Digital data isn’t really degrading. The truth is that it’s never been very good, and as much as brand marketers like to talk about the importance of big data and analytics, both are far from perfect in the digital space.
The Evolution of Digital Advertising report from comScore clarifies why digital data, particularly demographic data, is so flawed. Ironically, comScore uses data to make its points, but nevertheless, the points are quite valid and bear some consideration from brand marketers.
The problem with digital data for brand marketers is broken down into five key areas:
1. Misleading Click Data
Click-through data is rarely reliable. Not only is it filled with click-fraud, but comScore’s research found that exposure to display ads doesn’t just impact online sales. It also lifts in-store sales. In fact, comScore reports actual lift in offline sales is five times higher than the lift in e-commerce sales. Ultimately, relying on click data to measure the performance of a digital ad campaign is often not enough.
2. Inaccurate Demographic Data
Demographic data is often identified based on previous sites visited (using cookies, discussed in #4 below), content consumption, and registration data. However, comScore reports that this demographic data becomes outdated quickly and isn’t 100% realistic. According to the research, more than 60% of computers have multiple users. How can brand marketers be certain that their ads are being shown to the right users based on previous behaviors and content consumption when multiple people are using the same computer?
3. Misleading Impression Data
Impression data is misleading because it doesn’t account for some critical things. For example, 100% realistic impression data is impossible because some people leave the page where the ad appears before one second. Furthermore, some people will visit the page, but they won’t see the ad because the browser window isn’t set to full-screen. The ad might not be visible unless they scroll around the page (or the site uses a responsive design that resizes it to the smaller screen size). Furthermore, traffic fraud artificially inflates impression counts when no one actually saw the ad at all. The advertising industry has been combating traffic fraud for years, but it still happens every day—a lot.
4. False Cookie Deletion Data
As mentioned above, cookies are used to serve ads to the target audience, but many people, websites, and ad servers delete cookies. That causes higher than planned frequency because ad servers deliver additional, “new” impressions to people who deleted the cookies from their browsers.
5. Disregarded Platform Data
In the United States, more than half of internet time is spent on mobile devices, and that number continues to rise very rapidly. Similar results are being found in other countries around the world. According to comScore’s research, more screens equals more time spent on each device overall, but tracking ad campaigns across devices is far from accurate. Furthermore, processes for collecting and analyzing data are quite different from one device to another. For example, while 60% of computers are used by multiple people, it’s safe to assume that most smartphones are used by a single person. Understanding these nuances as you track performance between devices is equally important as learning how to track across devices.
What do you think are the biggest problems with digital data for brand marketers? Share your thoughts in the comments below.
Image: Janet Ramsden