Skilled Marketers Measure Twice Before They Act

Digital campaigns should be designed and executed with the same care and precision that a master carpenter designs a table or a bookcase.

Skilled carpenters will famously measure every millimeter of the wood they work with. They will do so multiple times before, during, and after executing their first cut.

Digital campaigns should be designed and executed with the same care and precision that a master carpenter designs a table or a bookcase. In fact, we are entering a time in digital advertising where it is going to be important that we “measure twice” before we act—or maybe even three or four times.

Audience-based targeting has established itself as a standard in digital display advertising, capturing over two-thirds of display ad buys in 2016. According to eMarketer, programmatic usage could grow to nearly three-quarters of digital ad spending in 2017.

While ad tech has been busy touting person-based one-to-one reach and sales funnels, our advertisers have been asking us for the same proof points required of direct mail, email, and store purchases. Is their digital investment paying off? Can you prove it?

Despite a spectacular average annual growth rate of 46% in the past two years, there is a growing buzz from CMO offices and agencies that more is expected.

A quick look into the “way-back machine” recounts the journey that audience targeting has been on, and why there is still much left to do until we are all master craftsmen of the digital space.

Where We’ve Been: Blips On The Radar

Audience targeting offered us the promise of demystifying the fuzzy audience segmentation schemes that ad networks were pushing. Designed as pools of anonymous users associated with anonymous traffic patterns, we wandered through campaigns trying to make sense of patterns that existed only as margins of less than a tenth of 1%.

This brought us machine learning and DSP optimization with a promise to validate these patterns, but all too often those patterns were obscured by fraud or “saved” by anonymous clicker pools. Onboarding hashed emails and logins brought a glimpse of clarity to audiences—and the shared portions of the walled gardens of Google and Facebook. But these audiences lacked scale and were rampant with unique user duplication.

As mobile impression counts began to accelerate, the necessity of cross-device mapping became a reality. What started as probabilistic device graphs graduated to deterministic mobile identities, and while still lacking in scale, mobile renewed our hope in a person-based ID.

As we inched our way toward improving audience transparency, campaign attribution flashed and faded in our consciousness as the act of modeling a depreciating cookie pool against incomplete IDs gave us nothing but thin data and a slightly better guess.

Today’s flavor of the year is account-based marketing. This shiny new object promises once again to exonerate the old marketing funnel, but mostly gives us fun new labels for a two-decade-old direct mail methodology.

The Realities Of A Common Identity

Despite the tongue-in-cheek criticism, the trajectory of history tells us that the economics of programmatic advertising works. The promise of audience targeting appeals to most advertisers because the cost basis is so appealing and it does a far better job of mimicking the direct marketing 101 standards than earlier digital strategies. However, this is the reality our industry has to face: Behind every digital campaign is an advertiser who is also investing heavily in other marketing channels.

The link between digital audiences and audiences in other marketing channels is supposed to be a common identity. Soon, being able to measure with deterministic reach and precision the impact digital has had on every other channel will be required. From ad impressions and website visits to email opens, visits to the mall, and purchases made in the store, digital spending must be just as quantifiable as every other marketing spending.

What Could This New Proof Look Like?

For most campaigns, multiple touch points are already in place that can be used to define the optimal campaign audience. Instead of selecting a bucket of invalidated “intenders” or just onboarding your CRM list and getting a 35% match rate, start by pulling the past few months of impression data from your website. Do the same for the last few online campaigns you’ve run. Retrieve the emails from your marketing mailing list, and pull your CRM postal list of customers who made a purchase both online or in the store.

We are reaching the point where you can map an audience across mobile devices and locations and determine up front which categories of media and websites customers prefer. Whether you choose to focus your online outreach to the locations that matter or supplement your campaign with direct mail and email, your prequalified audience should be in place at scale.

Like the skilled carpenter, the successful marketer in 2017 wants to be precise and avoid waste. Every audience should be measured at least twice before the first impression is served.