Chatbots have an operating system problem

Sophisticated chatbots have, in no uncertain terms, emerged as a reality that businesses and consumers must prepare for.

The media is all over this one. At the same time, looking at the spaces where these bots operate, the number of separate messaging programs with active users in (or headed into) the billions has created a fragmented marketplace.

At the top tier there’s WeChat, QQ, WhatsApp, Facebook Messenger, Apple iMessage, Skype, and Viber, with SnapChat, Line, Slack, Kik, Google, and Samsung holding the potential to join the billion user club if and when they see fit to make it a priority.

There are hundreds of messaging apps with small and mid-sized user bases, not to mention the vast audiences that prefer SMS, email, and services like Twitter for communication. It’s a good problem to have.

And yet, it is still a problem. It all adds up to a bewildering quantity of platforms for a business planning its own bot launch to consider.

Why it’s a problem

The situation with messaging apps today actually looks a lot like the mess of mobile operating systems available at the beginning of the smartphone era when Symbian, Maemo, MeeGo, Blackberry OS, Bada, Windows Mobile, iOS, and Android created a chaotic market. At the time, businesses had to pick their poison and swallow hard if they picked the wrong platform.

And the messaging landscape today? It’s even more challenging. Just about all of the major players above have created their own bot platforms or have announced plans to in the near future, which creates further fragmentation.

The mobile OS market at its most fragmented featured about 10 different operating systems. With messaging apps today, it’s more like 100, with four of those in the more-than-one-billion-users club (we’ll call them the B-messengers). Imagine a future where almost every business must create a bot for each major messaging app — the world literally may not have enough developers to quickly provide that many bots for that many messengers.

While emerging platforms make it simpler for businesses to work with bots and to deploy them across multiple messaging app platforms more easily, none of these solutions is able to address how to easily make bots on every platform.

Existing bot platforms function like a layer over a business’ back-end, aimed only at solving the narrow issue of bot development. What these bot platforms don’t do is deliver a true “digital core” to a company that can serve as a cloud back-end for the entire company.

There’s a growing need in this brave new world of bots to answer some key questions about function and sustainability.

• How can a business make bots that play well with all existing core systems and business processes?
• Where should bots be hosted?
• How quickly can the logic of a bot be changed after it has been created?
• How can a bot be connected not to a single messaging app, but to multiple platforms?
• How can we analyze a bot’s performance?

Businesses now stand at the beginning of a bot revolution, making it extremely important for companies to select the right IT strategy from the start. What businesses need is a single operating system for bots to emerge, one that makes it possible to very easily create, set up, and host bots for any messaging app or front-end.

What would a bot OS look like?

For starters, businesses would need to have systems that allowed the creation of back-end logic for any bot, the power to connect a bot to any internal or external service as needed, and the ability to react to users’ commands in real time.

This will require an event-driven architecture that anticipates and is able to respond to a tremendous breadth of scenarios. Once a basic bot is created in this operating system, it needs to be easily accessible via the desired messaging apps, web, mobile apps, and any other appropriate channel. It must be a full OS for the business it serves. A true “digital core,” if you will.

Comprehensive bot management goes beyond simple bot production and deployment. It needs to not only deliver fully capable bots to any external channel, but also integrate with ease to access all of a business’ core systems, preparing those businesses to thrive in our new bot-based reality. Welcome, bot overlords.