Emotions Run High with Virtual Assistants

It’s about a 4 min. read.

Authors
Chris Neal
VP, Technology & Telecom

The pace of innovation and disruption is accelerating. Just 10 years ago Uber and Airbnb didn’t exist and the iPhone was still a novelty shown off at parties by overenthusiastic tech lovers. Now, we have a hair salon receptionist convinced she’s speaking to a real person when in fact it was Google Assistant that was scheduling an appointment. While it might be hard for many of us to remember the last time we took a cab or used a flip phone, change is hardly straightforward and tech adoption raises critically important questions for brands.

Why do some people resist change while others embrace it? What emotions trigger true acceptance of a new technology and a new way of doing things?  What is that “a-ha” moment that gets someone hooked on a new habit that will be enduring?

To help understand consumers’ journey with evolving technology, we applied our BrandFx framework to the broad virtual assistant category—measuring the functional, social identity, and emotional benefits that people seek from Siri, Alexa, Google Assistant, Cortana, etc. I shared our findings from the identity aspect here.

And while each of these three benefit types play a role in adoption and use—the role of emotion is profound.

We asked a lot of people about how they use virtual assistants—from information seeking to listening to music to planning and booking a trip. Then we ran analytics on the overall emotional activation, valence, and specific emotions that were activated during these different use cases.

Our findings have broad implications for anyone in the virtual assistant category creating marketing campaigns to drive adoption, or product UX teams looking to design customer experiences that will deepen engagement.

Currently, virtual assistants are primarily used as information-seeking tools, basically like hands-free web queries. (See Exhibit 1):

Even though virtual assistants are evolving to do some pretty amazing things as voice-based developer communities mature, most people are only scratching the surface with the basic Q&A function. Asking Siri or Alexa for the weather forecast is a fine experience when they’re cooperating, but it can be extremely frustrating when you don’t get the right answer—like getting the current temperature in Cupertino when you live in Boston.

Meanwhile, watching TV or shopping through your virtual assistant turns out to be a much more emotionally rewarding experience, based on the analytics we ran. The problem for the industry as a whole is that these more emotionally rewarding use cases are among the least used VA functionalities today. Teams that market these experiences must motivate more consumers to try the more emotionally rewarding VA use cases that will deepen engagement and help form a lasting habit (see Exhibit 2):

Listening to music and watching TV/movies yields high emotional activation in general—specifically “delight.” Our driver modeling shows that feeling “delighted” is one of the top predictors of future usage intent for a virtual assistant product (see Exhibit 3):

As Exhibit 2 above indicates, using virtual assistants for scheduling and calendaring has overall moderate emotional activation, but is particularly good at activating feelings of efficiency and productivitythe single strongest predictor of use in this category.

Tellingly, however, the scheduling and calendaring function also over-indexes on feelings of frustration because this task can be more complex—currently AI and natural-language processing (NLP) technologies are more apt to get these kinds of requests wrong.

In general, “frustration” indexes high on more complex use cases (e.g., arranging travel, coordinating schedules, information seeking). This is a warning to the tech industry not to get too caught up in the hype cycle of releasing half-baked code quickly to drum up excitement among consumers. It also helps explain why younger demographics in our analysis actually experienced more frustration with VAs than older cohorts (contrary to my initial hypotheses).

Younger consumers are attempting to do more complex tasks with virtual assistants, and therefore bumping up against the current limits of NLP and AI more frequently. This is dangerous, because they are the key “early adopter” segments that must embrace the expanding capabilities of virtual assistants in order for the category to become pervasive among mainstream consumers.

Consumers will quickly abandon a new way of doing things if they get frustrated. Understanding and activating the right positive emotions and minimizing the negative ones will be critical as brands continue to vie for the top virtual assistant spot.

Interested in learning more about the emotional dimensions of Virtual Assistant users? Reach out to Chris Neal, CMB’s VP of Technology & Telecom.