Slacking in a Suit
My friend and I worked on a Messenger bot called Sunny. It was a chatbot that inquired about a user’s day and reminded them about past things they’ve shared in the chat. We envisioned a potential use case for people who experience depression or Alzheimer’s, but Sunny needed exemplary conversational skills so the UI needed to be very well thought out.
That’s the challenging part with chatbots.
Like the mobile app craze, every company wants to build a chatbot. It makes sense: adoption of messaging platforms, adaptation to customer preferences, and ability to automate task execution makes them favorable over clunky mobile apps.
However, without a clear purpose and strong conversational cues, the user journey can go in a million different directions, and that’s difficult to build for.
After working in the corporate world, I think the enterprise space is more predictable, and chatbots have an opportunity to eliminate painful processes. I like to think of chatbots as ‘efficient file explorers’ since they can navigate layers of interfaces in mere seconds while you toil through them.
An enterprise service like Skype for Business or Slack should build and sell an executive assistant chatbot that eases office tedium:
I need to find free time between invitees, go to a portal to reserve available rooms and send out invites with all the information. I would rather tell a chatbot about who I need to meet, where and when, and have it come back to me with the time, place and confirmations from invitees.
Time saved: 15–20 minutes per invite
2. Coffee & Meals
I try to place an order on the existing app, it fails to work ~3 times and I decide to take an elevator down. I then wait in a long line, place an order, wait for the food, and take it back upstairs. I rather order through my chatbot and have it tell me when my order is ready. It could send orders to smaller kiosks that can’t afford ordering apps, and compete with existing third-party apps.
Time saved: 30 minutes per food run
3. Booking Travel
I need to input my traveler’s information, go through pages of flights, hotels and rental cars on an online booking portal. I would rather tell a chatbot to book a trip to Atlanta and a hotel for 3 nights. It would meet company travel policies and send me the itinerary with a reminder the night before I’m flying.
Time saved: 30 minutes per booking
There are many other tasks a chatbot can do that saves digging through interface. Print the PowerPoint presentation titled Customer Experience. Remind me to follow up with Finance for updated numbers. Find me someone that can teach me about Blockchain at the office. The list goes on.
Integrating with the office ecosystem would mean jumping through bureaucratic and technical hoops, but the enterprise space is most feasible because of the purpose-driven requests. It’s also worth it — can you imagine all your employees having a personal assistant, and what that would do for their productivity and employee experience?