After observing the increase, the importance of voice user interfaces and personal assistants like Siri, Google Assistant and Alexa, I decided research something about this theme to my undergraduate thesis. Of theses assistants, I choose Alexa - specifically the Amazon Echo Dot device, because I saw the chance to propose something new and increase your competitive at industry.
To understand how the assistant work, I bought an Amazon Echo Dot and began using it everyday whit the objective of discovering its real concept, and why people wold buy it. During the period I realized that the assistant always needed a command to do something. After understand its behaviour, I thought that I cold propose something easier. 'Hey, I can proposal something better!'.
With my research, I found some academic papers about Anticipatory Design, that work to by understanding user behaviour, collecting data from their devices and anticipating the users' decisions.
So, after collecting the content and understanding better how Anticipatory Design works, the next step was create tasks for user tests. To keep the consistency between the tests, I used some suggestions from the News and Information section at the Ask Alexa folder included in its box. The selected task was:
Task 1 - 'Alexa,tell me the news.'
Task 2 - 'Alexa,will it rain today?'
Task 3 - 'Alexa,what's on my calendar?'
After this, I created artefacts sounds for Proposal A ( How Amazon Echo Dot work Today) and Proposal B (Anticipatory Design proposal)
The nexts steps was to create a test script, recruit users and finally, execute the tests.
Five users were recruited for the tests, and when each had finished, I used questions from System Usability Scale (SUS) to collect metrics, and validate which version was better according to the users.
Bellow, show the graph comparing the results of Proposal A vs Proposal B
- All result of Proposal B was above average with 68 points.
- The recruited users had the positive interest for Proposal B because your humanized interaction.
- Them also gave positive feedback that Anticipatory Design can be a differential for the future of conversational assistants.
Read now the full academic paper aproved in Interaction Latin America 2018