Can a lopsided focus on where ads run distract us from the real source of marketing waste? Jake
Moskowitz from the Emodo Institute uncovers the imbalance between media and data that’s limiting the
effectiveness of advertising.

Watch Now

For years we’ve developed media tactics and metrics to control the platforms, context and physical
location of ads. Today we can ensure that our ads run on the sites and apps we want, near the content
we approve of, in a place where they are sure to be seen, to deliver real — not fraudulent — impressions.
Unfortunately, our fixation with measuring and regulating ad placements comes at the expense of the
accuracy of the data we use to choose those ad placements in the first place!

When it comes to data, for too long we’ve just assumed it’s right. Without verifying information like
segment designations, bid request metadata, and attribution studies that measure how well campaigns
are driving toward KPIs, marketers are basing investment decisions on inaccurate information and coming
to inaccurate conclusions.

What we need is balance. It’s imperative to put as much time, energy and focus into verifying the data as
we do on where our ads run.

Video Transcript

Hi, I’m Jake Moskowitz, head of the Imodo Institute. I want to describe a dichotomy that we see in the marketing industry, that we feel a lot of people are missing. It’s a dichotomy between media and data, or another way to think of it is where versus how.

In the marketing industry we are obsessed with where ads run and we have tons of different ways to limit where ads run, measure where ads run, make sure ads ran where we wanted them to run.

Some examples include whitelists and blacklists of sites or publishers or apps where we either are okay with our ads running or not okay with our ads running. Another example is brand safety.

It’s about where ads are running with regards to the context of the information that’s being shared where our ad is running next and where our ad is running. Viewability is another example, so with viewability again it’s about the where our ads are running in relation to the screen and what’s being shown on the screen at some given time on a consumer’s device. So did I run above the fold, below the fold, did my ad ever make it onto the screen as the user was scrolling down the content? It’s again about where ads are running and finally fraud is also something I’d put in the category of “where,” in a sense that it’s about where my ads are running in terms of which devices it’s appearing on. Is it showing up on real devices or fraudulent devices? Is it showing up on real apps or websites or fraudulent ones? Are the impressions being served fraudulent in nature or are they real and seen on real devices?

All of these metrics that we’ve spent years obsessing over in the marketing industry have to do with where ads run. So the point that I want to make is that while we are so focused on the where ads run, we are spending far too little focus on the how ads are running. Not the where but the how.

Another way to think about that is the data that we base our decisions on and there are a bunch of different examples of data that we are using to base marketing decisions on. One is segment designations. When we see, for instance, a bid request in the programmatic ecosystem, we check first and see if that device belongs in a segment that we’re trying to reach. And, we make a bidding or a buying or a serving decision based on whether that device or browser or cookie or whatever it might be, is in a segment or not that’s data that determines if that device sits in a segment or not and so we’re making a buying decision based on data in that case.

Another example is metadata within bid requests themselves. As an example, let’s take lat-long data. Some campaigns are geo targeted so you either want a target within areas that our brand operates or geofence around stores or you want to hit people where they work a bunch of different cases for using lat-long data. All of those are metadata within a bid request themselves that the bidding engine is making a real time decision based on information in the bid request to determine whether we should bid on that bid request or how much to bid on that bid request. So, that’s again data that we are basing our decisions on.

A third category, maybe the most important of all is attribution data. Attribution studies, attribution tracking that we do to measure the extent to which our campaigns are working and driving the KPI’s that we’re looking to drive. Whether they be online KPI’s like conversions, or offline KPI’s like in market sales or in-store traffic lift or brand lift, any of these are things we measure using attribution studies. And, we use these attribution studies to determine should I move money from that platform and put it on this platform? Should I move money from this publisher to that publisher? Should I move money from mobile and put it into TV or TV to mobile?

All of these are big decisions that we’re using based on data. The problem with all of these examples with the “how” ads are running, is that all of them are so dependent on data. And, we don’t have good ways of figuring out if this data is accurate.

In fact, really the problem in this industry is that we just assume that the data’s right almost in whatever context we’re using data. Whereas on the media side we would never make the same assumption. We would never assume ads are viewable, we would never assume devices are non-fraudulent, we would never assume that a URL or a path is okay for our ad to run in or we would never assume that content is okay for our ad to run next to but on the data side we assume assume regardless of whether it’s segment data or bid request data or attribution data we just assume it’s right. And I think the reason why that is, is because when the verification revolution happened in the last five to seven years, the beginning of that a lot of media was bought on a managed service basis and therefore it appeared to marketers that most of the money that they were spending was on media and not as much on data. But today, as you all know, most media is run on a self service basis. In a self service basis in a programmatic environment, if you look at industry standard normative CPM’s or how much people spend on just the impressions themselves or the bid requests themselves, versus how much they spend on the data overlaying on those bid requests in terms of segments or premiums they pay because of data embedded in bid requests or the research studies they do on campaigns they run. It’s at least equal in terms of CPM that people are spending on data as they are on media and so we feel that it’s just as important to spend your time and focus on the “how” ads are running as it is on the “where” ads are running. And today that scale is dramatically tilted towards the “where” ads are running, the media side and we need to readjust and get them to the point where they’re even.

If we are spending even amounts on data as we are on media, then we need to spend just as much time and energy and investment and focus on verifying that we get what we pay for on the data side just as much as we are on the media side. So I really appreciate you joining us if you’re interested in reaching out to get more information on this or to have the institute do a course for your organization please reach out to us at Thank you.