Top

5 Reasons Why Chatbots Fail 

Jim Rembach | Oct 17, 2017 359 views 2 Comments

Share on LinkedIn

There’s five common reasons as to why your chatbots fail to live up to promises. If your chatbots are not living up to promises, don’t give up. It’s time to get smart.

Chatbot is the most complex area of Artificial Intelligence (AI). It is also the area where solutions may vastly differ from one to the next. You might have heard of a chatbot not living up to expectations or you’ve experienced it yourself.

If you have, you can easily point to five common reasons why chatbots fail.

  1. Poor chatbot solution
  2. Learning omission/poor
  3. Wrong Chatbot Generation
  4. Chatbot Hype
  5. Wrong Desired Business Outcome 

Poor Chatbot Solution

Chatbot is a hot space. There are hundreds, if not thousands of companies that want to sell you a chatbot for your organization. When anything is hot, there is also a corresponding issue of buyer beware.

Very few people that are responsible for the designing the customer experience have experience with chatbot technology.

I mention CX leaders because they are the ones that need to own the chatbot. After all, they own the customer experience and chatbots directly impact it. Also, when the chatbot fails, they have to own the consequences of that failure.



Allowing your chatbot to be owned by IT increases the chance of you obtaining a poor chatbot solution. Ask yourself, “Is IT our customer experience domain experts?”

You want to avoid this pitfall.

Learning omission/poor

There are two separate topics about learning, that are important for you to know. The first is simply, that you need to improve YOUR learning about chabots. It will help you to avoid many of the frustrations you’ve been hearing about. So, take the time to learn about the chatbot best practices

The other learning relates to teaching your chatbot. If you’ve been told that when a chatbot is an Artificial Intelligence Chatbot (AI chatbot) that all you need to do is build it and deploy it, that’s wrong.

Artificial Intelligence needs to learn. You can dramatically decrease your failure when deploying your chatbot by giving it a good head start. You have to train it, feed it data, and nuture it.

The solution you use (or build) will make all the difference in your ability to grow your chatbot to maturity. Is the process simple?

Ask yourself these questions about teaching your chatbot:

  • Does it require a lot of programming?
  • How will low/high volume impact learning?
  • How long will I need to nuture it?
  • What support do I have?

The time it takes for your chatbot to learn can take a few weeks to a few months. This variance is somewhat impacted by volume but more importantly it is impacted by the quality of your chatbot solution.

For example, is Deep Learning applied to the Dialog Manager? This will have a huge impact in learning and ongoing nurturing and maintenance of or your chatbot.

Wrong Chatbot Generation

Currently, there are three different generations of chatbots. Each can deliver business value. But scalability is a common problem. A 1st Generation chatbot is a very simple rule-based engine. You have very common questions, you can program very simple answers. These are easy to build and easily bought.

The promises may appear to get broken when a 1st Generation chatbot is pushed outside it’s boundaries. This can happen by expecting it to do more or by the customer thinking or expecting it to do more.

The 2nd Generation chatbot can do more. It leverages Supervised Machine Learning in addition to it’s rule-based construction.

The 3rd Generation chatbot adds Unsupervised Machine Learning in addition to Supervised Machine Learning and it’s rule-based construction.

3rd Generation chatbots are able to provide a customer experience that does not require as much customer effort to get problems solved faster. And we all know customers want speed. Speed is the primary driver for higher customer experience scores.

The reality is we all operate in an always on world. Your customer wants what they want, when they want it. And if you don’t provide it, they will go find it.

Chatbot Hype

It’s all the rage! Do you have one?

Media of all types and every analyst firm and consulting firm tells you, you got to have chatbots. Okay…you do. Customer experience is your competitive battle ground. And with the proliferation of mobile devices ,you need chatbots as an integral part of your digital transformation.

But all of this hype about chatbots has a slew of negative consequences. You end up experiencing pressure from all over.

One of the biggest pressures you might experience is from your executive leaders telling you, “I want this now.”

When you have that kind of pressure, you try to deliver. Anytime you need to make a decision in haste, it comes with a lot of negative consequences as well. None of them good. That is something everyone experience.

Wrong Desired Business Outcome

What do you want to accomplish with your chatbot?

ALERT: Does your goal match with your customers’ goal?

Let’s assume for the sake of this article that your expectations match with your customer.

Let’s focus on customer effort. We all know that effort is a huge contributor to referrals, loyalty and retention. So, you want to make things easy for them, correct? Definitely.

So, if your customer is using your chatbot to inquire about your products, do you want to make is easy for them to purchase them? YES

Then a 1st Generation chatbot will not deliver upon the promise.

Would like to give your customers the ability to do multiple transactions that might be unrelated to ease their effort? YES

Then a 2nd Generation chatbot will not deliver upon the promise.

Setting the Right Expectation

Too often a failure to live up to promises starts with getting the expectation right.

Your best bet is to know and understand these five common reasons as to why chatbots might not live up to you or your customers’ expectations.

Ultimately, the promise that must be kept is between you and your customer. So take control of AI before the alert is about you.


Please share:

  • “Alert – Chatbot are not living up to promises” – Click to Tweet
  • “You can easily point to five common reasons for chatbot failure.” – Click to Tweet
  • “Chatbots are a hot topic – buyer beware.” – Click to Tweet
  • “CX leaders need to be the ones to own the chatbot.” – Click to Tweet
  • “Allowing your chatbot to be owned by IT increases the chance of a poor chatbot.” –Click to Tweet
  • “Decrease chatbot failure by giving it a good head start.” – Click to Tweet
  • “Deep Learning applied to the Dialog Manager has a huge impact on your chatbot.” – Click to Tweet
  • “Currently, there are three different generations of chatbots.” – Click to Tweet
  • These are 5 common reasons why chatbots might not live up to expectations.” – Click to Tweet
Print Friendly, PDF & Email

Republished with author's permission from original post.


Recent Editor's Picks:


Categories: BlogContact CenterEditor's PickEnterprise Technology

359 views

2 Responses to 5 Reasons Why Chatbots Fail

  1. Dave Goodwin October 18, 2017 at 6:37 am (1 comment) #

    On-point Jim. I had not seen chatbot/virtual agents categorized by generation before; effective way of illustrating the stages from rules-based to true AI-powered.

  2. Jim Rembach October 18, 2017 at 10:49 am (1 comment) #

    Thank you Dave. I had a great teacher in Dr. Yi Zhang – http://www.rul.ai/fast. She has helped to guide my understanding and perspectives.

Add Your Comment (All comments are reviewed by moderator, no spam permitted!)