Throughout his career, Young-Bean Song has been at the forefront of analytic research and development in digital marketing. He currently oversees The Atlas Institute, renowned for its pioneering research in all things related to digital marketing. Young’s work also includes custom research for Atlas’ top agencies and advertisers.
PERFORM: There’s a proliferation of advertising channels on the Internet: search, display, rich media and social networks, to name just a few. How are today’s advertisers measuring their online campaign performance across channels?
YOUNG: When you talk about measuring the impact of digital marketing, there’s no shortage of data out there. The reality, however, is that people’s understanding of campaign performance is very limited. The industry standard that’s emerged over the last dozen years gives 100 percent credit to the last ad seen or the last click. Because of this, people’s view of what’s working and what’s not is framed in a very short period of time and only takes into account a single touch point. As this one touch point equals just one ad in one channel, clearly we have yet to measure the true impact of all digital marketing efforts. We’ve developed our Engagement Mapping model specifically to address this.
PERFORM: Audience or brand “engagement” is of paramount importance to big brands and their agencies. What specific types of engagement can be analyzed in the online world, and why are current models inadequate?
YOUNG: Until recently, online has been viewed as primarily a direct-response medium or channel. Brand advertisers are just now beginning to take online seriously and recognize the mainstream impact of the medium. The challenge is that, to date, they’ve had very little metrics that mattered to them, and the talk of measuring engagement has largely been theoretical. They’re looking for measures that have more to do with when, where and how consumers are engaging their brands than simply getting them to click on ads and purchase something from their websites.
PERFORM: So how is engagement measured under this new model?
YOUNG: We need to start with the tenet that not all ad touch points are created equal. Some ads are more engaging simply because they’re physically larger than others. Some are more engaging because their placements are more useful and relevant. There are myriad fundamental variables that can be measured in a standardized way to differentiate one impression or one interaction or one click from another.
For example, recency: We all know that ads that are closer in proximity to the sale have more impact and relevancy. The last ad standard can be viewed as an “extreme recency model” that gives all the credit to the last ad served. But what if the user were to see a dozen ads the day they purchased? Shouldn’t each one of those ad impressions get some credit? That data is available and useful to advertisers, but it’s being completely ignored and overwritten by the last ad model. Another example is reach and frequency. Brand surveys have shown that there’s a difference between reaching users once and reaching them multiple times. There have been similar findings about the impact of different ad formats. Banner ads, rich media, Web video and text all have different branding impacts, and even behavioral measures like click-through rates differ dramatically. Each of these elements can be isolated and measured and added to the engagement map.
PERFORM: How does a planner or analyst determine which of those factors are most important?
YOUNG: First, you have to understand whether the campaign is focused on branding, direct response or both. When you have direct-response goals, you’re going to care more about things like recency, and you’re going to care less about things like ad size and whether ads are rich or not. Another important consideration for direct response is the purchase cycle and whether what you’re selling involves a long purchase consideration or whether it’s a short, instinctual purchase. Ringtones are not purchases that people lose sleep over, whereas things like buying a new car or getting a new home loan require more time and consideration. The higher consideration products need longer frequency windows, whereas the low-consideration products depend more on recency. All this requires strategic discussions about your business as well as quantitative ones. There are a slew of analyses that can be conducted on your historical campaigns that can help steer many of the assumptions that drive reporting.
PERFORM: Engagement mapping seems to tie digital marketing more closely to the marketing fundamentals applied to traditional marketing. Is that a fair assessment? If so, what does that mean for digital marketers? What does it mean for more traditional marketers?
YOUNG: Given the way online success is currently measured, we shouldn’t be surprised that traditional advertisers see the online space as a purely direct-response world. But when you start talking to traditional advertisers and chief marketing officers (CMOs) about recency and frequency, and tying all these different touch points together, it changes the conversation completely. Online marketing becomes something that’s in line with how they view the world.
So these new standards that are emerging are not only going to be great for the people who have been investing in digital marketing for the last decade, they’re also going to help make this medium make sense to traditional advertisers. We’ll see more dollars come online as these new metrics become the new standard.
PERFORM: Will advertisers see a significant difference in advertising results when engagement mapping is applied versus the current “last ad” model?
YOUNG: The answer to that question depends on how much overlap and cross-channel synergy is occurring on a particular campaign. If you’re a large advertiser reaching a lot of people across multiple sites and multiple channels with high levels of frequency, you’re probably making some poor media decisions because of current reporting practices. For smaller advertisers, the change will be less dramatic but equally important from an insight perspective.
When people are being reached across multiple sites, that’s when that credit becomes fractionalized and shared across multiple sites and placements. Engagement mapping is not done at the site level, it’s much more granular than that. By granular, I mean at the placement and creative levels. As a result, you see conversion credit shifting from one place to another quite a bit more at those more granular levels. The more granular you get, the more credit you see shifting.
For example, in the diagram we’ve provided of an engagement map (Figure 1), Publisher A gets credit for about 5,500 conversions with the last ad model (on the left). With engagement mapping (on the right), the conversions go up by 47 percent when factors like reach, frequency, ad size and creative type are weighted and considered in the conversion process.
PERFORM: How big a leap is it for advertisers’ agencies to embrace engagement mapping?
YOUNG: From a workflow standpoint, there’s no additional work or cost. In fact, many advertisers today are already trying to use reach and frequency reports and intuitively give more credit to the rich media campaigns, even though they don’t appear to be performing on an ROI basis. Engagement mapping has standardized these efforts, creating a transparent, scalable system to accurately attribute conversions.