Are bots solving any problems?

I woke up to a series of Benedict Evans Tweets this morning about bots and service discovery on the internet. The gist was this:

  • Everyone’s talking about bots
  • Query whether moving services to bots solving any problems or just moving them to a new UI
  • We’ve mostly solved how to get things we know we want, but not of things we don’t know we want

He doesn’t explicitly say that bots don’t solve any problems but I think that’s the implication.

Most of the excitement about bots comes from the fact that they offer an extra level of convenience. You’re using Messenger to arrange a meeting with a friend and you can book an Uber right there without having to get your phone out and boot up the Uber app. Maybe there’s an AI running in the background that recognises you might want to book a cab and works out the start and end points so you can make the booking with one click.

That is all very cool, and I’m a believer in the power of small extra conveniences, but a small extra convenience is all this is. Pulling my phone out of my pocket to book an Uber isn’t hard. If all that bots are doing is offering a small extra convenience then for the mass market the user experience has to be seamless, and that means not learning new syntax or anything that is more difficult than using apps or the web. That’s a high bar.

The bigger problem today is discovering things we don’t know we want. I say this as an investor in several new ecommerce services and products that people love when they get their hands on them, but aren’t searching for on Google. Facebook has some good tools these days which our partner companies Live Better With and Lost My Name have used to good effect, and Instagram is becoming more useful, but it’s still a significant challenge. I’ve long been fascinated by the potential of influencer marketing as a solution but have yet to find a scalable solution in this space.

I’ve had a long day and prefer to blog in the morning when I’m more positive, but all of this has me wondering if messaging is just better email and that bots and conversational commerce have a long way to go to get from hype to mainstream reality.