“On TV And Video” is a column exploring opportunities and challenges in programmatic TV and video.
Today’s column is written by Nico Neumann, senior research analyst for programmatic marketing strategy and analytics at the University of South Australia.
For many years, advertisers and publishers have discussed whether digital media buying should rely on the same measures as traditional media, namely gross rating points (GRPs) and other reach and frequency estimates.
Indeed, several ad powerhouses, such as Facebook, TubeMogul, Comcast and Nielsen, have introduced GRPs specifically for digital ads and videos to enable a comparison to TV campaigns. However, is this the right approach? Are GRPs best suited to evaluate campaigns and buy media on?
GRPs have been the standard currency for broadcast TV as well as radio to plan, buy and measure advertising campaigns since the 1950s. Nevertheless, despite the historical success and the widespread use of audience ratings, it is crucial to understand the pitfalls of GRPs and reach measures in general.
TV Audience Ratings Can Be Inaccurate
We have seen fierce discussions about viewability issues in digital channels. Yet, TV audience measurement has its perils, too. For example, in the US about 30,000 households have Nielsen meters. This group is supposed to accurately represent the 110 million households with TVs. But analysts have frequently criticized the sizes of these samples as being too small, particularly when drilling down to segments, individual shows or local markets.
In addition, household members need to manually indicate how many people are watching TV at any time. Given that incentives are fairly low (typically reward points), one must wonder how often people adjust the meters to record who is really in front of the screen when one ad or show is aired.
Bear in mind that any small measurement error at a panel household is extrapolated to the population, so one can easily be off by thousands, if not millions. Hence, the question: Should a fuzzy measure like TV GRPs be the benchmark for all marketing metrics?
Reaching People Is Not The Ultimate Goal Of Organizations
Let’s be clear about the objective of communication efforts from organizations: They want to foster some action, such as a sale or behaviour change. In contrast, GRPs measure how many people of the target audience are reached and how often on a single medium, while neglecting cross-device connectivity.
This is one interesting diagnostic, but ads only represent opportunities to see a message. Whether someone paid attention and whether the perceived message was successful can only established by linking campaigns directly to sales or final actions.
And there should be no exceptions to doing so. Sometimes brand marketers and media planners distinguish between performance and brand awareness campaigns. Yet, we need to recall that even brand awareness has a direct impact on consumer choice – it’s what makes people keep buying a brand tomorrow even if it switched off all marketing communication. The long-term effect of advertising nurtures the baseline sales of a brand and can be estimated with various attribution or marketing mix modelling methods.
The Focus On Reach Fueled Ad Fraud
The historic focus on reach measures in media buying let many organizations use similar KPIs in the digital space, such as impressions served or unique cookies. Needless to say, these measures can easily be gamed or faked.
Cookies can be created and hit in an infinite number. Even if you use cross-device graphs, you may not be safe: IP addresses, user profiles and emails can be manipulated by software, too. In the digital world, it is extremely easy for bad actors to artificially inflate reach measures. However, one thing bots don’t do is buy products. Therefore, linking campaigns to sales through analytics instead of optimizing reach is more prudent from a fraud perspective.
Moving From Quantity Toward Quality
Concentrating on the pure number of people who were potentially reached does not consider that some ads may be more powerful than others. Let’s assume one ad reached 20 people, a second ad reached 50 people and your attribution solution suggests that each ad contributed to five converters. Since the first ad achieved the same outcome by reaching fewer people, it is more effective.
Only when linking advertising to the final goal of the organization, we do justice to the value of various creatives and the quality of different communication pieces. Of course, this example also illustrates that reach is still a necessary diagnostic to track, but it needs to be interpreted in the context of the final outcome measure and not by itself.
Make TV A Fully Accountable Medium
When TV advertising became popular about 60 years ago, audience ratings were basically the only available and technically feasible metric. Today, reach and frequency rather seem to be outdated ways to optimize media buying. Digital measurement has allowed us to analyze media campaigns’ impact on the final outcome, both short- and long-term.
While this has been done in only in small scale for TVs campaigns so far, set-top boxes, smart TVs and IP TV allow for integrating TV watching data (in combination with user logins) into the digital customer journey.
The more we can investigate all channels using individual exposures, the better we can evaluate their actual influence on customers from an omnichannel view. And TV may get a chance to prove itself, perhaps turning around the trend of losing money toward digital channels. For example, an attribution study by professors Bryan Bollinger, Michael Cohen and Jiang Lai found that TV advertising is by far the most effective medium in affecting consumer demand.
Learning From Programmatic
Media buying based on quality and sales impact will require a mindset shift among publishers and broadcast networks, who need to understand that the same spot and audience will have different value to different buyers.
This is the cornerstone of value-based pricing. Someone who is in the market for a car is more valuable to a car manufacturer than to other advertisers. So why should automobile brands not pay more for such a targeted spot? This is a positive aspect of the programmatic and addressable targeting paradigm, which any advertising medium should strive for.
Fortunately, we have the technology ready to change the TV and other media buying procedures toward a meaningful, accountable approach. Some cable companies have started offering targeting capabilities and measurements based on addressable TV, whereas bigger networks are hesitant to change their existing business model. Therefore, it is even more important to increase pressure from the buy side and overcome the tradition of reach-based media buying.
As the pioneering computer scientist and admiral Grace Murray Hopper reportedly once said, “The most damaging phrase in the language is ‘We’ve always done it this way.’”
Follow the University of South Australia (@UniversitySA) and AdExchanger (@adexchanger) on Twitter.