Are you demanding too much from your digital marketing metrics? If so, you are not alone.
“People are looking at metrics and asking certain metrics to do things and to give answers to things that they were never designed to do,” says Lauryn Bennett, head of brand at data analytics company Chartbeat.
It’s a practice that can have bad consequences. After all, companies are analyzing massive amounts of data points to understand how customers are responding to their marketing efforts. When the purpose and limitations of metrics are misunderstood, the data may not be telling the story you think it is. Asking metrics to answer questions they were not designed to answer can lead marketers to make poor decisions and take actions that are not supported by data.
Some of the metrics that are commonly used incorrectly are also among the most popular: click-through rate, page view, average time spent on page, number of followers and unique visitors. Below, we explain what each metric is designed to measure and describe some limitations, common misconceptions and misuse of these metrics.
Click-through rate. Click-through rate reveals how often people click on your ad or content after they have seen it. This metric has been both used and derided for years – and indeed its challenges have been well reported.
Brand marketers find it comes up short as a measure of exposure, as someone doesn’t need to click on a display banner to be exposed to the brand. Nonetheless, David Karel, chief marketing officer of the marketing technology company Bizo, still hears marketers wanting to use click-through rates to measure the success of a branding campaign.
Karel warned against using the click-through rate to measure how well an ad drew someone to respond. Some people just click on an ad and stop engaging, and their click shouldn’t have the same value as a click by someone who takes a marketer’s desired action, he said.
“You want to optimize your budget to the people who are taking the actions, like downloading a whitepaper, not the people who are just clicking, so clicking could be a false signal,” he said. Judging an ad by the number of clicks alone, he added, could lead marketers to spend more money on an ad that isn’t performing as well as it appears to be.
Number of followers. This metric reveals how many people subscribe to your content.
This social media measurement is often seen as a vanity metric, and there was a time that was important for Bizo. A couple of years ago, the company invested in Twitter advertising to increase its number of followers. Karel had worried that having a small number of followers might look bad for a company promoting its social media expertise.
Today, Karel doesn’t see much value in followers as a metric except that it gives him some sense that Bizo is increasing its social audience. What he really wants to determine is the lead quality of followers who are engaging with content – Web tools like those on Bizo’s platform can help him – and which content is most popular among them.
Unique visitors. This reveals how many unique visitors came to your site during a reporting period. This metric, which many Web analytics companies measure by the number of unique IP addresses requesting a Web page, has always had measurement issues. Cookie deletion and shared browser use create complications, said consultant Stephen Rappaport, author of The Digital Metrics Field Guide, recently published by the Advertising Research Foundation. In addition, marketers also need to contend with customers accessing pages from multiple devices that have different IP addresses.
But even a good tally of visitors still doesn’t go far enough for marketers these days. “You can just look at unique visitors and say, ‘We’re getting more traffic and this is good,’ but that doesn’t really help you because you’re not really understanding who those visitors really are – they could be first-timers or repeat or returning,” Rappaport said. “The important thing is to make sure that the measurement you are doing captures as much of the person’s context as possible so that their experience is tailored to them.”
A metrics duo: page views and average time spent on page. These metrics reveal which pages on your site are viewed most often and how long a specific page is viewed on average, respectively.
Marketers might think an increase in page views means a page is doing its job of having people reading its content, but that could be an incorrect assumption, Bennett said. “A page view just tells you if a page is loaded – if someone has clicked a link and landed on a page – but it doesn’t tell you that anything was actually consumed on the page,” she said.
Likewise, marketers may think a long period of time spent on a site automatically means consumers are highly engaged with the content, but that traditional notion of time on site just measures “wall-clock time, measuring when a page was opened and if or when someone clicks to another page or closes the page, not engagement in between,” Bennett said. Maybe, for example, they’ve kept the tab open while they’ve gone for a walk.
A different kind of measurement – like what Chartbeat calls engaged time – can help marketers learn if visitors are paying attention by determining if they are engaging with the content through interactions like scrolling down the page, moving the mouse and clicking the keyboard keys.
Engaged time can also help marketers understand how long visitors are actively engaging with the words on the page. That text could be sponsored content from a marketer, for instance, or news content placed next to an ad.
A brand might want to know if people are engaging with a section of the page with its ad in view for two minutes or 10 seconds, Bennett said. If it’s the latter, perhaps they clicked through to the page because they were interested in the headline but bailed quickly for some reason, she said, possibly indicating the content wasn’t relevant for that audience.