stats and google analytics is the first has the double of the second, in any case for such recent website they are not bad. My question is: the server stats says that it separates bots from the organic visits, if it does so, can i use this data as being accurate, or google analytics gives more precise data than the server?
I would pick one and use it to have a consistent measurement. Ultimately, it is very hard to get a truly accurate measurement of human visitors.
I noticed that if you put a file, like pdf, to media in WP and a link of it to a post, then Webalizer shows how many times it was downloaded (as long as it is among the largest kilobytes as webalizer only shows some number) while Google Analytics plugin Monster Insights will not show it. Bots do not typically download files, so at least these you get correctly from Webalizer.
I’ve always sensed the Google numbers were low, but now I’ve got some hard evidence. Thanks for that!