It’s a classic website conundrum. If you use multiple analytics tools on the same website they almost always report different numbers of hits/visits/etc., and sometimes the difference can be drastic! This can be annoying and inevitably prompts questions about whether one or more of the tools is inaccurate. But there are many explanations for your website analytics not lining up. Here are some of the most common reasons for differences across various analytics tools.
Internal Traffic Filtering
One tool may be including all your own internal stakeholder and content editor visitors while the other filters them out. This often happens by default in the internal analytics systems that come with CMS’s, which can exclude you from the stats if you are logged in to the CMS admin. Sometimes analytics tools are configured to automatically ignore traffic from specific IP ranges to help remove internal traffic from reports.
One tool may be including lots of bot/spider traffic while the other filters them out. Bot/spider traffic can be enormous, sometimes comprising a large percentage of the total traffic load on a website. Some popular analytics tools include functions that attempt to automatically filter some of these out for you. For example, Google Analytics includes a "Exclude all hits from known bots and spiders" setting that you can toggle. Even if you activate such a setting in all your analytics tools, they likely use different methodologies and blacklists to detect bots/spiders.
Server-Side vs. Client-Side
Differences in Page Coverage
Sometimes one tool is running on more URLs on your site than the other tool. For example, you may include Google Analytics in your site's main rendering template which applies to all standard CMS-managed pages, but there might be additional automatically-generated CMS pages that a server-side CMS-internal analytics package also includes in its record keeping that your Google Analytics code is not present on.
Duplicate Tracking Codes
It is technically possible to mistakenly include your tracking tools twice into the same pages, thus potentially double counting traffic to those pages. Some analytics tools may be smart enough to automatically detect and weed out these situations, but getting it right in the code is your best bet for reliability.
Differences in Domain & Subdomain Coverage
Analytics tools usually collect data for a single website (a single domain or subdomain) by default. It is possible to configure analytics tools to collect and pool data across multiple domains and subdomains, such as a main website like mysite.com, a blog like blog.mysite.com, and an e-commerce section like shop.mysite.com. If you don't set this up the same way across all your analytics tools it could cause discrepancies.
Some spikes in analytics stats can be caused by sneaky automated visits by bots whose goal is to advertise their own product or domain name in your analytics reports. Different analytics tools may detect and filter this type of spam differently.
Some analytics tools do a better job of identifying the same user as they switch IP addresses and devices, such as first visiting your site in the morning from their cellphone network and later in the day from their home network on their desktop computer. Server-side solutions that are aware of visitor login status on sites that have user login systems are especially good at understanding when multiple visits are from the exact same user. Client-side tools are at a disadvantage in this area, although some provide ways to improve the situation, but in general they are more likely to count the same person twice since that person keeps switching networks and/or devices.
Disabling for Privacy & GDPR
With GDPR and other privacy rights regulations growing in popularity worldwide, it is becoming common to purposefully disable data tracking tools like analytics until each visitor provides consent for it. If you don't treat this disabling in the exact same way for your multiple analytics tools you could see data collection differences. For example, does your client-side analytics get deftly loaded asynchronously as soon as the visitor clicks the "Consent" button while your server-side analytics has to wait until that visitor's next page load to collect data?
Different analytics tools have different delays between the time of data collection and the time you can see that data on nice reports. There is usually some delay, during which report data is aggregated, analyzed, and graphed. So short term differences in numbers may appear even though they get evened out over longer timings.
Your tools may treat HTTP redirects differently, with one tool counting all redirects as separate "hits" while another tool filters out all redirects and only counts "hits" for the final destination pages.
Front-End Magic & SPAs
Purposeful Under/Over Counting
Some tools purposefully under or over count some statistics for business or political reasons. For example, it is common for marketing tools to slightly over-count to encourage you that their marketing platform is working for you. Likewise, it is common for ad network tools to slightly under-count to discourage you from thinking they owe you higher payouts.
While there are many other reasons in addition to these for why different analytics tools might produce different numbers, the bottom line is that website analytics is not a simple straight-forward subject wherein every tool should be expected to produce identical results. The creation of an analytics tool involves hundreds of small judgements and logic decisions that will affect the reported outcomes and the sum total of these decisions can add up to big data differences.
At Hedgehog we have experience with many analytics and digital marketing tools. If you need help understanding your analytics data or honing your digital marketing to its fullest potential, contact us for more info.