The Practice of Awareness

Adam Heimlich
5 min readMar 11, 2019


by Adam Heimlich

The process for buying digital brand media is the same across the major agencies. Personnel routinely switch among them without retraining. Media vendors use the same RFP response template for all the big buyers. The entire branding subset of the media profession is united around a single way to choose placements, comprised of a few simple steps.

This process is not in the interest of advertisers. It goes awry at the very beginning: the core assumption is wrong.

The process calls for the division of client brand budgets into portions, allocated in advance to media properties. Websites, networks and exchange-buying partners are assessed on reach and CPM. Selected partners are each dealt a share of budget to spend on their own audience. In so doing, before the first dollar is spent, the typical campaign planner introduces unnecessary risk and forecloses on the opportunity to achieve optimal return on investment.

The process is built on the assumption that reach improves with the quantity of committed partners. That may be true in traditional media, but not in digital. Online, reach gets worse as you add commitments. Optimal reach requires a consolidated view of audience and centralized control of delivery. Marginal cost and incremental reach are what matter. There is no valid justification for selecting digital partners according to historical statistics and upfront pricing.

The failure to plan and buy with a dynamic view of the entire target audience drives considerable waste. Research shows that between $0.20 and $0.50 of every brand media dollar spent through the traditional process is wasted. (This estimate range excludes waste from fraud and non-viewable impressions.)

Digital media waste is not on most brands’ radar simply because executives are unaware of what ad tech can do for brand media. 99% of the ad tech conversation is about what it does for direct-response marketers, and others attempting to drive trackable short-term effects. In finance terms, ad tech revolutionized media by automatically revising valuations according to data feedback. The driver of value adjustment was the predicted relationship between any given placement opportunity and a subsequent conversion.

If your media goal isn’t tracked a thousand times per day, though, your agency will buy media without a value adjustment. That is, they will treat media as a commodity and buy from people they like, for the lowest price they can get.

To stop wasting money, brand media leaders need to start their own ad tech conversations. They need to start turning down impressions based on exposure data. More important than any algorithm, passing on impression opportunities is what DSPs and DMPs are for. It takes a fraction of a second and costs nothing. This has been true since about 2009. Brands can turn down 100% of off-target impressions, even if the user fell out of target parameters yesterday.

When the human on the receiving end of an impression opportunity fits the preset targeting parameters, brands still need to pass much more often. They should pass whenever any of their publisher partners just served that user an impression, to avoid annoying her. They should also reject users already beyond the threshold for effective exposure, no matter how recently the user passed it. And there’s yet another kind of waste: underexposure. Brand buyers should prioritize users just below the effective exposure threshold, to increase the share of their target audience that sees enough advertising to make a difference.

Material benefits flow from the ability to value-adjust media according to a stream of data on exposures against a target audience. No conversion data is required.

Brands’ typical digital exposure pattern is much different from what it would be with feedback and optimization. The antiquated process of commodity buying across established partners causes a significant share of the audience to see only one ad, while an even bigger share sees too many ads. Only a small slice falls in the optimal range.

It’s worth a moment to think about why this happens. We all know people who consume internet media obsessively. One of the ways digital advertising potentially improves on television is that big-spending brands’ frequency against people who watch tons of TV can only be absurdly high. But at least that couch potato doesn’t reduce your reach! In digital without global control over delivery, heavy users absolutely screw up your reach, because each digital impression reaches one person. Every time a heavy user looks at any site on your roster, you miss another chance to reach lighter users.

The opportunity to reap big rewards from disrupting the old brand media process is hiding in plain sight. Last July, during Proctor & Gamble’s earnings call, CEO David Taylor talked about cutting global user frequency from 20 down to 4, adding: “(We) plan to reinvest savings on wasted spend back into reaching actual customers.” Three months later, the company posted its strongest sales growth in five years.

What happened at P&G is atypical. The vast majority of large advertisers have not directed their buying partners to update antiquated procedures and stop wasting money. The reason has to do with the level of trust between advertisers and the firms that provide brand media services. It is very low.

Among both buyers and sellers of brand media, the preferred way to avoid getting ripped off by a partner is to have a lot of them. It’s one way to hedge your bets. Planners might request rates from 50 partners and deal out budget to the cheapest 10. When contracting for services, advertisers constantly pit agencies against their rivals and grind fees down. Agencies, for their part, are usually part of holding companies that serve their top clients’ competitors. There is little sense, on either side, of two companies in business together, pursuing common goals. Nobody has figured out how to value-adjust without trusting someone to measure value.

As distrust stifles innovation, focus on unit costs makes it hard to see the big picture of brand media value. No one gets fired for awarding budgets to digital media giants offering discounts from their rate card. The opportunity cost of staying within professional boundaries seems preferable to the risk of running even a single global frequency analysis. The latter would imply that one partner can be trusted to hold all others to account.

Without value measurement of exposures, it’s safer to stick with how brand media was purchased since before any of us were born. With value measurement and the cost adjustments it informs, the set of brand media offerings directly compared changes from 50 to more like 50,000. The cadence of decisions switches from once, upfront, to several times per second for the duration of the flight. Instead of requesting rates, you automatically find who has more supply of who you’re looking for on any given day.

If accounting or law or banking were as infected by malpractice as media, and it were as tough for brands to get expert guidance, it would constitute a crisis. Brand media buyers are not trained on ad tech nor bound by standards against bad advice. Try to hold them accountable, and they’ll reference their process. How many of your target prospects see too many ads? How many see not enough?



Adam Heimlich

CEO of Chalice Custom Algorithms, an optimization platform for ad delivery.