Text messaging as data input was a decision I made early as a result of a survey paired with qualitative research amongst peers which told me that despite its roots in sharing with others, text messaging is very much a private matter. As a result, if someone was to look at your phone when you were texting, you would feel “irritated” or “as if your privacy has been invaded”. It’s a faux pas to peek into someone’s phone without invitation – and so it’s a perfect choice for sensitive data input already embedded in social routine.

 

Up until this point, I had the technology for the text messaging service working, but I hadn’t explored the details of the user experience. My user test using MK2 prototypes  allowed me to get a good handle on exactly how the messaging portion of the tool should look and feel and some first end user input on the details of responses.

 

Things the text messaging service should do:

  • Introduce itself on first time use.
  • Give feedback on how it is interpreting the message.
    • Using graphics over text
  • Acknowledge that it has logged the message.
  • Allow users to adjust the interpreted levels.

 

Things the text messaging service shouldn’t do:

  • Pretend to be a human
    • Using text and trying to appear natural is misleading. It’s not a conversational text message service like Quartz.
    • It should acknowledge that it’s not human.
    • Special mention a Twitter DM conversation I had with Victor Loux (Social Digital 16 alumni)
  • Use numbers as ‘levels’ feedback.
    • Don’t want it to seem too scientific.
      • I had used percentages in previous prototypes.

 

With these requirements in mind, I have designed a text message introduction and response system which allows users to record their thoughts, see the analysis and adjust it if they wish.

 

Onboarding flow:

messageflow

 

Recording and adjusting:

recordingadjusting