What is this programmatic thing? No, really, what is it?

MediaCom's Willie Pang demystifies one of the industry's biggest buzzwords

Over the weekend, I read Professor Mark Ritson’s opinion piece on the flaws of Programmatic. His depiction of the ‘channel’ as being a dodgy $7 bill was, in my humble opinion, a step too far in demonising what is perhaps the most important technology shift in the media industry. Yes, Mark’s sensationalism makes for a captivating fiction read, but to me, it’s a little bit of the blind leading the blind.

By sitting idly and politely, it strikes me that the digital leaders (people and organisations) in the industry are hoping that detractors like Professor Ritson will simply ‘lose interest’ or that the industry will roll on and their arguments will simply become irrelevant.

The truth is that there is genuine confusion amongst marketers and it’s into this knowledge vacuum that the grenades are being lobbed and amplified. In an attempt to demystify it so that everyone can arrive at their own evaluations from the same set of understanding, let’s start with this:

What is this Programmatic thing? No, really, what is it?

Let’s start with what it isn’t. It isn’t a ‘channel’. It isn’t another line on a media schedule that sits below things like ‘search’, ‘video’ or ‘affiliates’. It isn’t a type of ‘inventory’ (i.e. it’s not a new type of online banner).

It’s also not particularly new. Google (more accurately, probably Overture) was arguably the first, way back in 1998-ish, to allow marketers to buy their inventory (links in the search results page) through an online bidding system. Put simply, the broad definition is that it’s a different method of buying media placements.

Every media placement across every medium will be bought programmatically over the next three to five years.

We, as the buyers, have access to live data on any given piece of inventory and can evaluate how much we want to pay for it. We decide based on a number of factors, including the audience, context, quality, commercial intent, or new or known customers, for example. It’s no different to deciding what TV, radio or outdoor ad we’d like to buy, except that you can make a decision in a matter of milliseconds.

Let me make this unequivocally clear, and if you have any doubts, stroll on over to the broadcast TV networks and ask them. Every media placement across every medium will be bought programmatically over the next three to five years. Yes, that’s every TV spot, every radio ad, every print ad, etc. Some of it will be real-time, matched to an individual user (ah, the dream of the true one-to-one conversation), while some of it will simply be smarter systems helping to make the buying and tracking easier.

So why is everyone so confused and scared, or both?

For a couple of different reasons. The first is cost. Professor Ritson’s somewhat outlandish example of 25 cents in the dollar going to actual media is a way of fuelling the mistrust. Can this happen? Yes. Should it happen? No. Another truth is this: if I was inexperienced or negligent, I could buy a magazine ad for three times what everyone else is paying. Mr or Mrs Marketer, if you really are spending $3 on technology for only $1 of media placements, please call me, I’ll help you to find out why.

The ‘why’ is important. One of my all time favourite movie scenes is from Brad Pitt’s Moneyball. It tells the story of an American coaching luminary, Billy Beane, who helped the Oakland A’s revolutionise the way professional sports teams invest in constructing their squads through statistical analysis. In the conversation, Billy asks his scouting team “what’s the problem?” and after a series of “same crap, different year” responses, he realises, “you’re asking all the wrong questions”.

And folks, we’re asking the wrong questions.

Rather than pull out a spurious set of self-serving statistics to demonstrate that the ‘right’ percentage is 10%, 15%, or 25%, I’ll reframe the question to be about proportionate value.

The point of programmatic is that every target audience set, down to an individual, can be valued based on his or her worth to me, as the marketer. If I’m in the market for a new car tomorrow, you can bet your bottom dollar that car brand X will pay significantly more to speak to me today than what sports brand Y would be willing to.

Finding me TODAY costs money

It takes technology, and people, and smarts, to make it possible. You need to find someone that might be me, triangulate it with other data sources until you’re pretty darn sure, then pick the right moment, in the right context, to give me the right message.

That’s a little bit cliché, but look at it this way: It took me 16 seconds to type this paragraph and in that same period of time, you will have made millions of decisions on whether you’ve found me or not and whether I’m ready to buy your car tomorrow. Millions.

Which leads me to the second reason for confusion. It works a like a treat for marketers who sell a widget. If you can track a response back to a purchase, online or physical, then you’ll spend as much or as little on the technology as dictated by the return. But industry pundits guess that 60% of marketing investment – some $8 billion in Australia – is spent in driving awareness and consideration.

Marketers who are tasked with building brands over time or driving behavioural change are asking themselves “can’t I just do the same thing I’ve always done?”. As a result, with a little fear mongering from the ill-informed, it’s easy to arrive at the conclusion that it’s all cost and no benefit. See Billy Beane’s statement above.

Advanced FMCG brand leaders are moving from just reach-based planning to ‘intelligent reach’.

Advanced FMCG brand leaders are moving from just reach-based planning to ‘intelligent reach’. I’m a big believer of Byron Sharp’s teaching on the importance of speaking to all buyers, in particular, light buyers in a category. And buying programmatically supercharges this.

It means that we can talk to all buyers with a message, a piece of content or an experience that they might actually give a shit about. The human logic is irrefutable, right? We can certainly talk to everyone but let’s take advantage of the capability to Professorive real personalisation.

Professor Ritson does raise a good question here. What should we be willing to pay and how do we measure success? In developing this framework, we are then able to get smarter in deciding which one-to-one conversations we value most and then maximise the returns of the technology investment. I won’t even start on the benefits of using this data to improve the customer experiences we deliver, and the impact on product design, more on this another time.

We absolutely should only pay for ads that are actually seen by a real human.

It’s convenient to drag in other important issues, like viewability and fraud, for instance, but this is not a fruit salad. We absolutely should only pay for ads that are actually seen by a real human. But it’s not a problem for programmatic. If anything, programmatic is the tool that helps make it better, more accountable. You can buy an ad on Foxtel programmatically today and it has absolutely nothing to do with a video ad you see on your Facebook feed.

My challenge to our marketing luminaries is to work with the digital and technology community to help co-create solutions to make it better. And to dispel the myths and make it work for all brands. There is no silver bullet, but I’m certain that trying to make one to shoot one another with is not quite the right approach.

Professor Ritson, the analogy may be more accurately presented as thus: programmatic is not a dodgy $7 bill, it’s Apple Pay.

Get insights in your inbox

Interested in the future of media and marketing?

Get actionable advice, insights and opinion from MediaCom experts every month

Sign me up

Sir Martin Sorrell on Cannes, the ad industry, and 2018 predictions
Cannes Lions 2017: The industry view