Messing about with bots (alpha)

It’s not that often that a new media comes along, so over the Christmas break I created my first chatbot. 

I’m not a developer, so I limited myself to playing with Chatfuel to get a hang for the basic principles of a bot on Messenger. It meant creating a Facebook page for my company, unfortunately, as Messenger bots can’t be linked to personal Facebook pages.

Test it please - I’d love your feedback!

It’s very much in alpha mode right now, so if you’re reading this I probably asked you to take it for a spin and give me some feedback. Thanks in advance ;) 

And if you’re really curious, there’s a little more info below on the What, Why and  How.  

However, I urge you to test the bot first, because: 

• it’ll make the following easier to understand, particularly as I’ve probably tweaked the bot a little since publishing this in late December 2016
  
• any feedback you give will be more ‘honest’ if you don’t know, as you begin, what this bot is trying to do.

So give it a test, and then come back here if you have any questions.

Goal: Offer Good Stuff To Read 

The bot builds on this Hub, and essentially fulfils the same purpose - it wants to help you find interesting stuff to read: 

Essentially it presents Good Stuff To Read corresponding to the user’s input - i.e., the user feeds it a keyword, and it’ll come back with the three latest articles which I’ve Hubbed (read, recommended and published here) and tagged with that keyword. 

Basic assumption: you need the same interests as me! ;)

So if you were to answer its Welcome message, above, by typing in “news”, you’d get (enlarge this):

It also provides a bot-version of Top3ics, my newsletter. Whereas the emailed & online version might present anything from 10-20 interesting resources relevant to three topics, the bot just presents the best three articles I’ve recently Hubbed (i.e., one per topic), as well as links to discover more.

Instead of subscribing for an irregular email of interesting resources, in other words, users can simply ask my bot at any time for whatever interests them. The Top3ics content is updated manually, but that takes no more than a few moments, and adds a new wrinkle to my personal content strategy:

Note that these screenshots are taken from my explanatory post on the Chatfuel community. At that point I was pulling those stories from this Tumblr, but that meant the user was being sent here, when they presumably would prefer to go directly to the full article (”Resource”, in the above botflow). It also meant that they weren’t getting any images. 

Since then I’ve fixed this by pulling the content from my personal Diigo library, although I still send users here for 'read more’ links (generally tag menus). 

Takeaway: bots are easy if …. 

Bots are easy … if you already have a content strategy, underpinned by a well-organised content database with online APIs and/or RSS feeds.

If not … well, how can you expect a bot to provide anything useful from a disorganised digital pile of static content?

Wishlist: Faceted Search via Bot

I wish I could code. If I could code, my Hub would not be on Tumblr, and would offer users faceted search, allowing them to combine tags together to refine their search and find the most relevant resources from the 2640-and-counting resources I’ve curated here.

And if I could code, the Hub would have an army of Bots (Messenger, Twitter, etc.) that would do exactly the same thing. Faceted search via ChatBot is a no-brainer: instead of just presenting 3 resources relevant to the user’s keyword, I would also present Related Tags, allowing them to refine their search, drilling down through the mountains of content to find the diamond they need. All with one thumb, as they wait for the bus, or a job interview, or a sales call, or a date.

If only I could code.

Further reading

• More stuff about bots
• About this Hub
• For those who discover this later: if there are any other posts about this particular bot, they’ll be tagged mybot.