Bringing brand data to programmatic

The question of how we measure advertising success is not new.Whether it is about an immediate response, building long-term brand measures, or somewhere in-between, is an age-old debate.

Right now the trend is toward short-term behavioural response as the key measure of success for advertising. This is not just because of digital innovation. Sales promotion and direct marketing have successfully pushed the performance marketing approach since the 1970s.

Programmatic media buying has underpinned and reinforced this approach. With more and more media bought and targeted using automated systems, the question of what represents success has become more urgent.

The central argument for programmatic and the data-driven media buying revolution that powers it is the ability to hit the right person, at the right time, with the right message.

At Millward Brown, we believe successful advertisers need to use brand relationship data to inform the inputs to the campaign as well as using it to assess the long-term contributions to brand measures. In particular, brand data needs to play a key role in both targeting and creative development.

To examine how brand can play a role in programmatic media buying we need to address each of the elements – time, person and message – in turn.

What is the right time to advertise?

Today’s programmatic systems use behavioural ‘big data’ that can be tracked in real time as proxies for interest and intent. They are often selected because they are what we can measure, rather than what we should measure.

For example, searching or shopping may represent a “declaration of intent”, while site visitation history may identify an interest.

This emphasis on the behavioural indicators of context can go badly wrong. If I search for a Volkswagen Golf, is it because I am in the market for a new car, or because I just happen to like old Golfs?

If I am shopping at a retail location or online will I be more open to messages regarding the brands that are available? It may be true sometimes, but often such ads will be seen at worst as intrusive, or at best clutter/wallpaper.

This narrow approach to identifying the right time or context ignores the learnings of generations of media planners, who have long used factors such as social setting, mood, emotion and intrinsic media perceptions/usage to refine placement.

We cannot know what moods and emotions are being experienced by audiences consuming media, but we can make assumptions about increasing the probability of tapping into relevant ones by selecting key programs or dayparts.

Brands that have associations with happiness and conviviality would typically be placed alongside TV programs with similar associations, or where people are likely to be watching in groups, for example sporting events or The X Factor. Social consumption of such programmes may enhance the mood and receptivity to advertising.

The fact that some media are more influential than others, and that this influence may vary at different points in the consumer decision journey can be important. The presence of a brand in a specific media channel may have a great influence on its inclusion in the consideration set.

The programmatic search for the moment of truth is based on the idea that the closer to the decision a message arrives, the better it is at influencing decisions.
That people “make their mind up at the fixture, or even when they are searching”.

The truth, of course, is that ads closer to the point of purchase do not drive the decision to change behavior alone, but “harvest” attitudinal response created much earlier.

The misattribution of brand power in favour of location and activity represents a bigger issue than even “last click wins”.

Who is the right person?

Programmatic systems promise the earth about reaching the right person. They tell advertisers about the hundreds of data points they collate to ensure that messages are highly targeted. To be really effective, however, we need to understand the relationship that the target has with the brand being advertised and not rely on behavioral proxies of interest and intent.

We know that a strong brand relationship will influence the attitudinal response to advertising. It may also shorten decision-making time for purchases. The frequency with which we advertise to such consumers will be lower, and the messages we convey would be different.

For example, a consumer who is a fan of Apple will only need a small number of messages to ensure they stick with it when their mobile contract comes up for renewal. However, someone who is not an Apple fan would need considerably more effort to ensure that the latest iPhone earns a place in their consideration set.

Getting the message right

We have already noted how aligning content to the mood and emotion of consumers will impact the effectiveness of advertising. Similarly, adapting the message to suit the brand relationship of the target will also have an impact.

However, for all the talk of programmatic media buying having the ability to adjust the message according to the target, little brand data is currently feeding this process.

The role for brand data

Understanding the brand relationship is fundamental to both targeting and message development. Without this, it would be very difficult to target the right people and tailor the right message. Purely focusing on the right time and, worse still, proxies for right time, will not create effective marketing.

Our challenge is to make available brand relationship data that can facilitate the right approach in a programmatic context.

As researchers, we are concerned about the importance of understanding brand outcomes of advertising. For marketing directors, however, the lure of short-term sales and behaviour changes are more tangible and translate straight to the bottom line.

We need to focus more strongly on how we can provide inputs into the advertising planning process and demonstrate the value of brand relationship in targeting the right person with the right message.