Not All Chatbots are Created Equal: Picking the Best Bot for Your Business
So, you’ve decided you want to incorporate chatbots in your customer service mix—great idea. But before you start considering options, it’s crucial to understand the different use cases for chatbots and determine which one is right for your company’s specific needs and goals.
It’s a mistake to think of bots as a one-size-fits all solution. Take the case of Nordnet and SEB, two leading Swedish banks, both of whom incorporated the same virtual assistant, Amelia, on their websites. Nordnet tasked Amelia with speeding up new customer onboarding and improving customer satisfaction but ended up “firing” Amelia after getting underwhelming results. SEB, meanwhile, was having success with Amelia, but was using it in different ways, to achieve different goals. The takeaway? Even best bot may not get results if it’s not the right fit for what your business wants to achieve.
Broadly speaking, there are three main use cases when it comes to implementing chatbots for business..
- Questions and Answers
The simplest use case for bots is answering basic, routine customer questions. Basic bots pick up on keywords in customers’ entries and respond with canned answers, written in advance by humans. These scripted bots ultimately aren’t much more than interactive FAQ sections, but they can be effective in intercepting simple inquiries and keeping live customer service agents free to deal with more complex issues. They’re also easy to get up and running and relatively affordable.
As a result, a lot of companies start with this use case, but many quickly find that they want to ramp up to more powerful forms of AI. Scripted bots force customers into a constrained workflow, which quickly breaks down when the customer goes beyond those scripted answers, leaving them frustrated at best and annoyed at worst. (Remember Microsoft Office’s infamously annoying and useless “CLIPPY”?) With sophisticated AI chatbots becoming more and more common, customers can often tell when they’re interacting with a scripted bot, and will seek human help as soon as possible, defeating the purpose of having a bot in the first place.
- Contextually Aware Conversations.
Increasingly, companies want to do more with chabots than just deliver canned answers to simple questions. They want to provide value to their customers with bots that can answer complex queries, in human-like interactions where meaning and context are preserved from response to response. To do this, you need chatbots with and Natural Language Understanding (NLU) capabilities that can leverage structured data to have natural conversations. These AI chabots can understand the intent of a phrase or sentence and respond to that intent, not to a keyword.
For a sense of the difference between these AI Chbots and scripted ones, consider a question like “Where’s your nearest location?” A scripted bot, triggered by the keyword “location,” might respond with a link to a store locator tool, or a list of stores, where the customer could figure it out themselves. A conversational chatbot, on the other hand, could respond by saying “I’d be glad to help. What’s the address where you are right now?” Once the customer responded, the bot could say, “OK, are you looking for a store with a drive-through, or walk-in only?” and use the customer’s response to provide the appropriate answer. It’s a much more helpful, human-like experience.
- Tasks and Transactions.
The most complex scenario is using conversational AI bots like those above to actually accomplish tasks, not just have conversations and provide information. This is a much bigger financial and logistical undertaking for companies, because it requires linking the bot software with various back-end systems. If, for example, an airline website wanted to use a chatbot or virtual assistant to book flights for customers, the bot would need to interface with a whole network of back end systems, from customer identity verification to schedules to fares to frequent flier miles systems—and do it all quickly and securely. Companies considering more advanced use cases like this, which require a big outlay of time and money, need to be sure the ROI is worth it.
The potential of bots to enhance, streamline and improve, and streamline the customer experience for your business is limitless. But make sure you’re choosing the type of bot that not only matches your company’s goals now, but that’s robust and flexible enough to ramp up as your goals change.
To learn more, check out our latest whitepaper: Conversational AI Solutions — The Happy Medium with Bold360