Evolving Programmatic: Navigating the Breakers and Riding the Wave


 

As the book (and documentary) “The Human Face of Big Data” so eloquently illustrates, there are both massive benefit and massive risk associated with the relatively new capabilities developed for both storing and processing big data in real time. On the upside, these technologies are being applied to almost every facet of daily life in ways that improve life for the masses; from high impact things like understanding and adapting to changing climates, to finding cures to deadly diseases – to the more trivial by comparison, like finding patterns in shopping habits and delivering relevant and timely marketing messages across consumer devices. The risk is that with so much data about our lives being collected, those who seek only to profit personally from it and care about little else, find this level of exposure ripe, low hanging fruit for fraudulent activities.

Speaking only of marketing now, the fact is that as we go about our daily lives, we are leaving behind a “trail of digital exhaust” that is being collected and stored indefinitely. While this practice is not new (credit card companies and search engines have been doing it for decades) the ability to mine, analyze and react to events that produce a usable digital signal in real-time is still essentially in it’s infancy. We’re still getting our feet under us when it comes to making sense of these signals in a way that is valuable to both our quality of life and to commerce.

Characteristics of programmatic graphic

The challenges we face are many, but to start, most of the datasets we would want to leverage in this capacity are siloed (for security/privacy or proprietarily) and not easily connected and activated upon in a real-time manner. When we try to overcome this hurdle in one area, the ever-growing walled-garden media properties prevent broad analysis, which significantly obscures true insights and ability to activate against those insights in an automated fashion. We also need to be tech-savvy enough to spot high quality machine learning versus that which is not even truly programmatic, then, use that technology effectively to employ data science and rigorous experimentation processes against our assets and institutional intelligence once we’ve separated the wheat from the chaff, so to speak.

These are crazy, exciting times and as with the growth and evolution of any ecosystem, the pendulum swings most wildly in the early going. As marketers, we are being held to higher standards, never seen before in the realm of advertising. Loose “marketing” math and extrapolated data are less acceptable now that we have the means to be more precise, often at the expense of efficiency and return on investment. The opportunity for an advertisement to be seen has been replaced with asks for in-view guarantees (can you guarantee that I didn’t use the restroom during your 5 million dollar, big game TV spot?). Advertisers are being forced (by some) to pay for fraud mitigation along with their technology and media investments – because like with credit card companies, it might still be more cost effective to pay for fraud than avoiding it or removing it altogether.

While we’re approaching convergence of broadcast and digital, there is still much to be done in terms of process and technology to make the final leaps from fragmented, multi-disciplined, channel based approaches to holistic, cohesive and relevant campaigns that make the best use of available technology and talent while driving clearly measurable return on investments. Because, at the end of the day, what we want as consumers who give our data freely to companies, who in turn productize the data and us – aside from implied security of said data – is a return in quality of life (read: service, convenience), just as brands who invest in the burgeoning ecosystem expect an higher rate of return in terms of revenue and measureable brand equity.

Predictive analysis and inductive statistics graphic

As with any ecosystem, if an individual player is a part of the value chain – adding to the collective, more than they are taking out – they will not only survive, they will thrive. Those that are only takers will be seen as invasive or parasitic and will either wither from neglect, or be forcefully cut out and removed by those that are negatively affected by their activity. No amount of marketing can change that, it will only delay the inevitable. As brands and marketers there are a few things we can do once we have navigated the breakers (fraudsters and fakers) and aligned our internal resources to ensure we are going to be able to stay upright as we ride this wave into the future.

First, clean and consistent process for collecting and consolidating data across our media related activity and resulting sales. Next, automating tried and true analysis techniques to model and train the vast sea of data that our media activities generate in order to uncover the insights needed to make smart optimizations, or better yet, automate those as well. And finally, employ experimental design across all media investments to make sure we are balancing our investments in a way the will drive us forward with enough speed to keep the wave from crashing down on us.