Data-driven advertising: A force for social good

The deployment of data in advertising has tended to emphasise new ways of reaching consumers, technology changes and evolving regulations - it’s much less common to see examples of how leveraging data in advertising can be a force for social good.  At Mindshare we follow a goal of Good Growth, meaning its enduring, incremental but also looking at social impact and climate impact. With this mindset we have asked our data and tech communities to focus on how these assets can drive social benefits.

What has come from that is our Precisely Human Intelligence tool suite which is changing the way we think about audiences and audience data, moving us beyond traditional socio-demo and bringing us much closer to the mindsets that drive behaviour.

Bringing this to life in a great example is the partnership between Walgreens and Choreograph, GroupM’s data & technology specialists, to roll out Covid vaccines in the USA. This show cases how data can be used to go far above and beyond traditional ways of addressing audience in digital and, moreover, how such data can be a force for good. Walgreens was not exempt from the pandemic-induced challenges of supply chain disruption, the accelerated shift to online and changes in consumer behaviour, but in the midst of this they also had a key role in the distribution of Covid vaccines.

The scale of the challenge was daunting: deliver 30 million Covid-19 vaccines in only a few months, with a hard deadline of 31^st August 2021.  Trust in the vaccines had to be built, appointments made, and vaccines administered – all in an environment of significant supply chain challenges and where misinformation abounded.

To deliver on this challenge Walgreens partnered with Choreograph, taking on the task of identifying and reaching target people open to vaccination but for whom the right message at the right time could help make the difference between them getting the vaccine or not. What’s more, the targeting strategy needed to align with vaccine supply availability across a huge geography – and all without using any personally identifiable information or sensitive personal health data.

In a great example of how disparate data sets can be brought together to offer new and valuable angles to advertising, Choreograph combined four elements within their data strategy:

First, an approach was needed to predict who would be most likely to consider a vaccine at any given time. The outcome was a model that ingested real-time signals to predict who would want a Covid vaccine - and was able to do so a week before they would otherwise have taken any action. This enabled Walgreens to reach the right audience at the right moment.

Next, the model needed to predict supply chain fluctuations so that messaging was not delivered to people who were not able to access a vaccine appointment due to restricted supply in their local area.

Thirdly, those individuals who were highly likely to never accept a vaccine needed to be removed from targeting to ensure maximum impact by reducing wasted advertising.

Finally, important data sets around demand signals, local infection rates, testing rates and other sources needed to be factored-in in order to ensure the campaign took full account of the context of the pandemic as it evolved.

The result was a campaign that drove a vaccine appointment rate 40%+ higher than previous activities, through 50% more engagements delivered at a CPA 20% lower, demonstrating the power that overlaying different data sets can have.

The sheer number of different factors to consider (for example likelihood to consider a vaccine was scored against 125 different attributes for each individual) along with the supply chain and external dynamics, would have been operationally impossible not all that long ago and remain far beyond the capacity for any human operator to make sense of and optimise towards in real-time - but predictive modelling approaches like the one developed by Choreograph allow advertisers to take in far more variables than ever before and deliver advertising that is ever more efficient.

The Walgreens example shows how data sources not typically brought together can be used to deliver a highly targeted campaign that drives great results. It is just one example – but many brands are sitting on valuable data sources within their wider business that are yet to be exploited for marketing. Meanwhile, increasing use of cloud, APIs, clean rooms and other technologies has made exploiting external data sets beyond the business significantly more accessible. Bringing these data sources together to inform advertising can unlock value, efficiency and greater ROI while creating a differentiated approach using data not available to competitors.

To return to those changing industry dynamics, the challenge today is how to achieve a data-driven strategy like this within the context of technological and regulatory change, ever-evolving consumer behaviour and increased consumer sensitivity towards how data is used. This isn’t a zero-sum game though, we can still deliver well-targeted, personalised campaigns within the new ecosystem we’re moving towards and, if anything, factors like the demise of cookies have provided the impetus needed to explore new ways of addressing users.

The successful data-driven approaches of the future will be those that embrace ethical considerations like consumer sensitivities around data usage and harness previously untapped data sources to build innovative, differentiated approaches to engaging consumers.

Morys Ireland, Head of Data & Technology Services