Sometimes you create something which you think will work and ultimately it doesn’t. The project I created called BnBChatbot is one of those projects and unfortunately now I am signing the official death warrant for it. It was a cool idea and I enjoyed spending time learning about chatbots and the potential they have. Messenger marketing on Facebook has huge potential but this project just didn’t get the traction needed to justify me ploughing further time into it, so it is going on the scrap heap.
Why Kill BnBChatbot Now?
It is a valid question and the truth is, I stopped using it for Roomfilla. If I am not using it for my own project which I thought I built it for to solve the issue I was having, then perhaps the issue didn’t exist or the product I built, didn’t solve the issue as smoothly as I thought. With Roomfilla, a percentage of the business is through upsells of additional services such as transfers, tours and all that good stuff. We invest quite heavily into these sites such as Explore Phi Phi, Kathmandu Bus, Siam Ferry and more peripheral websites.
I spent quite a lot of time building such additional service booking functionality directly into the chatbots I created. The chatbots however were simply not converting as well as the previous funnels that were created. We saw a dip in Stripe transactions during the months when I was heavily forcing all our guests to use bots primarily. We then did a hybrid approach of bot and old style funnels and saw these conversions increase again but it started to become apparent, that why bother with the bot if we can automate the other areas of the communications into such a way that makes sense.
Are Chatbots a Gimmick?
Possibly. They are new, they are exciting and they are different. We had a lot of users interact with the bot during our months of testing and found that guests did enjoy clicking around and probably clicked around more than they needed to purely because it was interactive and different. Chatbots are still not the norm and I think that some people who interacted with the BnBChatbot were having their first interactions with a chatbot like this in this setting which was novel for them. Novelty didn’t lead to increased sales or dramatically reduce our workload so unfortunately it led to the question – what is the point of keeping this going?
BnBChatbot was initially built so I could learn more about chatbot frameworks and how they work. It was built to solve our needs of guest communications and as a potential product we could sell as a secondary service. Building bots for other customers is a truly custom job. The more I interacted with interested parties, the more I found out that everybody has their own systems and wants a bot to work specifically in those systems. Multiple language requests, a bot feeding directly into a complex system and other custom integrations were asked for. There didn’t seem to be a way to mass market these bots which is what I wanted to do. Price low, create a sort of bot template and sell to the mass market. That did not exist and I do not want to be in the business of being a bot builder. There are plenty of those already existing.
A Phoenix Rises from the Ashes
Side projects like BnBChatbot are great. All in all, it was a fun weekend / evening project which I tried to not let consume too much of my idea or focus from my main work with Roomfilla. That was great and it is fun to build something like this from scratch, it stimulates the brain and teaches you a whole bunch of things as you solve the challenges and learn about a whole new area of tech. Side projects often ensure you develop new skills which you may not know at the time, can be applied to other areas of business and life. BnBChatbot is no different, it was made with the intention of ‘how can I automate guest communications?’. We may have just solved that in another way, so thats excellent!
Guest communications is such a tricky thing to automate as guests ask all sorts of questions and there isn’t a one size fits all answer to these questions. The best way to do it when managing a large portfolio of properties is to provide as much information as possible along with the communications and speed up things such as foreign languages being automatically translated to English and have them all appear in the place you spend all your time, for me this is Slack.
With a bit of magic from Zapier, Slack, Gmail API, a teeny Python script, email parsing and a database app. We have created something far better than BnBChatbot ever could have been. Far more efficient. Far more advanced. Far more.. just works. I’ll be going into this in more details in upcoming blogs when I start a series on ‘How to create a DIY Multi Inbox for your vacation rental properties.’