Plus ça change, plus c'est la même chose...

The more things change, the more they stay the same. It is as true in audience measurement as in so many other areas of life. Traditionally, the definition of the ‘audience’ has been somebody with an ‘opportunity’ to see (or hear) an advertisement. Digital media has promised far greater precision than this. But arguably they too are delivering ‘opportunities to see,’ and not verified ‘views’.

So for television, the opportunity arises if they are in the room with a television set tuned to a channel where a commercial plays. For radio they normally should have recalled listening to a station for a number of minutes during a given quarter hour. And for magazines and newspapers, they have to remember having read or looked through any part of any issue of a particular title. These ‘opportunities’ are the basis for media planning and trading.

It is easy to see why audience measurement developed like this. Any TV or radio broadcaster or any newspaper or magazine publisher is selling space into which an advertisement will be slotted. Whether or not that advertisement is successful obviously depends on it being seen or heard. But, equally, it depends on whether the audience is in the market for the product being advertised, whether the creative execution has impact and whether they are in a receptive state of mind to take note of it amidst all the other content bombarding them.

Digital media have taken aim at these metrics, calling them vague, indirect and impossible to use for calculating or improving ROI. Digital media, in contrast, are inherently measurable. Every page load of an ad is captured. The ‘journeys’ (sites visited) before and after an ad has been viewed can be tracked. Messages can be targeted precisely at people whose online behaviour indicates have a high likelihood to purchase. We are, in short, approaching a measurement nirvana.

If only it were so simple. We now know that the digital audience count includes many who do not or cannot see the ad which they are attributed as having seen. One reason is fraud – a study in the United States last year estimated that 11 percent of display ad views and 23 percent of video ads were watched by ‘bots’ – essentially machines programmed to act as if they were people.

Viewability remains an issue. An advertising banner can appear anywhere on a web page. But many web pages are bigger than the screen on which they are viewed, meaning users cannot see all the content at once. If they don’t scroll down – and the ad appears outside the viewable space on the browser – it will not be seen. If a user has multiple windows open, but is not looking at the website where an ad has loaded (i.e. it is not in the active browser) this too might be counted but never seen.

There are also cases where ads or videos are stacked on top of each other or rotated very rapidly so that real people looking at the page cannot actually see any of them. In all these cases, there is no opportunity for people to see ads; they are invisible.

Technology and methods exist for subtracting all these essentially false hits – we can isolate and exclude fraudulent hits in many instances; we can count only those ads appearing in an active browser and we can define ‘viewable’ impressions in various ways which increase the chance that they were actually seen.

But even having verified human eyes on a screen does not guarantee they will be attended to or even seen. There are many ads competing for attention. Often, users are moving quickly through websites looking for specific information and not attending to the ads at all.

Sir Martin Sorrell recently dubbed Facebook’s ‘three-second rule’ as ‘ludicrous’, This is where the company counts as viewers anybody who happens to encounter - as they are scrolling through the site - a video for at least three seconds (which will only play with sound if they are clicked on).

So, when it comes down to it, is a digital ‘view’ any closer to being a precise measure of the number of people seeing an ad than an ‘opportunity to see’ measure that captures the number of people in a room with a television set tuned to a channel when an ad appears?