Sign up to get full access to all our latest content, research, and network for everything customer contact.

How Do I Know if My Company Needs a Chatbot?

5 questions to ask and answer before making the investment

Matt Wujciak


Chatbots can be expensive and complicated to build and maintain, but many organizations are foaming at the mouth to deploy them. It’s one thing if you’re Facebook and want to quickly acquire a startup that builds chatbots to aid a new product (just a casual new cryptocurrency known as Libra), and another if you’re, well, anyone else who can’t afford such a costly endeavor. 

Whether you’re looking to replace agents or improve them, the ultimate goal is to increase the quality of customer experience over time. And while bots may be a throwaway investment for billion-dollar global brands, many smaller companies are actually ready to empty their pockets in pursuit of this shiny new toy. Here are a few questions to keep in mind before you find yourself deploying your customer service team to aid faulty chatbots instead of the other way around. 

Read More: CCW Digital Special Report - Chatbot


Do I have the scale that makes this level of engineering worthwhile?

First and foremost, it’s important to recognize the scalability of the company. Are you receiving thousands of customer chats or calls via contact centers with hundreds of agents? Remember, it takes transcripts from that volume of interactions to generate data needed to properly train the A.I. that powers the virtual agent. Yes, bots relying on cognitive, machine-learning technology can evolve on their own, but they still require massive amounts of existing data to get started. 



What would the customer most likely be trying to accomplish? 

If you were one of your brand’s customers, and you were told that a chatbot has now been made available to aid you in some way, where would your mind wander first? What issue would you be hoping to address (or pain point would you be hoping to avoid) with an automated platform?

The intent of a telecom provider’s customers, for example, might include “fix my nonworking service,” “reset my password,” “help me move,” or “upgrade my service.”

Once you understand these common issues, you’ll know where bots can make the most impact within your customer experience journey.

Keep in mind that even the most advanced chatbots can require hundreds, if not thousands, of interactions to perfect their responses; the number rises exponentially for complex issues. It therefore makes sense to start with straightforward intents that the bot can quickly understand and address.


What level of authentication is required? 

Authentication is tough when running an application on someone else’s platform. That’s why the most advanced banking skill on Alexa, Capital One, not only requires a login to set up but can deliver no more than your current balance and the last five charges on your bill. If you’re a Capital One customer and want to do anything sophisticated, you’re better off on their app or website than talking with their bot on Alexa. It’s not you, Alexa. It’s Capital One. Now play Old Town Road.


Do I have a specific project to pilot?

Trying to launch a bot with full company knowledge from step 1 is a kamikaze mission. 

Companies that have undergone the guinea pig era of chatbots and come out alive have found it useful to start with piloting a specific issue.  Once you identify a common pain point, you can begin to start thinking about conversation design and triggers. For example, if you identify a problem in your company where customers have trouble with navigating your website or Facebook page, or are unsure of what to purchase, you may have the chat pop up after the user is on the site for over 20 seconds without clicking a particular keyword. More manageable from a data standpoint, a pilot program also helps you assess the impact of the bot while containing any unexpected consequences or fallout.

As you learn more about how the bot is working (and as your bot learns more from its interactions), you can roll it out to other use cases and departments.


What is my time frame? 

As noted earlier, bots may need to field thousands of inquiries to build the data processing required to perfectly understand a customer’s intentions. Only you (by virtue of customer and agent insights) can answer the question: is it really worth the effort? If you’re a small company fielding a few, yet critical, inquiries a day, you’re likely going to offer a more efficient customer service experience fielding them with a human touch. If you’re Facebook, you might sit back and enjoy the ride.