The tracking trap in digital marketing

Columnist David Rodnitzky discusses the power and peril of measurement and explains why both online marketers and brand marketers need to find the right balance.

Chat with MarTechBot


In Google’s early days — when online marketing was still a novelty avoided by most large advertisers — then-CEO Eric Schmidt acerbically quipped that “the last bastion of unaccountable spending” was the corporate marketing department. Schmidt’s comment was no doubt intended to contrast the measurability of online marketing against the opaqueness of traditional media.

In the first decade of the new millennium, Google was incredibly effective at spreading this narrative. As a result, media dollars flowed from traditional to digital. Today, digital media accounts for more than 36 percent of all media spending.

Many online marketers, however, are so obsessed with measurement that they absolutely refuse to invest a dollar that can’t be clearly measured. This dogmatic adherence to tracking will push marketers into a corner, much as a dogmatic dependence on branding is also a career-limiting path.

Embracing fuzziness for fun and profit

The historical problem with brand marketing was the notion that half of marketing works, but marketers didn’t know which half. Online marketing has its own historical problem — 100 percent of the marketing might work, but growth eventually stalls, because there are only so many measurable clicks available.

Ideally, we want the growth potential of brand marketing, along with the predictability and measurability of online marketing. Realistically, I think the proper path forward is to improve the measurability of a brand but to hold it to a lower standard of measurability than online marketing.

Improving measurability is already happening. In-store beacons help measure the impact of advertising on retail; probabilistic and deterministic attribution can show how offline advertising impacts online purchases, and vice-versa. TV, direct mail and radio are slowly becoming programmatic and addressable.

Consider a radio campaign. In the campaign, you can promote an offer code that can be redeemed at a website for a discount (deterministic tracking). You can also use attribution to correlate the time and geography of radio spots with spikes in online traffic or direct URL type-ins (probabilistic tracking). Lastly, you can limit the radio spots to a few states to evaluate incrementality.

In a perfect world, the campaign would generate enough code redemptions to be deterministically ROI-positive, though even with a discount code, many users will convert without the code. The probabilistic attribution might show some lift, but conversions from radio are often time-shifted and can even be geo-shifted because listeners are driving.

So that leaves us with geographic incrementality — a mushy metric. This type of data is not going to provide a lot of insight into which messaging performed best, or which radio stations drove the best conversions. Instead, we are left with a question that will make online marketers uncomfortable: Did my spend generally work or not?

If you are an online marketer, you are used to knowing which query, at which time, on which device, to which ZIP code, on which day of the week, with which ad text, which offer and which landing page performed best.

But if we can directionally conclude that this radio campaign worked at a high level, that should be enough for us to declare victory.

The bottom line is that we found a channel that drives results. If online marketing has reached a point of diminishing marginal returns, the options at this point are to stop growing or invest in new channels that drive less measurable but incremental growth.

Change or die

I used to predict the demise of brand marketing. My vision was a future where everything was completely measurable. Online marketers like me would treat all channels as “zeros and ones,” and channels that couldn’t offer our required granularity would fade away.

Mea culpa. Radio, TV, display advertising, billboards — these continue to be effective at introducing new customers to products. The future will not be devoid of top-of-funnel marketing spend. It will, however, be more measurable. Online marketers will need to play in this space to survive.

And what of the bastions of “unaccountable” brand marketers? These marketers will need to embrace the drive for accountability championed by online performance marketers. Simply driving “lift” without measuring, say, “revenue per lifted consumer,” will no longer suffice.

We are seeing a convergence between online marketing and brand marketing that will make both types of marketers uneasy. The marketers who thrive in the future will be able to sift through reams of data to find minute optimization opportunities, while simultaneously championing channels that drive directional, but largely unmeasurable, uplift in performance. Depend too much on tracking, or avoid it entirely at your peril!

Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.

About the author

David Rodnitzky
David Rodnitzky is CEO and co-founder of 3Q Digital, a marketing firm with offices in the San Francisco Bay Area and downtown Chicago. David is the founder of the LinkedIn Online Lead Generation Group, an advisor for Marin Software, and a regular contributor to the 3Q Digital blog. He can be found at numerous speaking engagements across the SEM community.

Get the must-read newsletter for marketers.