Mike with the donation confirmation, after I had used his Echo device to successfully donate $10 to charity.
Me: “Alexa, donate $10 to St. Jude”
Alexa: “Do you want to make your $10 donation now?”
Me: “Yes”
Alexa: “$10 will be sent to St. Jude Children’s Research Hospital using Amazon Pay.”
Thursday morning, $10 was successfully transferred from my coworker’s bank account to St. Jude Children’s Research Hospital. Don’t worry — I paid him back… Mike and I work together at Vocool, a voice-tech company helping businesses engage with consumers on Alexa. As you might imagine, our office is littered with almost every Amazon Alexa and Google Home device. I had just learned about the new feature that Alexa rolled out that lets users donate between $5 and $5000 to charity simply by using their voice. I tested the interaction on Mike’s Echo (which is apparently tied to his personal Amazon Pay account). Surprisingly, it worked! If you are curious to see it in action, watch the video below for the full sequence (WARNING: mute nearby Alexas!).
But why was I never required to verify that I was the owner of the device?
Shortly after, I actually started again using the $5k amount, and nerve-wrackingly responded with a “No” when Alexa asked me to confirm the donation(silently pleading that she wouldn’t mistake my response for a “yes”).
Me: “Alexa, donate $5,000 to St. Jude”
Alexa: “Do you want to make your $5,000 donation now?”
Me: “No!” 😬
Voice platforms provide developers with opportunities to create refreshingly simple, yet powerful applications. Donating to charity using voice was exciting and the ease-of-use actually might make me do so more often. But how can we build conversational experiences that balance brevity with security in a development environment lacking a fingerprint reader?
Finding the optimal balance between simplicity and complexity is a key issue for designers of all sorts. This is especially true for the emerging role of conversational software designers (VUI / VUX) as voice platforms such as Amazon’s Alexa and Google Assistant grow in popularity. (We’re actually starting to bring on English-majors to lead voice design.) From Amazon’s perspective, creating an extra authentication step creates friction, and may slow users down, reducing the likelihood of a user completing a donation. Yet, without this extra step in the conversation, we see potential risks like this.
“As you add capabilities to your skill, make sure you don’t introduce unnecessary pain points or friction.” — Alexa Skill Builder’s Guide
Interestingly, Voice Purchasing is enabled by default when you set up an Echo device.
Given this, it would be nice to see Amazon:
What other ideas do you have to improve this experience?
If you would like to lock your account down to prevent unintended access, I recommend navigating to the Voice Purchasing section of the settings in the Alexa app and taking the following steps:
After you’ve added some security, don’t forget to try out the feature (on your own device) and donate! Here is a list of the currently supported charities.
https://pay.amazon.com/us/alexadonations
Don’t have a favorite? Try mine:
Give hope to kids with cancer_Families never receive a bill from St. Jude for treatment, travel, housing or food - because all a family should worry…_www.stjude.org
TL;DR: If you have one-click purchasing set up on your Amazon account and own an Alexa, you should update your Voice Purchase Settings in the Alexa app now. Otherwise people can make large purchases from your account.
We are all-in on voice technologies. If you’re interested in connecting to talk about how you can engage audiences with voice, please reach out!
If you enjoyed this, feel free to clap 👏Follow me on Twitter.