Bots. What are they good for?

March 21, 2016

Let’s walk through use-cases that are well-suited for bots

Things are about to get intense in the next month in the world of bots and chat:

  1. March 22–23: YC Demo Day will feature many bot, chat & AI startups
  2. April 12–13: Facebook may open Messenger to bots at F8

Let’s recap:

  1. bots are going to be huge
  2. good bots are hard to make

We’re missing something extremely important, though. Exactly what are these bots going to do to improve our quality of life?

  1. One-off or infrequent mobile transactions. Consumers want to transact with business via their mobile device, but often either the business website isn’t mobile optimized or requires the user to download an app first (which the consumer isn’t willing to do). This is a sweet spot for a bot. Examples: buy a ticket for an attraction, download a fast food coupon.
  2. Reduction of large or fragmented choice sets. The abundance of choice leads to anxiety and even depression in people according to the Paradox of Choice. A bot can gather or recall information (likes, dislikes, budget, timing, etc.) from its user, and then pull together a small set of 1–4 recommendations that meet the criteria. This makes the user disproportionately happy; you’ve saved them time AND relieved them of decision fatigue. Examples: choose a movie on Netflix, pick a restaurant for dinner.
  3. Enabling access to mobile commerce for small business. Should small business owners invest in a mobile app? Let me save you the trouble: NO! It’s not worth it, no one is going to use it. It’s inevitable that messaging will replace the phone the interface to local business.
  4. Customer support enhancement and automation. Customer service follows the Pareto principle. Most people have the same few problems. This makes customer support a good candidate for automation via bots.
  5. Proprietary enterprise solutions. Imagine you need to arrange a parking pass swap for your 200 person company. You’d be insane to build a website or heaven-forbid a mobile app! A shared spreadsheet might work, but maybe a chatbot in Slack could be just as easy, and more fun.
  6. Reimagining consumer mobile apps as bots instead. Not many users are downloading your mobile app. Maybe it’s better suited as a bot inside Facebook Messenger instead? You’ve reduced the install friction by 10X which can push your product into viral territory (k>1). Alternatively, it may be smart for startup founders to validate their assumptions with a bot-first strategy.
  7. Financial planning with actionable advice. A bot can be programmed to have visibility into multiple financial databases and unify the data into a common interface. The bot can answer basic questions: Should I invest in X? Do I need to transfer money Y?
  8. Travel accommodations and planning. Travel planning is particularly difficult. There’s a reason there used to be travel agents. Consumers have to make multiple decisions across multiple service providers, with prices and options changing every day. Messaging and bots can unify the experience into a single user experience that works well on mobile.
  9. Interactions between people and their local government. What better way to help build our smart cities, but by providing low friction access to its inhabitants? People can report issues and interact with services in real-time.

Did I miss any big ones?

If you’re interested in trying to build for one of these use-cases, Meya is now accepting beta signups to build & host your amazing bot creations.

Erik Kalviainen

I make a bot building platform called Meya. Past: ProductWiki founder. Product @ Google. U of Waterloo Engineering.

You might also like