Discrepancies in Website Tracking – When Paid Tracking and Google Analytics Don’t Add Up

by Marjorie Steele, Editor at IQS

This is a question which is often raised by our advertisers. IQS offers free Hitslink website tracking to all companies listed in our directory, not only so that listed companies can track the progress they're making through being listed on IQS sites, but also so that our manufacturers and service providers have an effective tool for measuring their overall success on the web. Naturally, many of the companies which come to be listed on our sites already have some form of website tracking (although you’d be surprised to find out how many companies have no website tracking at all). While some have paid tracking programs, most of the companies that have tracking come to us already set up in Google Analytics, Google’s free website tracking service which goes hand-in-hand with Google’s Adwords program.

So, inevitably, we are posed with variants of this question: “I’ve been tracking with both Google Analytics and the Hitslink tracking you provided. How come the results from these two are different?”

Unfortunately, there’s no single answer to this question, and it’s difficult to find any satisfactory answer without turning into a super SEO nerd. In an effort to make this explanation as painless as possible, we’ll break down some possible reasons there may be for discrepancies between website tracking into a few main categories.

Unique Visit Definitions: Different website tracking services define unique users differently, and this can be one of the biggest contributors to tracking discrepancies. In an article I posted last week I made a distinction between “unique users” and “page views”, or “page per user count”. It’s important to know how your website tracking is defining these terms and what criteria it uses to measure them.

Google Analytics tracks “unique users” in half-hour intervals, which means that if a user comes to a site three times within the same half hour without closing his browser, GA will register one visit. However, if the user closes his or her browser and comes back to visit later in the day, he will be recorded as a new “unique user”, raising GA’s unique visitor count. Hitslink uses both cookies and IP addresses to track unique visitors by day, month and all time, so that one person visiting your site in a day, no matter how many times he visits, is counted as one user. This can give the appearance of lowering user numbers, but it’s all a matter of how you do the math. It can also dramatically change your page view (page per user) count.

Cookie-based vs. IP User Agent-based Tracking: In this regard, we’ll call Google Analytics the Cookie Monster. There are three main methods of tracking: cookie-based tracking, which relies on a user’s browser to retain the tracking cookies it is given; IP address tracking, which looks at site users’ IP addresses; and a combination of IP user agent and cookie tracking. Cookies are tiny bits of information that get stored in your browser’s files, telling your computer which sites you’ve visited, what your preferences are at certain pages, et cetera. Unlike your IP address, which is kind of like an online name tag, cookies can be deleted by users.

The difference between cookie-based and IP-based tracking is how unique users and visits are counted. Website tracking services which provide in-depth reports will try to distinguish multiple visits from the same person as one “unique user” so that you’re not counting the one person who visited your site 30 times in one day to look at a picture as 30 different people, you’re only counting him once (as one “unique user”). Think of it as counting customers at a grocery store; counting the gross number of grocery items sold is a much less effective way to gauge marketing effectiveness than counting individual shoppers and taking note of what each shopper buys. If you’re using cookie-based website tracking (like Google Analytics), any user who has disabled their browser’s cookies or recently deleted all cookies for security purposes will be counted as a unique user each time he visits your site, instead of only being counted once.

Real-Time vs. 24 Hour Updating: One of many people’s complaints against Google Analytics is that it does not reflect real-time tracking results. Google claims its analytics are updated every 3-4 hours, but it can take as long as 24 hours for GA to refresh its statistics. Hitslink and many other in-depth analytics operate in real time, so if you just posted a new blog last night and are checking your stats this afternoon, Hitslink will (hopefully) give you higher numbers than Google.

Timezone settings can pose a similar issue. If two website tracking products are set to different time zones, the lag in registering results can create a difference in numbers.

Enabled Cookies, JavaScript & 3rd Party Images: Some browsers allow their users to block cookies, JavaScript and 3rd party images. This is often done through firewall settings and browser settings in for security purposes, but in the process of blocking “bad cookies”, many cookie-based and IP-based analytics are blocked from collecting data from users as well. Tracking JavaScript content and images spidered from 3rd parties often get blocked by certain users’ firewall settings. There’s also the question of whether cookie-based tracking uses 1st or 3rd party cookies; these make a difference in that 1st party cookies (given directly by domains, e.g. www.yoursite.com) are not blocked as commonly, but they don’t track a user’s path across the web very well. 3rd party cookies are picked up by users from outside domains, making them easier to block but also better tracking devices. For the curious, Google Analytics uses 1st party cookies, and Hitslink uses 3rd.

Click Fraud Reporting: While researching this topic, I stumbled across a blog posted by a company which had done a study on this issue. They used two different analytics programs to measure their conversion rate on a pay-per-click program over a short span of time. Their in-house tracking recorded almost ten times as many clicks as Google Analytics did, and when the company broke down the data manually, they found a large portion of clicks Google had discarded as fraud. Click fraud certainly does exist (if you don’t believe me, read this article), so be sure you know what capacity your analytics program’s has for avoiding this type of fraud.

Filtering Data: Differences in tracking data very often boil down to one simple, overlooked factor: how is your data being organized? Analytics services, whether free or paid, not only use different definitions, but they may filter data differently — often at your instruction. Website tracking services can be set to sort and filter the data it receives in different ways (as we saw earlier with unique user counts), so before you cancel your account with a website tracking tool that doesn’t match up, check to see how you’ve filtered your data. If you tell one tracking tool to count each visit as a unique user and another tracking tool to count unique IP addresses as single unique users, your user counts are going to be very different.

It’s a pain, but any web-savy marketer would tell you that analytics are an essential part of your business campaign, and unless you want to throw your dollars spent on web marketing down the toilet, it’s best to get friendly with your website tracking. If numbers don’t add up, sit down and sort them manually. Examining each click, cookie and IP address can be tedious, but if it helps you better understand how your site’s traffic is being measured, then you’re one step closer to understanding how to optimize your site.

VN:F [1.9.10_1130]
Rating: 0.0/5 (0 votes cast)
  • Facebook
  • Twitter
  • LinkedIn
This entry was posted in Industrial Marketing and tagged , , , , , , , , , , , , , . Bookmark the permalink.

Comments are closed.