Designing the Oracle Voice User Experience: Oracle Shares the Lessons
By Ultan O'Broin-Oracle on Dec 16, 2013
Oracle’s Mobile User Experience (UX) team has been exploring voice technologies as they evolved. Our interest intensified with the release of Siri intelligent voice assistant on the Apple iPhone in 2011. By converging several technologies, Oracle has designed a mobile voice solution for our Oracle Sales Cloud customers, Oracle Voice. And, more is to come!
Oracle Voice enables users to talk to the Oracle Sales Cloud; speaking naturally to view, edit, and add notes to customer opportunities. Whereas Siri enables users to interact with personal data on their phones such as contacts, settings and calendar, the focus of Oracle Voice is to enable users to interact with their enterprise sales data as part of an overall task flow.
Oracle Voice user interface. A clear UI and underlying technology that recognizes the names of important objects in the task flow are some of Oracle's shared UX design insights.
- More and more sales reps are using voice technologies to get their work done productively. Expect enterprise use cases to increase.
- Voice to text is only part of the technical solution. Natural language processing (or NLP) and understanding users’ context are important related technologies that we had to develop in order to provide a voice solution.
- Understand what enterprise users do, the when and the where, of being mobile. Support only such users most frequent and basic tasks. Voice is not for everything.
- Make voice usage a hands-free operation. And don’t forget any legal requirements, for example when driving.
- Voice recognition must understand user data, such as the names of important objects in their task flows and the relationships between the objects. For example, voice must recognize the input of proper names, such as customer names, that are part of the sales cloud.
- Users will want to use voice-based search to find key information. For instance, users will want to just say the name of a customer in order to see opportunity details returned. Provide for fast search and a way to integrate the results.
- Make the UI clear so that users know what task flow is being completed. Misrecognitions of voice inputs do happen, so provide an ability to correct misrecognitions easily and to continue.
- Keep voice interaction flows short. Remember, a human is talking to a machine that understands enterprise data but hardly anything else, until it learns it. It is not a normal human conversation (yet!) so flows must be as succinct and efficient as possible.
- Although some users may have had only basic experiences with voice recognition in the past, most users that we bring into our usability labs are now surprised at how well the current-state of the-art technology works and helps them to complete simple activities much more quickly (such as when dictating by voice rather than typing a note). With voice recognition accuracy improving steadily, be positioned to respond to more new scenarios of use by having your voice UX roadmap ready.
- Add some personality to the voice interaction. Experiment with sounds for the microphone interaction and the opportunities offered by the many natural-to-machine type voice outputs now available. Personality and emotion adds to the voice user experience. Careful use of humor and an aspect of fun has its place in augmenting productivity on the go.