Generative Research Documentation

My generative research on the Starbucks register is finally done, and it’s been a lot of fun watching the novice failing on the tasks.

Here’re the four tasks the users are asked to complete:

Task 1: Awake Tea Latte Task 1: Check out an Awake Tea Latte and pay with Student ID

Task 2: Coffee and Credit Task 2: Check out a Blonde Roast Coffee and pay with credit card

Task 3: Starbucks Mint Task 3: Check out a Starbucks After Coffee Mint

Task 4: Gift Card Reload Task 3: Reload $15 on a Gift Card

Here are some things our users have complained during and in between the tasks: - The big yellow “manager” button doesn’t do anything; - The “scannable” button does nothing; - Colors are very unattractive;

I also gathered some feedback from both the expert and the novice after the tasks:

  • Expert
  • The interface is not easy to use; it was very hard to avoid making mistakes at the beginning, and even after she has worked there for a year, she mess things up occasionally.
  • The buttons are disorganized, and there’re random TBD buttons in the system.
  • Some items are not in the system (Mint, for example), so she has to manually type in the amount.
  • The Trenta-sized drinks are in the “brewed coffee” category, which is very counterintuitive.
  • Gift card checkout procedure is annoying: if the customer is using the mobile app, she has to manually punch in the card number to the gift card POS.
  • Novice
  • It’s very hard to find where the items are on the register, and checking out any complex item is hard.
  • Many buttons don’t do anything at some times.
  • The flow is very poor, and feedback is confusing. There’s no prompt on what to do after you perform an action.

And here’s the final video documentation!