11
June
2020
Spotting Fraud and Reconciling Ad Traffic with Google Analytics
In this series of posts, I’ve addressed some of the reasons why what Google Analytics reports may not align with the numbers you’re receiving from advertising partners. In this edition, I’ll address two more potential causes that may be more difficult to resolve.
Clicks Counted as Direct Traffic
A lot of display ad traffic happens on mobile devices, specifically in apps. This can be problematic because when a user clicks an ad in an app, their browser has to open—the website doesn’t open within the app.
Sometimes, Google Analytics gets confused about where this user came from and just dumps them in the “direct” traffic bucket. In other words, Google Analytics thinks, “Hey, this person just opened up their browser and went directly to this website.” It isn’t able to see that they really came from an app, specifically an ad within an app.
This recently happened to one of our clients. A long-standing vendor reported 17,833 clicks for a given campaign. During that same time, Google Analytics was only attributing 5,734 users to that vendor. That’s a discrepancy of almost 68 percent—far larger than what we had ever seen from this vendor.
Closer examination showed that this client had a HUGE number of direct visitors for the campaign period. The number of direct users was about nine times higher than the site average. After accounting for all of the previously mentioned reasons for a discrepancy and talking to the vendor, we determined that Google Analytics had been counting in-app clicks as “direct” traffic.
This had not happened in previous campaigns.
We told the vendor to stop serving our ads in apps and we’ve since seen a much smaller discrepancy between the vendor’s report and Google Analytics, as well as normal levels of direct traffic.
Is It Fraud?
If you’ve ruled everything else out, it’s sad to say, but there could be fraudulent activity going on. There are two possibilities:
There were invalid clicks on your ad. Most ad servers have safeguards in place to filter out invalid clicks (like those by a bot). Typically, your vendor will deduct these invalid clicks from the total in your report. However, Google Analytics reports all users. This is an instance where the number of Google Analytics users may be higher than the numbers reported by your vendor.
Your vendor has inflated the number of clicks. This could be some kind of glitch with their reporting system OR an intentional action to make the campaign appear to have performed better than it did.
It’s a sad truth, but there are vendors in the marketplace who are less than reputable. It’s not the most frequent reason behind reporting discrepancies, but it does happen. By understanding the logic behind the reports your web team and your vendors are delivering, you will be better equipped to protect your budget and yourself.
This post is part of a series on digital marketing analytics. To read the others, follow this link or subscribe to our blog to get updates when new posts are available.
Loading…
21
May
2020
Diagnosis and Treatment of Web Analytics Errors
Multiple digital campaigns running on different platforms can make measuring results more complicated than we’d like. The challenge is particularly acute when comparing your own Google Analytics to reporting from a third-party partner. In a previous article, I discussed how one of the main causes of discrepancies is conflating Google Analytics users with third-party clicks. While that’s a common issue, there are several other reasons outside data that may differ from your own metrics—and sometimes wildly.
Google Analytics Is Not Firing
In order for Google Analytics to count a user, the Google Analytics tag must fire, which happens as the web page loads. This means if a visitor clicks your ad and then quickly hits the back button your vendor will most likely report a click while Google Analytics will show nothing.
In addition, if the visitor prevents the page from fully loading by quickly moving to another page or by pressing the browser’s “stop” button, Google Analytics may not fire. Your vendor would report a click because the person did click your ad but Google Analytics would show nothing because it never fired.
Google Analytics Is Not Set Up Properly
If Google Analytics is not set up properly, it may not accurately count users or document where they came from. This could be something as simple as Google Analytics not being installed on your landing page or something more complicated, like a cross-domain tracking issue.
Traffic Is Being Sent to The Wrong Page
This sounds silly, but if you’re seeing large discrepancies, always ask the vendor to verify the URL they were sending traffic to. We’ve seen cases where clients have accidentally provided invalid URLs or URLs to a different website and therefore traffic was going to an unintended location.
Not every tactic we roll out as marketers is going to work perfectly the first time. To get the best value from your marketing investment, you have to make choices. And to make the right choices, you need data you can trust.
And confidence that it’s being presented accurately.
This post is part of a series on digital marketing analytics. To read the others, follow this link or subscribe to our blog to get updates when new posts are available.
Loading…
14
May
2020
Why Don’t Vendor Numbers Match Google Analytics?
Author:
Alex Webb
Help; the clicks my vendor is showing don’t match what I’m seeing in Google Analytics!
Unfortunately, this is a very common problem. In fact, we’d say it’s the norm!
Usually, the number of clicks a vendor is reporting is significantly higher than what you’re seeing in Google Analytics, which can be quite distressing. However, there are a myriad of potential causes behind this discrepancy, and while more than one factor may be at play, let’s take a look at one of the most common issues.
Clicks Are Not the Same as Users
If you’re running display or pre-roll video ads, the vendor is probably delivering those ads over an ad server or ad exchange. The vendor is most likely tracking clicks, which are calculated based on server logs.
On the other hand, Google is tracking “users.” A user is not necessarily the same thing as a click. A user is a unique person that has come to your website. To put it more accurately, it is a unique device and/or browser.
Let’s dig a little deeper.
Scenario 1: Let’s say John is surfing the web using his iPhone and a Safari browser. He clicks your display ad three times.
Your vendor would count this as three clicks, but Google Analytics would only count this as one user. That’s because the first time John clicked your ad and Google Analytics fired, he was “cookied.” A code snippet was appended to him so that Google could recognize him. Now, any time John visits your website on his iPhone in Safari, he’ll just be counted as one user unless he clears his cache or the cookie expires.
Scenario 2: Let’s say John is surfing the web using his iPhone and a Safari browser. He clicks your ad two times. Later that night, John is on his laptop in a Chrome browser and clicks your ad once.
Your vendor would count this as three clicks, but Google Analytics would count this as two users. Once on the iPhone in Safari (even though he clicked the ad twice) and once on the laptop in Chrome.
Unless John changes his device and/or browser or clears his cache he’ll continue to be counted as one user in Google Analytics … but your vendor will count a new click each time he clicks on your ad.
Conflating clicks with users is the quintessential apples/oranges issue in third-party reporting. But, it’s far from the only possible cause of reporting discrepancies. In an upcoming post, I’ll discuss some other common causes and how to use your web data for the best results possible. If there’s a particular issue your brand is wrestling with, give us a call or join the conversation on LinkedIn or Facebook.
This post is part of a series on digital marketing analytics. To read the others, follow this link.