Content Bots, E-Commerce bots, Application Bots

In my last post I made the comparison between the Facebook Messenger Bot platform and the Facebook Apps platform. In this post, I’m making a different comparison — bots vs web sites.

Historical development of web sites

In the 90s, web sites were mainly content sites. Sure, there were a few places you can order things online, but online shopping was mostly a niche and full fledged applications with functionality (think Facebook itself, or Myspace) was mostly non-existent.

E-commerce gradually became more mainstream, and it took a good few years before the public is more accepting of it and put away skepticism. (a lot of people felt uneasy about putting in their credit card information over the internet back in the early 2000s.)

Around the same time, full fledged web applications started popping up everywhere. At first, they were simpler and resembled regular content web pages, but people created a variety of applications in the form of web pages, including games (e.g. Kingdom of Loathing, which launched in early 2003 and still exists today) and social networks (e.g. Friendster). Then AJAX came along and all of a sudden, web apps felt just like native desktop applications with just as rich of functionalities and UX.

The trend for bots

I believe bots will go through a similar trend, though it may happen a lot quicker. (or not — it would be because of failure to gain consumer traction in that case.)

From the community of bots enthusiasts and creators, I feel that most people are in the “content web sites” stage of bots. Tools like ChatFuel and Octane.ai provide very easy ways for non-developers to create content bots, and even people in the mainstream like celebrities are creating them or having it created for them. I call these content bots because in essence they simply contain content for users to browse (albeit in a chat interface, instead of a web page interface). They’re both created and consumed like content web sites. In a way, services like ChatFuel and Octane.ai are equivalent of tools like Microsoft Frontpage and Dreamweaver — they allow non-coders to create content sites/bots easily.

Then, there’s e-commerce and full fledged applications. I mentioned in my last post that it’s clear Facebook intends for the Messenger Bot platform to be used by companies like airlines and online shopping companies, for them to send things like order receipts, flight check-ins and do customer service. If it wasn’t obvious enough from their keynote, they even provide API like specific templates for airline boarding passes!

In terms of e-commerce, I believe companies will pick up on bots tech over time and deliver all of what I mentioned above through bots. There will be more and more development in the area, but they’ll mostly happen within corporations themselves, or contracting out to dev studios. They need to and will happen, but it’ll be predictable and less exciting for the entrepreneurs and side-project enthusiasts.

Then we have full fledged applications. This is where I believe the current green field is. It’s funny to talk about bots being “applications,” because the community typically likes to pitch bots against “apps” (mobile apps, specifically) currently. What I’m talking about is bots that do a lot more than content, and a lot more than e-commerce. Bots that take heavy backend coding, and do heavy lifting, like what Kingdom of Loathing and Friendster did on web pages in early 2000s.

As far as I can see it, there are very very few developers working on creating full fledged applications on the bot platform. (at least the Facebook Messenger Bot platform, which is my primary interest and knowledge area right now) For every 10 developers working on the bot platform out there, it seems that 9 of them are only interested in creating the tools. Tools like ChatFuel that helps others create content bots, and tools like bot analytics, or bot design utilities. Tools are fine and all, and I believe everyone is learning from the past few generations of platforms (web, mobile) where a lot of the winners were people and companies who created great tools. But you also need creative developers who actually come up with ideas and create real application bots too. When 9 out of 10 developers are making tools and only 1 of them is making actual application bots on the platform, the balance is a bit skewed.

The good news is, for those of us who’re actually interested in making the apps, not the tools, this is a green field ripe of opportunities. Now, as I also mentioned in my last post, I believe the Facebook Messenger Bot platform isn’t quite there yet in terms of providing all the ingredients needed for bots to grow rapidly. But that’s not to say we can’t get started now, and that we can’t creatively use what we’re provided with now to try and do it.

My last experimental project, Coffee Break (mini-games bot), was an attempt at that. All I really did was to put in a few simple Javascript-based games, put in a leaderboard to store users’ high scores, and a score sharing function. I believe there’s an opportunity for a lot more on this platform, despite the limitations I mentioned.

In conclusion, I feel that we now have a decently good platform to build apps on top of, but with very few developers taking advantage of it. We’re likely one killer app away from the Messenger Bot platform taking off on its own, and everyone rushing into the platform to develop their killer apps. This will be what it takes to bring Messenger Bots beyond the content sites stage that we’re currently in. I’m looking forward very much to the arrival of that killer app.