Verity Gill, Digital Director at Ebiquity, explains why advertisers should work with publishers and agency partners to develop metrics that measure those harder-to-find outcomes in place of readily-available outputs.

Digital media promises so much, particularly in terms of measurement. The wealth of data available offers the opportunity to track customer journeys with forensic accuracy, making marketing activity suddenly much more accountable and transparent. And while there has been a lot of progress in recent years in the quest to understand target audience behavior, many advertisers have also been prepared to settle for metrics that are easily acquired rather than those that measure actual impact.

There are two major challenges that face advertisers looking to create clarity around digital measurement today. The first is the ‘walled garden’ placed around some of the biggest user data sets potentially available, held by platforms such as Facebook and Google. The second is the apparent resistance to change from some of the more traditional players having migrated to digital – including broadcasters – in enabling third-party tracking across video content. The challenges of developing accurate and truly meaningful metrics of online effectiveness have led some agencies to offer (and some advertisers to accept) simpler output metrics. One notable ever-present is the ubiquitous metric of ‘reach’. We believe that the time is right for all parties to dig a little deeper.

Advertisers looking to increase brand awareness through digital advertising have been encouraged to use impression delivery as a measure of increased reach. This is logical enough, and resonates well with marketers familiar with TVRs and print media impressions. The challenge is that digital impressions only form one element of the awareness equation.

A familiar refrain is heard from advertisers across the world asking why, having invested in a campaign to deliver so many millions of impressions, these impressions are failing to have any meaningful impact on brand health measures or brand awareness. They’ve been reassured by industry experts that “these things take time”, “the medium is young and everyone is still learning how it works … so you need to be patient”.

So how do advertisers get closer to meaningful measures of digital advertising performance? What is the best way to measure outcomes and not simply outputs?

Reach is where the digital media industry started in terms of measurement and evaluation. Advertisers had grown used to the term representing more than just a simple number of impressions in traditional media. In TV, for instance, reach is a much more structured and nuanced KPI than it is currently in digital. With TV, a reach figure gives a percentage of coverage against a particular target audience – say 75% of 18-24s hit with a particular spot. With TV, reach is also accompanied by a measure of frequency over a particular period, too – so 75% of 18-24s hit an average of eight times over a two-week campaign, for example.

With digital, reach is currently simply a measure of the number of impressions potentially delivered. No frequency, no demographic detail, and no time period comes as part of this package. A campaign with a reach of one million impressions is just that: the campaign had the potential to be seen a million times. It’s just a measure of the number of shots fired. Reach in digital needs to start evolving. And fortunately for advertisers, it is.

The step on from simple reach is unique reach – that is the number of different individuals (or, more correctly, individual devices) who saw an ad. Unique reach considers the campaign from the audience perspective. No longer is it just the number of shots fired, but how many ‘users’ were hit by those shots. We use the term ‘users’ loosely here, as unique reach is based on cookie- or device-level data – not the ideal user-level data. But at least with unique reach we are one step closer to traditional media metrics that de-duplicate activity, representing both reach and frequency, like GRPs and TVRs, for instance.

As anyone who has ever invested in digital advertising knows, viewability is a major issue and one that can negatively affect a campaign’s reach and therefore its impact and outcome. The next stage on from unique reach is therefore unique viewable reach. Unique viewable reach is a measure not just of how many impressions were delivered to unique IP addresses or cookies, but of how many of those impressions served were actually delivered on the screen.

From an effectiveness perspective, the standards of viewability are still relatively low compared with those of traditional media. The Internet Advertising Bureau (IAB) counts an ad as viewable if – in the case of banner ads – 50% of the pixel surface of the ad is seen for one second; with video ads, an ad counts as viewable if 50% of the pixel surface is viewed for two seconds. Putting this into context against TV is important; counting ‘opportunities to see’ on the traditional media means that fractions don’t count and success is all about the quality of placement (e.g. position in break, peak access) as well as the fact that the audience that has been reached.

Issues facing digital that are absent from those measuring TV include: ads being played below the fold; ads auto-playing but not being engaged with; ads being served and counting as impressions, but not actually loading; and the intentional layering of ads – as well as apps – being served on top of one another for fraudulent purposes. So, while a million impressions may have been bought, only half of these may have been served to unique users, and only 40% of these may have been actually viewable. And yet advertisers may have been charged for a million impressions, irrespective of whether they were served to unique users or if they were even seen. This is why viewability is such an important issue for online display and video advertising compared with, say, TV.

Other issues of viewability that are now, by and large, well tackled, include:

  • Ad-blocking: advertisers are typically not charged for ads that are served but blocked by viewers’ ad-blocking software.
  • Player size and inventory type – users often don’t have their video players optimized for user engagement, and so ads and branding can easily lose impact. Ad servers now provide player-size reports which give an understanding of what proportion of ads were seen at what size, and there are also opportunities to evaluate real engagement with formats through eye tracking technology, such as that provided by Lumen Research. 89% of global advertisers recently surveyed by Ebiquity and the World Federation of Advertisers (WFA) said they intended to increase spend in online video advertising in 2017[1], meaning that it is even more important for them to embrace the latest measurement tools available.
  • Skipping – ads served on YouTube, Facebook and other popular video sites are often skippable, but advertisers can also see reporting on the proportion of their video ads that were skipped and at what time point, which provide incredibly valuable insight at a user and creative level.
  • Sound-off – it had been suggested that sound makes very little difference to a user’s engagement with an ad, and so having sound switched off by default did not affect brand engagement. Publishers are now moving away from this position because they have been persuaded that ads made with sound need that sound to be played to have their intended impact.

The state of the art in terms of reach – and the ideal KPI for where the digital advertising industry is today – is human, audible, viewable, individual, on-target reach; HAVIOT-R. Catchy I know. This is a measure of the number of ads served to the right kind of real person, targeted by the brand, and those people have actually seen and heard the ads. Independent ad verification platforms including MOAT, IAS, and Double Verify provide the majority of components for this advanced reach metric, but most of the industry has yet to embrace these opportunities fully.

The necessary progress from simple reach is symptomatic of an industry growing up fast. Advertisers, agencies, publishers, and platforms are all playing a part in creating more meaningful KPIs that reflect what advertisers are looking to achieve with digital advertising, not simply what can be measured. Every player in the media and marketing ecosystem has a role to help take measurement to the next stage of its ongoing evolution. Specifically:

Advertisers should ask for more. They should look at every KPI they are offered and ask whether it’s the best option that is technically possible, a meaningful measure of what they want to achieve, and can also be consistently and continuously reported on.

Publishers and platforms have a major responsibility to allow third-party tracking. It’s very encouraging that both YouTube and Facebook have responded to the challenge thrown down by P&G’s Marc Pritchard earlier this year and are introducing third-party verification. All publishers must also evolve their inventory supply in line with findings of the tracking in place.

Media agencies need to ensure that they’re sharing everything they know with their clients and should work closely with publishers to enable them to track genuine marketing effectiveness.

Together, all parties can deliver and now are delivering better, more meaningful metrics.