Understanding where to begin with chatbot technology takes monumental effort. Many chatbot implementations fail because business leaders neglect to ask the right questions. Other times they fail because they make the wrong assumptions.
That’s why you need to have a basic understanding of artificial intelligence and its limitations, and you also have to consider long-term maintenance since chatbot platforms can’t fix or update themselves. A backup plan is also required in case interactions get complicated and users want to speak to a live rep.
Failure to consider any of these can make you ask the wrong questions and in turn cost your organization hundreds of millions of dollars in unforeseen costs, not to mention it can leave customers with a sour experience.
So today, we’re going to set the record straight by debunking some commonly held myths regarding chatbot technology.
8 Common Chatbot Myths Debunked
Myth #1: Chatbots are only used for technical support.
Reality: Chatbots have had their debut on the technical support frontlines. Today’s bots offer more promise than early technology and are expanding into new areas such as sales, customer training, and customer success.
Pacesetter sales organizations are currently using chatbots to assist sales reps with qualifying and understanding their prospect’s business challenges before scheduling initial meetings. Meanwhile, education teams are using them to assist customers as they progress through online training and customer success organizations are allowing customers to renew contracts themselves through chatbot portals. Their use case is widespread, and the applications are endless.
According to the results of TSIA’s 2019 Technology Stack Survey, 25% of surveyed Sales Organizations, for example, have adopted some sort of chatbot platform while 55% plan to invest additional funds into them within the next two years.
Myth #2: We can build a chatbot ourselves.
Reality: You can certainly bring together resources to build your own chatbot, but why would you when you’re way better off going with a third-party solution provider? If building chatbots isn’t one of your company's core strengths, dedicating too many internal resources will only result in significant investments and a lot of rework, not to mention the costliness of having to make fixes later enterprise-wide if things go wrong.
So, unless AI is a core competency of your company, look for “off the shelf” chatbots with embedded AI and machine learning. This is going to allow you to realize results faster, and leverage libraries of terminology and common problems right away. In other words, don’t reinvent the wheel.
Myth #3: A chatbot will fix our broken website.
Reality: An easily navigable and comprehensive website is almost always going to be the quickest and most surefire way to demonstrate tech savviness and improve self-service. It’s also going to be an indicator of where a company’s priorities are. For example, having a great website is a crucial step in showing that your organization is truly customer focused.
In TSIA’s 2018 State of Knowledge Management, Distinguished VP, Technology Research, John Ragsdale, covers six fundamental ways that business leaders can dramatically improve self-service:
- Google Index Your Content
- Offer Unified Search
- Enable Multiple Paths to Content
- Embed Self-Service into Your Products (which can include your chatbot once live)
- Create a Video Library
- Make Your Site Mobile Friendly
Overall, unless you have a mature content structure, search capabilities, and high usability, it is unlikely that a chatbot will be successful, so make sure you have the fundamentals down.
Myth #4: We need a team of data scientists to implement and manage a chatbot.
Reality: This goes back to the second myth about using packaged solutions. Thanks to recent advancements in chatbot technology, teams of data scientists are no longer required to effectively manage or even implement a chatbot.
Off-the-shelf chatbots come with business-user targeted controls for building common problems and machine learning without the need of a developer or a data scientist. This means your existing team can manage the maintenance of the system, and not have to wait for the availability of your company’s limited and in high demand data science resources.
Myth #5: Customers don’t want to use chatbots.
Reality: It’s estimated that 50% of employees are now comprised of Millennials or younger generations. These generations happen to be the most comfortable interacting with chatbot technology and they happen to be the same individuals searching for answers through your website.
Customers specifically and increasingly prefer chat and digital chat and that is the experience customers now expect to have. Overwhelmingly, customers prefer self-service to assisted service, both for service and sales. It is true that many customers are not chatbot fans because most frankly don’t work, but they will try it. If the customer has a good experience, they’ll keep coming back.
Also, be sure to offer a warm transfer to a live support or sales agent if the chatbot isn’t able to effectively assist the customer. This is especially important during complex interactions such as those that are sales related, or support issues that are too technical or involved for self-service.
Myth #6: If we have a search option, we don’t need a chatbot.
Reality: Introducing effective cognitive search capabilities to your enterprise is crucial in helping find content or optimizing deflection rates, but it won’t fix everything. The real problem is that customers often search for a single word or a short phrase, and even the best search technology doesn’t have enough to work with.
A chatbot introduces a conversational experience, so it can prompt the user for additional information in order to do a better search. This makes all the difference, as customers are more comfortable with this natural interaction that chatbots offer. That’s all a chatbot really is: a conversational search experience.
Myth #7: We know what customers want and can script it.
Reality: Predicting what your customers are going to ask is nearly impossible. Customers often go to your website specifically for questions they would never have called you about, so all the reports and FAQs from your sales or service call center aren’t necessarily what people are looking for with self-service.
Additionally, scripted flows are frustrating when the option you want isn’t there. So while you may want to build some scripted flows for common problems, such as password resets, problems processing a payment, or filling out an online form, these should be options the chatbot recommends based on what a customer asks, not by presenting them with three things they can do and if that’s not what they want, tough luck.
Myth #8: We should focus initially on our more complex issues.
Reality: Designing chatbots to handle very complex issues from the very get-go is a recipe for disaster. It’s easy to think chatbots can easily script diagnostic steps to troubleshoot and resolve sophisticated problems. Afterall, we’ve all been promised the amazing conversational and linguistic capabilities AI and NLP are supposed to offer.
But, in reality, you should start with the easiest and most repetitive problems, and as your library of terminology and FAQs builds, you can introduce more complex capabilities. TSIA Members who have tried to tackle complex issues first with a chatbot have failed, as the implementation wasn’t mature enough to support it.
Conclusion
If you’re interested in increasing your organization’s case deflection or really just trying to understand a bit more about how chatbots work, be sure to check out our new report, The Chatbot Comeback. TSIA highlights recent technological advances in chatbot technology and identifies some of the risks inherent with implementing such tools. We also look at how chatbots can be leveraged in other areas of the enterprise such as sales, customer success, and customer training; so there’s something here for everybody.
Also, feel free to contact our Membership Development team if you have questions about how you can leverage chatbots to drive business outcomes.
Finally, stay tuned for our upcoming research report The Path to AI, where we’ll reveal best practices and technology companies can implement for a positive customer experience and improved automation.
Thanks for reading!