The following article was written by Tom Riordan, the head of advanced measurement at Adobe Advertising Cloud, and was published as sponsored content courtesy of Adweek.
Do you have a framework in place to measure and optimize your ad campaigns to the metrics that matter? And is it making you smarter as a marketer over time?
As more and more brand marketers up their media measurement game, it’s critical that they match their metrics to their goals. Unfortunately, many buyers stick to KPIs that can be easily gamed, and don’t convey true advertising value to the media buyer. And that can make it a challenge to optimize your performance and get the most out of your digital ad spend.
The challenge can be overcome if you take a step back and answer these four simple key questions:
1. Am I buying what I think I'm buying?
This question relates to media quality. Issues such as viewability, fraud and brand safety have traditionally been the questions that keep advertisers up at night. While optimizing to these metrics feels like table stakes with current ad cloud technologies, small changes to the math equations we use as in our north star metrics can have big payoffs.
Let’s start with viewability. Determining if an ad is viewable has been a metric the industry typically has been obsessed with. And while 100 percent viewability is feels like a reasonable goal, it’s not always the best measure of actual viewership. Volume metrics such as the actual number of people who could view the ad is are more meaningful. Therefore, you may choose to look at viewable CPM, which takes into account viewability and as well as the ultimate cost and total count of those views.
Next is brand safety: Do my ads actually display in an environment that I’m comfortable with? Did I buy the apps, websites and TV programs that I intended to buy? Then there’s fraud: Did I actually serve ads to human beings?
Most buying platforms now account for these kinds of issues through features related to transparency. And while the tools don’t catch every infraction, you can now react in real time directly on these platforms.
2. Am I spending the right amount on a given tactic or channel?
Next comes the challenge of balancing reach and frequency on each advertising channel. Have I reached a point of diminishing return on a given tactic or channel? Is there room to scale? Will adding dollars to a particular tactic provide incremental value to the brand or will I just be reaching consumers I’ve reached elsewhere?
This is another area where accuracy metrics can be problematic. Ad campaigns optimized for high on-target-percentage or conversion rates are inherently limited in the scale of impact they can have for a particular brand. Brands with a higher tolerance for “missing” on an impression-by-impression basis will actually end up with more unique “hits” on customers they care about. Consider updating metrics to reward unique reach, by looking at things like Cost per Human Reached or Cost per Unique Lift Point. (We suggest shorthand like “Quality CPM” to make talking about them a little easier).
3. How is my ad working?
This question attempts to get in the mind of the consumer: How is my ad working? What are the changes in perception that are being created that may lead to behavioral changes?
Clarity of your measurement is critical here. It’s not unusual for media buyers to focus on goals that are in conflict with one another—I want to drive reach and frequency; I want to drive trial and sales. It is critical to know what the campaign is trying to achieve and aligning those metrics against those specific goals.
This sounds simple but is often overlooked. Efficient Conversions are the choice objective for performance advertisers nearly across the board; but brands looking to drive trial by new customers should really emphasize increasing the total amount of consumers who buy (Conversion Rate). Whereas for brands harvesting an existing known customer base, an efficient volume of conversions (Cost-Per-Action) is more appropriate.
You can take this a step further by enriching that “Action” at the end of Cost-Per-Action. Brands focusing on harvesting known revenue should be counting transaction totals from each action. Brands driving trial and exploration may want to look at pre-purchase metrics like dwell time and total site actions, which can signal product exploration and intent. And if you’re like the majority of advertising brands in the world, most of your key consumer actions happen offline. Using offline attribution feeds—such as point-of-sale data, loyalty card data or a home scan panel— has become widespread for solving this problem, and many brands now operate with an “always on” framework for incorporating offline behavior.
But this richer understand of behavior still falls short of incorporating sentiment, your “glue” metric. Understanding how sentiment can relate to behavior – rather than treating it as an end goal in itself – can lead to a much clearer understanding of how your ads are working in the wild and bring your media and creative efforts together. For instance, drawing a correlation between the spikes in brand metrics as well as buying behaviors that result from seeing a particular ad can provide insight to that brand to incorporate into future creative processes.
4. Are incremental actions being taken?
Lastly, the hard part. Brands need to ultimately know if their ad campaign is causing behavior that would not have happened had they not advertised. This is tricky given the concept of “attribution” suggests that consumers who purchase after seeing an ad were influenced by that ad – a notion that unfortunately is not always true.
Take the now-aging example of the shoe ad that follows consumers around throughout a purchase process they had already intended fully to complete, regardless of the additional ads. These ads are there to annoy you for one very specific reason: attribution models will reward them for dropping a cookie on you prior to making a purchase.
What if we could get a window into which consumer actions were going to occur anyways, and then remove those from the equation when assigning credit? This is the world of incrementality testing that is quickly emerging, powered by a new set of toolsets and a growing understanding of the limitations of performance metrics used to date. This part starts to feel like science class, with proper control/exposed groups, “sugar pills” or placebos, and the need for a new class of media analyst to interpret and respond to results. If this feels new and challenging to you as a brand, you’re likely not alone. Emphasizing the importance of incrementality testing tools with your technology partners and beginning to build your team to incorporate experts on advertising experiments are a must for brands looking to maximize their ad impact.
Final thoughts: If you’re going to succeed with advanced measurement strategies, you need to have the right framework in place. By asking yourself these questions—and adjusting your KPIs to enable that provide the opportunity for real-time campaign optimization on the metrics that matter —you’ll be creating the foundation you need to get the most out of your ad spend.