“I may not be able to to accurately describe it. But I’ll know it when I see it.” This was the famous phrase used by United States Supreme Court Justice Potter Stewart to describe his threshold test for obscenity in the famous case Jacobellis v. Ohio (1964).
He may as well been referring to measuring the results of a typical marketer’s online spending.
While marketers are quick to acknowledge the impact and potential of digital and social media, a large majority are not so sure about the effectiveness of their own efforts. For example, in a new survey of 3,000 marketers conducted by Social Media Examiner, only 22% said they think their Facebook advertising is“effective.” Similar numbers have been reported regarding online display advertising.
Yet, the marketing dollars keep flowing.
Why are marketers continuing to up their online budgets at a time when they’re unsure about how that spending affects the bottom line? It could be that on some level, they “sense” it’s working, even though (as in the obscenity case above), they can’t tell you how.
The main problem to measuring results of online media is that many of us tend to view each element of online spending (social media, paid search, SEO, display, mobile, email) as a separate silo. We’d like to see accountability for every dollar we spend.
Unfortunately, that’s not going to happen. The customer’s “path to purchase” has changed from a predictable funnel to an “Influence Loop” with many new possible touch points and opportunities for information and influence, from social conversations, authority and customer review sites and long-tail search terms to the “old school” media of TV, print, outdoor and radio.
So is there any way to measure your program’s effectiveness? A popular approach some progressive marketers are taking is a metric called Effective Cost Per Action (eCPA). Rather than focus on outmoded measurements such as clicks or “likes,” eCPA is built around a “multi-attribution model.” In short, it doesn’t give “credit” for a sale or conversion to any individual click or action, but acknowledges that the overall “mix” of elements may have contributed along the sales path.
Under an eCPA model, you first need to define what counts as an “action” (examples: a sale, an inquiry, a download, a video view, a completed survey, an opt-in, a review, or any combination). Then you would divide your advertising-plus-online spend by the number of defined “actions,” regardless of where your analytics say they came from (yes, even those actions from repeat customers, who at some point are “influenced” to purchase again). This will give you a “cost per action” figure that can be used as a benchmark, either by looking back at past years or quarters, or for reference in coming years.
But how do you know if your Facebook ads, for example, are pulling their weight? You can always take a look at the traditional engagement statistics, but with an understanding that such actions do not necessarily indicate a desire to buy. Trying to attribute incremental sales to a single touchpoint (“Last Click Attribution” for example) totally ignores how today’s mix of mass, social, digital and mobile “mesh together” to create a much fuller experience for the consumer.
Perhaps eCPA is not as comforting as a spreadsheet populated with “click-throughs” but it certainly addresses the reality of the New Customer Path.
If you’d like to learn more about eCPA or are interesting in trying it out, feel free to drop me a line. Either leave a comment or email me at mickey @quisenberry.net
Posted by Mickey