Back to Basics: Knowledge and Power in the Algorithmic Age

Ask yourself how well you know your target audience. Not how well your demand side platform (DSP) knows them, and not how well you know your customers – I’m talking about your potential target universe : all of the people you’d like to speak to about your products and services. Do you know how many of them are in the U.S.? Gender split? Average age? Favorite TV shows? Which marketing medium they trust the most? Favorite thing to do online? Activities they do offline? How many kids do they have? Pets?

Don’t feel bad if you can’t answer these questions – in the age of programmatic digital display and social, most advertisers are actually further from understanding their target audiences than perhaps ever before. In some ways, they don’t need to know them: platform algorithms can find your best audiences (in terms of digital engagement or response) in fractions of seconds and optimize entire programs to them. My fear is that we’re A/B testing ourselves into happy complacency. Digital ads and marketing technologies are powerful stuff, but they are increasingly closing the blinds on what we actually know and understand about targets and what makes them tick. We’re substituting performance for knowledge and it’s beginning to blind us to basics.

Let’s take a simple example: you know your most recent digital program “reached” 100,000 unique users. You know it drove a ton of efficient traffic – maybe it was your best campaign ever! You may even know that the average frequency of exposure to those 100,000 was 6x – because you served 600,000 highly targeted impressions. You also know you have 10,000 customers. Does this mean your prospective target universe is 90,000 or 100,000? Either might be right. Problem is that your next campaign might serve a million uniques at the same or a different frequency. The point here is that in most of the current ecosystems, we never get quite enough information to say how or if we’re really reaching our potential audiences. That is to say, does my customer base, or my media programs represent 10% of the audience opportunity or 80% of it? If it were my company, I’d want to know. And honestly, I’d want to know even more than that.

Most of the intelligence we get from A/B testing our way through digital and social platforms are behavioral in nature, and there’s an obvious propensity for the algorithms to seek out that 2% of the population that would even click on an ad in the first place. Back in the early days of digital advertising, we’d call this “spray and pray” – we’d buy as many impressions as we could and simply shed the impressions that didn’t yield desired responses or actions. No one appeared to be too bothered by the fact that we bought those ineffective impressions in the first place, and folks were satisfied that agencies seemed to be proving their worth by “optimizing” away from those impressions week over week. The only real difference today is that this all happens really fast – like in nanoseconds. Ever notice that your typical paid Facebook campaign has a super efficient CPM (cost-per thousand impressions)? That’s because Facebook has to serve an insane volume of impressions to elicit the response that they sold to you. And they don’t share back much about those responders and the cohort they came from – that’s up to you to define and articulate.

So what happens if my audience likes to shop offline? Or if they spend more time watching TV than surfing the Internet? What happens if the things they have in common with each other, and perhaps with you, aren’t captured by the algorithms or (worse) are captured but not shared back? Then all you can do is optimize to the behavior, fumble your way through messaging, and try to determine the context in which that message is delivered.

This is where Ideas Collide steps in. It’s not that audience intelligence isn’t available or that it’s wildly expensive – it’s that agencies and clients have allowed the power of performance to act as a proxy for knowledge. We’re pleased to announce that Ideas Collide now has access to GFK MRI – a long time gold standard survey of American Consumers. It is projectible to the entire US adult population and allows us to define, size and compare target audiences utilizing demographic, behavioral and psychographic parameters. It helps us predict where to find your ideal audiences and gain insights into what we should be saying to them. We’re committed to taking back the high ground on knowledge and it starts with understanding your best prospective target – right down to the cereal they eat and the cars they drive. Because even in the face of advanced technology, one thing remains true: knowledge is power.