Impressions are a surface metric that don’t provide much actionable information. PR professionals are starting to use “opportunity to see” as a more accurate descriptor of the metric, because that’s really what the number calculates: the number of people who had the potential to read the story, hear the message, or see the Tweet.
Still, impressions continue to persist as a unit of measure in PR and communications work, and even advertising and marketing use the term.
So, what is an impression, and are they remotely useful, or just one of those numbers that seems to stick around out of habit?
Impressions are everywhere
There are media impressions, advertising impressions, mobile video impressions, impressions in affiliate marketing, and marketing impressions. There are social media impression figures too—Facebook, Twitter, and Instagram—all of these platforms offer measurement of impressions.
If this is such a worthless metric, why is it so widespread?
Impressions are widespread for a few reasons—they are fairly easy to calculate, don’t require much explanation, and they’re generally big numbers. People spending money on programs like to see big numbers.
The problem, of course, is that these really aren’t very good numbers on which to base a conclusion of whether a program or effort is working.
Which means that impressions are sort of like those weather forecasts that go out a month or more ahead in time. They are interesting and do contain some small nuggets of information, but you should use them as rough guidelines—not for detailed planning.
When we consider the “opportunity to see,” it makes sense to take three things into account: the medium (print, online news, or social channel); the audience; and the content.
These elements all matter in how, when, and where the opportunity to see is presented. In this post, we’ll cover the medium—how impressions are defined, and what to watch out for if you choose to include impressions as a metric.
A straightforward way of calculating impressions for print media is to just use the circulation rate. A circulation rate is defined as the number of printed copies distributed by a particular publication.
If someone is a print subscriber to a newspaper or magazine, the potential for an article to be seen by that subscriber is fairly high.
This is also what I would consider to be the most transparent and logical way to calculate impressions.
At some point, the concept of “multipliers” started to be applied to print impressions.
Since print materials can be passed along, the reasoning followed that more than one person could receive an impression of an article. If a household of four has a subscription to a daily newspaper, there’s a chance that all four could be reading the paper every day.
Essentially, multipliers are used to estimate this “pass-along” rate.
While the underlying theory is hard to object to, multipliers have always made me a bit queasy.
Sure, that People Magazine in the dentist’s waiting room might have 10 people pick it up over the course of a week, but there’s no guarantee each of those people read the magazine cover-to-cover (in fact, it’s unlikely—depending I suppose in part on the efficiency of the dental practice).
A newspaper is even less logical to use a multiplier on, as some people read the front page and opinion sections, while others go straight to the sports page.
Essentially, circulation numbers are a broad indicator of how many people had access to the content. We have no solid data on which stories people read.
Multipliers are sketchy; if you’re going to use impressions at all, go with straight circulation numbers, preferably audited numbers from one of the independent organizations that validate these types of numbers, such as the Alliance for Audited Media.
Websites of news organizations, blogs, magazines, cable and broadcast TV news channels and so on all receive web traffic and therefore generate web traffic numbers. One of these is known as “Unique Visitors per Month,” or UVPM.
A UVPM is the total number of visitors to a website per month; someone who visits twice, or multiple times, is only counted once.
Because the detail on a site’s actual traffic numbers is accessible only to those with access to the analytics linked to the site, general traffic numbers are what is typically readily available to a PR practitioner. So, we might know that the New York Times had 72.9 million unique views in January of 2016, but only the Times knows exactly which articles were read.
This presents a digital version of the problem noted for print: we do not know which stories are read, but with UVPM, we do know how many people potentially saw a story.
You start to see the problem with using UVPM when a story hits several major online publications.
For example, if you have a client story that gets picked up widely on a number of major newspapers’ online properties and a few blogs with large readerships, it’s entirely possible to have UVPM numbers in the hundreds of millions.
I once saw a report that had combined UVPM and impressions number nearing 4 billion. It’s ridiculous to report a number that suggests nearly half the people on the planet saw this particular content/story, but that’s what happens when these types of numbers are used.
Social platform analytics are improving over time, but impressions are still sort of a flimsy metric. On Twitter, an impression is defined as the number of times users saw a specific tweet. In Facebook, it’s the number of times an ad is viewed the first time. On Instagram, it’s the total number of times a post has been seen.
Knowing and understanding how each communications medium defines impressions is important, as it provides the context through which you should view the numbers.
Impressions—the “opportunity to see” an article, post, or social share—are a very generalized metric.
Having an understanding of how many people had the opportunity to cross your content is useful to know, just as it’s important for a brick-and-mortar retailer to know what the foot traffic is in a store: it gives you a general idea of what the upper end of your audience might be.
The numbers that actually matter: those who read and absorb the content or take action are your true metrics.