Sprint's crowdsourcing interface

As the only analytics representative at the Boulder Digital Works (BDW) this August, my experience and takeaway was unique. It reinforced why I think digital media is the best communication medium in which to have a conversation with the customer. Some my favorite creative examples I saw at BDW were Yahoo! Bus Stop Derby & Sprint, both of which leveraged crowdsourcing.

From a client’s perspective, digital allows you to get ideas into the market quickly and efficiently. If the concept doesn’t work, you move on…it’s all about “small ideas.” But HOW do you know whether something worked or not? Measure it! You can measure almost anything when it comes to digital media. So the question is: Why didn’t anyone talk about what impact these initiatives had on the client’s bottom line or objectives? The truth is that even if the objective might be awareness building, clients will almost always look for ROI.

Wouldn’t it be more compelling to tell the audience that not only can you have an innovative execution, but you can also measure the impact it will have on the brand – be it by lift in brand likability or lift in sales? I’ll be the first to admit that some of the ideas and usage of emerging technology can make measurement challenging, as we have no historical context. In such cases, you could test and learn, instead of always testing and winning, or you could start collecting the data for future benchmarks. One of the reasons why I liked the Yahoo! and Sprint creative is because they showed how some new ideas can drive brand likability, even though measuring the ROI might not have been the key goal.

Yahoo! Bus Stop Derby

Agencies are doing a lot of cool work for multiple clients, much of which is meant to drive awareness. We might not be able to leverage traditional digital response metrics for these executions, but a good analyst will work with you to recommend measurement solutions that will allow you to measure the effectiveness of your campaigns using new techniques. These solutions could include using engagement scores, creating a consumer engagement index or using a traditional media mix model with a twist. For example, we have clients who are leveraging QR codes. Instead of just giving the client reports on the traffic from QR codes to their app, the value is in showing the engagement these codes drive for the client. It’s about focusing on micro-conversions to show the significance.

Next time you think up an innovative idea, huddle with your analyst to come up with a measurement solution. We are in the era of creative measurement – it’s a challenge for us as analysts, but we embrace this challenge.

But, sometimes campaigns are not focused on direct response, but rather influencing upper-funnel stages such as awareness and consideration.  When this is the case, a best practice is to conduct pre- vs. post-campaign brand survey studies, but we can also utilize an audience engagement construct based on online behaviors that act as proxies for awareness and/ or consideration and be leading indicators of future economic value (leads, transactions).

There are a number of behaviors that can make up an engagement score – rich-media ad interactions (window expansions, video plays, utilizations such as calculators or games), homepage visitation (especially via post-impression response^†), landing page visitation, store/ branch locater activity, consumption of high-value content (reinforcing campaign messages), etc.

However, not all of these behavioral interactions are equally valuable depending upon campaign goals and the interaction’s respective place in the purchase funnel – visiting store locater results pages is an intuitively stronger indicator of purchase intent likelihood than visiting the campaign landing page or even the brand homepage or ad unit interactions.

To account for this differential value of interactions, each behavior can be weighted, respectively, to reflect each one’s relative value (say, on a scale from 0-2, or whatever is appropriate for the specific brand’s business).

Once a methodology is established, the engagement score can be utilized as the basis for gauging campaign responsiveness (via engagement yield = engagement score/ ad impressions) as well as financial efficiency (via cost per engagement = media cost/ engagement score) to guide campaign optimization decisions.

So, the next time, your client has a digital campaign focused on influencing upper-funnel activities, and you don’t have the time, funds, or interest in formally evaluating brand awareness and consideration measures from a brand survey study, consider behavioral engagement as a good, easily available success proxy.

What experiences have you had in using online behavior measures as gauges of success for upper-funnel digital campaigns…? What else would you consider…?

^†Post-impression (PI) response represents the target audience being exposed to online display advertising (OLA), not clicking on the ad (banner), but at a later time visiting the brand’s site and exhibiting the campaign’s goal behaviors (homepage visit, priority content consumption, order submission, etc.). PI response typically accounts for 60-90% of total campaign response.