The age of the programmatic CMO – why aren’t we there?
Let’s face it: for all its noise, the programmatic engine is stuck in first gear. With its detailed data on individual customer interactions, this engine has the power to unlock a long-awaited transformation of marketing from black art to data-driven science. But so far, the transformation is barely underway.
Instead, programmatic has delivered improvements to the mechanics of ad purchasing, and not much else. That’s none too surprising, actually, since most of the data never makes it from marketplace to marketer.
The advent of programmatically purchased digital impressions offers today’s CMO a world of information unimaginable to previous generations of marketers (I know because I was one of them!). Who received each impression, when, where and how? Did they respond in any way: view it, click it, interact with it? This is precisely the kind of data we need to truly understand our customers and craft marketing interactions that are welcomed by them, not ignored or blocked. Why, then, is this critical data not getting to marketers?
All too often, we don’t even ask for it. Many marketers think this is a level of detail best left to the agencies, trading desks and DSPs running our campaigns for us. This is a shortsighted perspective, and leaves the history of your customer interactions solely in the hands of others, to whom you are now increasingly dependent. The periodic, summarized reports on campaign performance typically provided by these intermediaries are a far cry from the granular, searchable data on which data-driven marketing depends.
For another thing, marketers may not even realize what they’re missing. They may be working with a data management platform and assume they’ve got all the data they need. Isn’t that what DMPs are for? Indeed, DMPs play a vital role in programmatic marketing, helping marketers profile their desired audiences and purchase those audiences in digital marketplaces. But most DMPs were not designed to manage the granular impression-level data needed to track and analyze audience behavior – that’s a technological horse of a different color. In fact, the technology needed to do this job has only recently surfaced, with advancements such as streaming data stores and huge reductions in data hosting costs.
But there’s one more reason why programmatic data can’t yet be effectively put to use by marketers: today’s world of programmatic is a Tower of Babel. For all the talk of Google and Facebook squaring off for dominance, in reality the programmatic marketplace is incredibly diverse and growing more so each day. One large DSP buys from 50 different SSPs; yet another buys from 60. Larger publishers want to run their own private marketplaces.
For each consolidation, another new marketplace appears: new countries, new ad types, and new publisher consortia. While there are standards for how to place bids in an auction (thank you, IAB), there are no such standards for reporting. Each exchange has been left to invent its own way of reporting on available inventory and transaction results. So even if the data were freely flowing from marketplaces to marketers, that data would be hard to make sense of, coming in dozens of different formats and content types.
So, what’s a CMO to do? To start with, it’s time to start a dialog with whichever service providers you work with to plan and run your digital campaigns: agencies, ATDs, ad networks, or DSPs. They may not be able to deliver the data you’re seeking right away – they may not even have it themselves. But don’t give up. Change only happens in the presence of heat and pressure.
Second, decide how you will store and access the impression-level data once you get it. In some cases, you may choose to let the data reside in a system built and operated by a third party, such as an agency. If so, make sure you are given full, granular access, not just periodic reports, and make sure you have appropriate ownership rights to the data from your own campaigns.
If you decide to build the system yourself, help is available from a growing number of providers who offer streaming analytics at scale. Just make sure to take one bite of the data elephant at a time. Start with a limited set of data sources and a well-defined, constrained set of use cases to support. Otherwise, the volume and complexity of the data may swamp your project before it even gets off the ground.
Finally, become an advocate for reporting standards on programmatic events. CMOs need to weigh in to assure that these standards get built and adopted by the ad tech industry that must provide the data conduits from programmatic marketplaces to marketers. Those standards must be designed to allow the right information to flow to all the parties who need it: DSPs, trading desks, agencies, and most importantly, to the marketers who paid for the impressions and ultimately have a right to the associated data.
The age of the programmatic CMO is finally within our sights. In the end, it will help us do what marketers have always tried to do: understand our customers and figure out how best to engage them. But with the rich data now available to us, we can achieve a whole new level of outcome. Getting there won’t be easy, but the results will be well worth it. Now it’s up to each of us to get there.
Deb Rieman is executive chairman at Metamarkets