About the Talk
March 24, 2018 12:00 PM
Chicago, ILChicago, IL
"assistant who could answer the most frequently asked questions, provide guidance to visitors, and free up staff to perform other duties. The project would also create a unique ohpportunity for developers, designers, and data scientists to converge around a shared product addressing a pressing need.
While AI and automation has made great strides, gaps remain. VAs are uneven in their ability to satisfy user needs; a 2017 Business Insider article unfavorably described using an Amazon Echo as the verbal equivalent of command-line prompts. Adding to the burden, both the Echo and Google Assistant require users to find and install third-party voice apps to expand the device’s capabilities. But many of these are “zombie apps,” according to a VoiceLabs report, with only 31% having more than one consumer review. Such data suggests that although thousands of apps are available and accessible, the overwhelming majority are not heavily used.
To unlock the promise that VAs offer, designers working with these interfaces must develop new capabilities and often, apply an approach to interaction design that goes beyond the visual and embraces integrating new disciplines. …Furthermore, the VA’s capabilities must be aligned with addressing real user needs.
This session will focus on exploring what a user-first approach means for designing VAs. Using a case study of one company’s experiences building a “digital concierge” through a unique, integrated digital experience platform combining the usually separate tasks of search, personalization, natural language processing, and task automation, the session presents key considerations for VA success.
Such factors include:
- Adopt a user-first approach. There is little value in developing a solution for a problem that doesn’t exist. VAs should be built from the outset to solve real problems.
- Integrate information architecture and UX. Ensuring that indexing, automated curation, behavior tracking, and analytics are deployed in relation to content enables the VA to surface the right information at the right time.
- Flex multidisciplinary knowledge. VA projects also offer unique cross-team opportunities for individuals from varied disciplines, from design to data science, to collaborate together.
- Plan for continuous improvement. Consider “before” and “after” states. Measure the user experience and how users are behaving before the VA is deployed, after its initial deployment, and across subsequent deployments."