The Data-Driven Art Of Digital Marketing
Digital marketers today have an abundance of data at their disposal, and columnist James Green notes that determining how to best utilize all these data points is truly an art form.
I love tech and data. In fact, there is nothing I like more than utilizing data to make informed decisions. Magnetic (my employer) has run at least a hundred thousand campaigns, and we are accountable for every last one of them. The metrics and analytics we see for each campaign flight provide a compelling view on how to best optimize each campaign for maximum ROI.
What makes the conversation interesting is that sitting right next to the science of marketing is her sister, the art of marketing. Figuring out how to best utilize all these data points is truly an art form. Let’s delve into some of the finer arts of digital marketing.
Multiple Devices Means More Data
As digital marketers, we’ve shifted from targeting a shared family computer to targeting a single person that uses multiple devices. The obvious impact of this evolution is that many companies now have more user profiles than there are consumers. In response, marketers must apply a cross device targeting strategy, giving advertisers the ability to measure for cross device attribution.
So far, so great; however, if you target across devices, but your attribution model can’t give credit for that mobile ad that drove a desktop conversion, then your campaign will appear to be performing poorly. This shortcoming makes it impossible to optimize campaigns properly.
We are huge advocates of fractional attribution, and marketers can work with external companies to achieve this. However, if it’s not possible for whatever reason, your results or return on ad spend (ROAS) will improve by turning off cross-device targeting and keeping your mobile and desktop campaigns siloed until your attribution problem is fixed.
The Most Important Metric: ROAS
Return on ad spend (ROAS) is the most fundamental metric. It is also the one most open to artistic interpretation because you can’t calculate ROAS without making assumptions about attribution. Even if you use the most successful attribution system in the world, you still won’t achieve perfect results. This is because different campaigns perform differently during each stage of a consumer’s purchase journey.
For example, branding campaigns are not going to get the same result as a targeted, lower-funnel sales promotion designed to move goods immediately. Further, we commonly see campaigns use the wrong metric to measure the desired result.
Andy Frawley wrote an interesting article about how we should move away from simply measuring ROI. He proposes a new measurement: Return on Experience and Engagement. Broadly speaking, Frawley addresses how measurement metrics are often too narrow. For example, performance marketers hold their measures too tight, e.g. last click before purchase gets 100% credit, and awareness campaigns look at a set of digital metrics rather than a broader change in consumer intent.
Dynamic Creative Is Key for Personalized Ads
Almost any programmatic campaign today uses massive amounts of data and extraordinarily sophisticated algorithms to decide whether or not to deliver an ad. But, there are few campaigns outside of site retargeting that use data and apply it to ad customization. I suspect that much of this comes from industry inefficiencies: creative agencies are separated from media buying organizations.
That said, you don’t have to be a huge retailer with millions of products to use dynamic creative. Small to mid-sized retailers can also take advantage of data to help customize their creative messages, creating a more personalized marketing experience, which ultimately will lead to better results.
Don’t Rely Too Heavily On Click-Through Rate (CTR) As A Metric
If you ask an algorithm to optimize to a CTR goal, it will optimize to the consumers who click more often. This is just an arbitrary section of the general public, the same as people with brown hair, those that are 5’9’’, or people from Fargo, North Dakota. These people may or may not have anything in common with your best customers.
We’ve all talked about the collapse of the purchase funnel and how a consumer’s path-to-purchase has become fuzzy and difficult to follow. What has yet to happen is the widespread use of dynamic campaigns that run from awareness to purchase. The industry is still under the assumption that campaign strategies should be siloed: my awareness campaign will run here, my in-market campaign will run there, and I will use site retargeting to close the deal.
The future of our industry is combining these separate strategies into one. And technology will infer by the actions you take where you are in your decision making process. Many companies can do this for marketers today, but few actually take advantage of this.
We know that in our digital indulgent world, some consumers will take thirty minutes to make a decision while others will take thirty days. Some people will only shop online, while others prefer the in-store experience. Most people will be seen online and offline using multiple devices.
Data and device information can be used to figure out which ad to show and what the message should be. This can be fully automated from beginning to end, but you’ll always need a partner to make sure that the measurement, analysis, and messages make sense together, performing in harmony. That’s art.