Investors Are Going Botsh*t Crazy

Bots. They are everywhere in startup land. It feels like a gold rush — a bot rush. And, there is a backlash, as well — a botlash. Whether it’s more and more people using Slack and other messaging apps and come across these new bots and features, the trends set by Y Combinator’s Demo Day (where, by my count, four very interesting companies launched in this space), or one of the 101 pitches investors see from founders, bots are everywhere.

In this rush, there is surely opportunity, and also noise. At a high-level, “bots” are attractive because it is a lightweight format by which a startup or messaging app or Amazon Echo can deliver services to users on the simple end of the spectrum, and perhaps one day automate, predict, and conduct work for us as it learns from our interactions and behaviors, as well as learning from other APIs they interact with.

The noise created by today’s frenzy makes it hard to find signal. As a I result, from an investment point of view, I see the bot world in three (3) categories:

1/ Vertical-focused B2B platforms w/ SaaS business model: This is targeting a specific end users (usually businesses) that need to create a bot for news, or commerce, and pay the startup for a suite of tools to do so.
2/ Selling picks & shovels: When you sense a category is going to be big/important, but don’t know what the end-users will use, one can invest in the developer platforms and services that the bot creators will need to make their bots.
3/ Direct to consumer (or prosumer) plays: This can be seductive because messaging apps are huge (easier to get users, distribution), but also dangerous because today’s platform can become tomorrow’s gatekeeper.

How do I think this will all unfold? I have no idea, but I think the three investment strategies above make sense, all for different reasons. I have an investment in 2, looking at 3, and looking for 1. I’ll let you know when I find out.

=====

Separately, on the consumer side, I’ve been thinking about what I want personally in a bot service. I wanted to write that below in case anyone has built it, is building it, or is looking for an idea. Here’s what I want:

1 – I want to create my own personal bot, Semil-bot.
2 – On web or mobile, I go to your page and authenticate with Facebook. This has to be ID-based.
3 – I connect a range of services and accounts, beyond financial like with Digit etc. Definitely Gmail and Google Calendar, like Paribus.
4 – If I give Semil-bot the keys to my digital kingdom, I want it to not only dive in and learn everything about me beyond Facebook, I want it to crawl a network of specific APIs for financial services, insurance, digital media, social networks, etc. and constantly be on the lookout for my best interests, to save me time and money without me having to think about it or execute any actions.
5 – Semil-bot would send me a daily update email with actions that it’s taken an easy way to reverse, pause, or suspend commitments. It would work for me, on my behalf, 24 hours a day. I would pay $10 or even more for this service.
6 – But, this isn’t just a weekend hack project. As the system gathers more information about me and other APIs, it needs to learn and form a real brain. It needs to learn over time and make inferences. It needs to anticipate things. It needs to deliver a service that saves me both time and money, reducing cognitive load and becoming a trusted digital agent to act on my behalf.