I continue to be extremely intrigued (like the rest of the market!) by the potential for chatbots to revolutionize B2C, business enterprise applications and possibly even B2B interactions. As several other pundits have already opined, this is likely one of the biggest opportunities for developers since Apple opened up their App store. Time will tell how this really plays out, but this is a great time to get in on the ground floor and
As I’ve already written, something inside me really wants to do a start-up. However, I’m not too interested in going the VC funding route. I’m also not interested in scaling massively without a strong fiscal foundation. Given my personality (conservative!), I’m probably going to try to boot-strap this, scale accordingly, and try to build something that can hopefully be profitable. Definitely not easy, and not saying this is the “right” way to do it, but it certainly appeals to me. This is why I really enjoyed reading how MailChimp made it from start-up to a real business in the most un-Silicon Valley way. So this will remain back of mind as I analyze possible chatbot ideas.
I’m currently pondering a few simple business models:
- Build bots for local businesses
E.g. restaurants. Provide their customers another interface to interact with the brand- ask about the menu, place orders, reservations, learn about specials, the most popular menu items, earn/redeem points, etc. The most important benefit would be introducing new customers to the brand — not quite clear how bots would enable this yet.
The same could be extended for retail, etc.
Financial model — one-time fee for bot development, and a monthly hosting fee. Tiered service model, something like Prototype, Mid and Pro.
Expect a lot of competition, so the advantage goes to the early movers. It’s still early days so now might be time to strike.
- Related to above, build bots for local franchise stores
Grocery stores, fast food, retail, etc.
- Similarly, build bots for bigger brands. E.g. GQ magazine, Nike, etc.
The benefits of having a bot are obvious. But you don’t just walk into Nike asking for a contract to develop a bot. Need some serious credibility to get into that game. That’s why I like the local business angle — learn to walk before you can fly.
- I wrote about this earlier, but build a restaurant recommendation bot. Others have already developed similar bots, and it’s going to be hugely competitive. Also, not quite clear how you make money here? Ads?
One theme that cuts across all these ideas (besides them being related to food!) is competition, and a question of how to differentiate. Since everyone will have access to the same platforms, and the same NLP libraries, this will not be a source of differentiation. An area for differentiation might be the machine learning applied to the underlying data to provide more insightful answers to users. Another thing that bothers me is that this ends up trading apps for bots. Will users want to interact with so many bots, i.e. a different bot for every brand? The advantage bots have over apps is of course having less friction as there’s nothing to install, etc., so maybe that won’t be a huge concern. You can pull up the appropriate bot in the bot search window and then shut it down — very unlike apps.
- SaaS application like Salesforce for SMB
Start with a simple tool that allows tracking of leads, opportunities, issues. Provides insights via machine learning. Bot interface to interact with underlying database.
Now that I’ve got some basic ideas flowing, time to start playing with the technology and build a prototype!