Notes from Day 2 of UXLx 2018 — Designing voice interfaces & Crafting discovery phase
The second day of UXLx had two workshop slots, and each slot has five workshops to select from. My notes from this day are from workshops about designing voice interfaces and discovering problems in the first phase of any project.
Designing & prototyping voice interfaces — Ann Thyme-Gobbel
„You need to help your users when they got lost in the conversation. If you don’t, they get angry and don’t use your voice thing.“ — Ann Thyme-Gobbel
Summary
Workshop description promised to cover basics and concepts behind designing voice interfaces, and I think that this was delivered. I really liked examples of conversations that Ann was showing to us and discussion during the workshop was interesting as well. This workshop also made me realize that it can get really complicated even when designing very simple thing to do (say).
What do I want to remember
- It’s better if people are encouraged to talk in shorter sentences. Every word is an opportunity for error when analyzing input.
- You can never be 100% sure that something was recognized correctly.
- It takes 2 to have a conversation. Even conversation with robot should be dialog, not monolog of any side.
- It’s important actually to hear what you designed. Seeing it on the diagram or written is just not enough.
„The problem with the speech is that it’s always floating. Now it’s here, and now it’s gone.” — Ann Thyme-Gobbel
- If you need to confirm something back to user, response should be always specific. The user has only your voice to be sure that everything was set right. For example, when setting time, don’t respond only „okay, it’s done.“ Use longer and more concrete „Your timer was set for 10 minutes“ instead. Reassure your users that you understood them correctly.
- On the other hand, always think if the response is needed. When the user wants to turn lights on, she can see that something with lights happened. It’s probably not needed to add reply „Your lights were turned on.“.
„I don’t like if it sounds too much like the human, it sets the wrong expectations.“ — Ann Thyme-Gobbel
- Be careful with voice intonation of your voice UI. You can’t rely on it much and it can create a lot of problems for users. Intonation can help with adding focus to something but clarity always has a priority.
- When deciding what a correct response should be, think about consequences of the following action. It can be okay to call someone without further confirmation because a user can always cancel the call. But when you send e-mail, it’s not possible to take it back and consequences can be fatal.
- Expect unexpected. Even the most simple answers like „yes“ or „no“ have a lot of possible ways how to say it.
- Unexpected input should be handled gracefully with appropriate certainty. Never blame the user.
- Picking information from a list is tricky. Longer lists create huge cognition load for users. Try to avoid it if possible; users should have control over the information flow.
- A fantastic way how to test early voice prototypes is Wizard of Oz technique. Overview of other methods.
Resources
- [book] Designing Voice User Interfaces by Cathy Pearl
- [book] Wired for Speech by Clifford Nass and Scott Brave
Crafting the discovery phase — Dan Brown
„All the decisions we make on the start will affect everything else later in the process. It will help you keep focus throughout the process.“ — Dan Brown
Summary
A terrific workshop which I really enjoyed. It was a lot about the importance of proper planning and being sure that we keep our focus, and we don’t get lost in the process. To use vocabulary from this workshop — I gathered a lot of information, now I need to process them, explore ways how to incorporate them and then focus on revisions and refining. And I am looking forward to it after this workshop. Inspirational.
This I want to remember
- We can’t fully understand the problem before we start to actually solve it. It’s ok to refine ongoing solution when we find something additional to the problem we are trying to solve.
„In the world where we rush to deliver any solution, we usually shorten our time for thinking about problems.“ — Dan Brown
- Discovery can be divided to two most important parts — framing problems and setting direction.
- Problem statement draws attention to the part of the world that can be improved.
- Project objectives establish basic constraints for our discovery effort.
- Contextual statements elaborate on the ecosystem in which the product will live.
- Principles are guideposts for how to achieve our goals.
- Concepts are the central framework or big idea that ties the product together.
- Models are representations of aspects of the experience to evaluate design decisions.
“Framing a problem gives design team a shared focus.” — Dan Brown
- Plan of discovery phase is important for realizing a scope of the project.
- Well written principles can provoke conversations and can be important guidance for decisions later in the process.
- If working with other team members (which is every time, probably), it’s good to have shared pool of knowledge. It is set of things what we know, so everyone on the team has the same view.
- Problem Statement Worksheet is a great way how to formulate a problem.
- Discovery phase is great opportunity for creating the bridge between the design process and what business expects from us.
- Be sure that you have defined what outputs will be generated from every discovery activity.
Resources
- [book] Communicating design by Dan Brown
- [book] Designing together by Dan Brown
- [book] Practical design discovery by Dan Brown
- [tool] Surviving design projects, card game by Dan Brown