Affectiva, an MIT-linked SaaS startup, uses technology to observe and analyze physical responses to determine emotional states. Positioned for media testing, Affectiva’s Affdex product uses webcams and facial recognition technology to test advertising spots for large clients. Gone are expensive and inconvenient test labs agencies use to evaluate TV ads—participants just watch from their computers while Affdex records their reactions. The company uses sophisticated algorithms to interpret expressions and head movements, ascertaining emotions and their level of intensity in real-time.
Sounds like Madison Avenue marketing hooey, right? It turns out greater emotional response leads to better business results, and the science backs it up. In a 6-week case study, AOL used Affective technology to test a new online media advertising format (IAB portrait) versus traditional online advertising formats. The results were impressive in how the audience reacted:
· Portrait format attracted attention 35 percent faster than competing formats, created 81 percent more attention, and 95 percent more fixation time
· Live media metrics showed interaction rates rose 4.5x – 7x
· Facial recognition analysis showed Portrait format lowered negative emotions by almost 40 percent, with less user frustration, and fewer frowns
And in business results:
· Users indicated they would visit the Brand site/Facebook Fan page 49 percent more often
· Users indicated they would recommend the brand or product 46 percent more often
· Purchase intent rose 263 percent
What does this have to do with cloud computing? Researchers have proven emotions are a primary driver in customer loyalty and retention, so new emotional response measurement approaches may be critically important. Many cloud computing companies already collect information on what customers think (via satisfaction surveys), what they say (via text analysis on social media sites), and how they behave (via application monitoring tools), so understanding how they feel may not be far away. All of these attributes, especially used in combination, can create powerful customer churn prediction models. When lowering churn even a few percentage points dramatically increases customer lifetime value and per-customer profitability, understanding and addressing the causes of churn depends on good measurements.
A practical application may be using Affdex as a training tool with willing new customers enabling their webcams during the onboarding process. Imagine recording new customer orientation sessions with Customer Success Managers to provide feedback on how to improve their critical first customer interactions (see “What Starts Right, Stays Right”). Managers and staff could then optimize customer engagement and ensure they begin the relationship on a very positive note.