• by Fred Askham , January 7, 2019

We have all known for a long time that while digital media budgets are increasing year-over-year, the scale of the investment does not match the scale of Return On Advertising Spend (ROAS) numbers.

So why don’t ROAS and sales line up? Because the ROAS numbers reported are wrong. Businesses know it. CFOs know it. And marketers know it. It’s the comfortable lie we’ve all been living with for far too long — and it’s time to make a change in 2019.

The belief that every sale that can be tracked to digital media is wholly and fully due to the consumer seeing media is an inflated and often false perception. Forget about ad viewability. Forget about consumer purchasing habits. Forget the millions of reasons this is a horribly inaccurate assumption.

If marketing touches someone before they purchase a product, marketing gets credit, period. It is this comfortable and convenient lie that creates wildly high ROAS metrics and a general lack of trust in the value of digital marketing. And that is why the lie is comfortable. It’s simply easier to believe it than to challenge it.

Since the 1990s, digital marketing has advanced and flourished on its ability to track direct sales, and for good reason. It’s enticing to everyone involved to be able to say: “Our media drove this-many-sales and therefore, our campaigns are ROI-positive and should receive more budget.”

More and more advanced systems, like cross-device tracking platforms and cookie-less tracking systems, have shown the full effectiveness of digital media dollars. All these systems are built to attribute as many sales to digital media as possible, justifying their existence by being predicated on the fact that better tracking of end behavior allows for more informed decisions on how to spend media dollars.

There is an implicit understanding that implementing these measurement systems will increase the measured ROAS associated with digital media campaigns, thus supporting bigger media budgets.

This has led agencies and brands to regularly tout metrics like $4 ROAS on digital marketing spend with 30-day look-back windows. To translate that for those less-marketing inclined, a $4 ROAS is akin to claiming a 300% ROI on a 30-day investment period. (A $1 ROAS is equivalent to breaking even.)

Red flags should be going up in your mind right now — major red flags.

If marketing actually produced a 300% ROI in a 30-day investment window every business would be dumping as much money as possible into digital programs. Even at an average of a 5% return on a 30-day investment period (a far cry in scale from the ROAS metrics we are accustomed to seeing), marketing would be an insanely profitable endeavor

yearly compounded return even at that “modest” rate would be 79%, which beats almost any investment that can be made in traditional equities markets.

But times need to change with 2019, and we need to get past the comfortable lie we’ve blindly accepted for far too long – to our own detriment. Whether you’re a CEO who wants to align your marketing department’s incentives with your own, a CFO who wants to better understand the financial returns from marketing, or just a marketing professional who wants to make a real impact on your business.

There are a few new considerations that might make a huge difference in getting closer to the truth, among them, causal attribution, a bright spot in an otherwise dim set of misleading, antiquated measurement tools and systems.

Causal attribution enables the ability to report the incrementality of the sales tracked by digital media on a 1:1 basis, thus providing highly granular insights into how to improve the efficacy of digital media campaigns in near real-time, without the need to spend large portions of budget on non-working PSA ads. 

In addition, causal attribution systems utilize metrics like viewability and third-party data segments to construct highly matched test and control groups naturally from media delivery logs. Via Rubin Causal Model (RCM), causal attribution accurately controls for natural consumer behaviors (i.e. new offers entering the market, competitive product price changes, and seasonality).

It finds and matches extremely similar consumers whose main difference is that one set saw advertising and the other set didn’t, so the difference in behavior is attributable to that media exposure. Think of it like a giant, super deep A/B testing system. The metrics returned are truly representative of the effectiveness of converting consumers via paid media. 

The hard part, ironically, is that the information produced by causal attribution is significantly more accurate, resulting in less attractive, yet more honest metrics, typically somewhere between 70%-90%.

But sometimes, getting past an inbred industry lie and a lot closer to the truth comes with a tough dose of reality. As an industry and as marketers, it’s time to embrace that reality in 2019 and get on the right side of digital marketing history.