Connected Devices Weekly Update/2016-07-26

• Sandip went over Vaani on the Monday morning call!
• Prototype is in Berlin!
• We are currently finding users
  • https://www.surveygizmo.com/s3/2942818/Vaani-local-short
  • People between 25 and 40 years old, living in North America urban areas, who work full time and run a 3 to 5 person household, are involved in grocery shopping (primary or shared responsibility), and use smartphone to create shopping lists
• Bi Weekly review is coming up
  • Please join us to hear the results of our survey about Vaani Local to users
• Currently looking into the best way to set up geofencing
  • We would like to collect emails as the call to action (as in lean startup training) but are not able to based on legal
    • Is there going to be an official format for all of the teams?
    • We would love to hear other team's solutions

An always-on voice-enabled smart screen device to keep track of the activities, events, and to-do lists of family members. Initial focus for the prototype is on Reminders only.

Hypothesis Tested by a Study

The Cue team tested the following hypothesis in a follow-up study over the past few days that included busy family representatives from the first study:

• By creating a voice enabled solution to create reminders, lists, coordinate schedules, we will reduce the time taken to perform these tasks by 20%.
• By offering a solution that provides an alternative way to create and manage reminders, lists & schedules using a smart display, families would be able to accomplish their tasks 20% more times than before.

Study Interviews

• Study participants were asked a series of questions focused on three primary areas: reminders, lists, and event and activity coordination of family members and those outside the family. These interviews were recorded and notes taken for later synthesis of data collected. The interviews began at the end of last week and took only a few days to complete.
• Interview questions were designed to collect deeper insight into what types of reminders, lists, schedules people are currently using and managing in their lives.
• The study also included exercises where:
  • participants were asked to set two reminders using the method they normally use. They were then asked to use a prototype to set the same reminders using voice. Each of these exercises was timed so that a comparison could be made post-study about the potential time savings that might be realized.
  • participants were asked to set some example reminders of their choosing verbally to gain an understanding of the speech patterns they use to set reminders. This exercise was repeated for to-do lists and to schedule events both within the family & outside the families.
• Participants were also asked to enter a few prescribed reminders verbally into the prototype and then were asked for feedback on the use of voice to create the reminders.
• At the end of the interview participants were asked questions to gain design insights into what would be important to them when setting reminders, creating lists, and coordinating schedules, both from a voice perspective and also from a visual perspective.
• The next step for the team is to synthesize the data and determine if the hypothesis was proved to be true or not. Input will be evaluated for what is important in the design of Cue. This evaluation will also result in the formulation of additional hypotheses to be tested.

• Bulleted list item
• Bulleted list item

We have been working on 2 ideas for a wearable device that lights up/buzzes when people I care about are available.

• Idea 1 is a device that lets me set/unset my status for my friends to see.
• Idea 2 is a device that lets me request the current status of my friends - my friends can then send back an auto-reply that means "Busy now, txt you later."

Priority assumptions to test - see here for all.

• We believe people have 5-6 priority people/things that they care about at a time (both ideas).
• We believe that the number of desired auto-replies is small (idea 2).

Our next experiment is a survey to learn about the following. Launching this week.

• For both ideas, how many people are considered "priority people"? This has design implications for the hardware and software.
• For Idea 2, what are the types of auto-replies people would need to make this device valuable to them? This also has design implications.

Also working on visualizations of the device interaction. We plan to use these in future experiments prior to building the actual device.

• Example for Idea 1
• Screenshot from interactive HTMl client

• Ongoing experiment (https://trello.com/c/fdvHl50c).
  • Hypothesis: Average users want to have access to hyperlocal environmental data
  • Deliverables: conversion rate and a number of early adopters
  • Metrics: 1) Sessions 2) Subscription rate 3) Referral
• Next steps
  • Follow up measurement, analytics on the current experiment (target done July29)

• Bulleted list item
• Bulleted list item

• Experiments in Progress:
  • Prototype Device
    • Testing the assumption: We assume that we can identify some percentage of food via image recognition
    • Targeted to be given out to three users this works to conduct experiments with Image recognition.
    • Trial agreements finalized.
    • Initial experiment will test what 1) what percentage of food items we can recognize and 2) how often users are interested in getting recipes.
  • Facebook Ad
    • Testing the assumption: We assume that people want recipe ideas that they can make with their existing inventory
    • Currently blocked on collecting email addresses to test the hypotheses via CTA.
    • Unless we hear something back today approving it, we will go with surveygizmo without email addresses.
  • Video
    • Testing the assumption: We assume that people want recipe ideas that they can make with their existing inventory
    • Team is working on a script to test user interest for the idea. Filming scheduled for Thursday.
• Update with the PIB next week.

View Source

• After talking to Viewsource team and knowing the painpints in the last view source, we have proposed them some use cases/experiments focused on the "discovery" to be run in the next View Source event in mid September. Magnet proposal for View Source and requirements
  • Schedule discovery has been proposed as the primary experiment: Attendees would be able to open up the application and see what is happening around them. Since the View Source is mainly a one track conference, the big benefit would be seen during the Discussion Corners as attendees wander through the event space.
  • Secondary experiment: broadcasting information about the presenters. Ideally it would be related to the event in some way, such as their internet-accessible presentation or some sort of biography page. This way attendees would not only see who is presenting now, but would get a quick way to learn more about the presenter or see their presentation.
  • Regarding the availability of the apps on the stores at the beginning of September, we are still nailing down with Marketing and legal teams the details. We need to make a decision about the branding of the app and the account used otherwise we will not be in time for View source (planing and milestones spreadsheet)

Development

• Basic Notifications landed in iOS and Android (some videos demos are included in the links)
• Already landed in iOS the improvement about keeping history of beacons already discovered, so users are not notified about them. This week we will continue integrating this feature in Android and doing bug fixing
• Most of the team are attending to the lean startup training this week so we hope we can get some insights and run new experiments to validate our assumptions.

• Working with Product strategy to come up with a succinct view of what moving forward looks like
• The team has created a mana to pull together information about competitors
• The work will start this week and be ongoing through out the quarter

• Bulleted list item
• Bulleted list item