I got another question on my Drift bot recently:
should i use drift or dialogflow? which is best?
As it turns out, this is a false dichotomy.
Not only can you have both, it might even be a good idea to do so.
And in addition, I have created a proof of concept of this on my website. The Drift chatbot already has a Dialogflow agent powering it. Unfortunately, it is still very much a simple agent because I haven't had the bandwidth to create a more complex one yet.
Drift has its own chatbots ...
Now, you may or may not know this - but Drift has its own chatbots. Except, they are not really "intelligent" in the sense of NLU-powered Dialogflow bots.
Rather, they use conditional logic trees to get people to click on buttons to carry out a conversation.
Even if they were to add some basic keyword matching like Chatfuel does, that will still fall well short of what you can do with a proper Dialogflow bot.
... but NLU powered bot frameworks work differently
NLU powered bot frameworks work in a fundamentally different way to Drift and Chatfuel style conditional logic chatbots.
And if these companies decide to implement more "powerful" bots, they will have to replicate Dialogflow features such as entity extraction and context management. If they do that, then it will make their drag-and-drop visual editors quite a bit more difficult to manage.
A couple of days back, a reader wrote this in to me about Dialogflow:
Thanks for offering lots of useful stuff about Dialogflow. That tool looks that easy when you start with it and turns to be really difficult after a number of steps.
In other words, even Dialogflow, despite its appeal of visual editing and what looks like push-button bot deployment, actually gets pretty complex once you start creating any kind of complex bot.
It is a good idea to understand when you should use Drift, when you should use Chatfuel, when you should use Dialogflow as well as understand the pros and cons of each tool. While I don't have a comprehensive course covering all of this, I do have a somewhat simpler course on this topic. It will get you started with answers to some of these questions.
Interestingly, after assessing all these tools, I have come to the view that if you can pull it off, you will get the best combination of power/flexibility/autonomy if you extend an existing WordPress live chat plugin. It requires some serious development resources, but there are really nice benefits you get from the WordPress platform out-of-the-box which can help you handle some tricky bot management questions.
If anyone is interested in learning more, leave a comment with some information on which live chat plugin you use. If it doesn't involve too much work I will write some articles about the integration. Note: if the live chat plugin doesn't already support extensions, (while it might still be a great plugin for the live chat use case) it won't be a good choice for integrating a Dialogflow bot.
- Using Collect.chat for preNLU bots
- Reader Question: How to get some sample training data for Dialogflow?
- Getting the top 3 (or top N) intents in Dialogflow: An experiment
- Dialogflow Regexp (regular expression) entity
- Using Dialogflow for educational bots
- Chatbot makes $3 million in sales in 1 hour
- Reader Question: Intent based FAQ bots vs knowledge based FAQ bots
- Should you use Dialogflow from scratch?
- 3 ways Airtable can speed up your Dialogflow prototype development
- Automatically generating a rich Dialogflow website chatbot