There is a role for bots in this world. Yet the size, shape, and scope of that role across industries is still being heavily discussed, experimented with, and iterated upon. Just to be clear, this is referring to a bot in terms of automation software that runs through the web, which is admittedly a lot less sexy than those fully fleshed-out robots of Westworld. But for businesses today, the idea of internet bots is sexy in its own right.
Bots should theoretically bring about huge cost savings and reduce inefficiencies in businesses because they ideally decrease the number of humans that are necessary for various business operations. But at the same time it is safe to say that bots are not the end all be all for independently maximizing the bottom line. There are a lot of factors that come into play, and viewing bots as a one-size-fits-all solution for automating different aspects of any business is an incredibly flawed way of looking at it.
Figuring out how bots can help optimize the user experience is an important consideration. But it requires taking a deep dive into the user experience to really understand where bots can help and where they will hinder.
The way Amazon has designed its customer experience serves as a great example. When a customer is using the Amazon mobile app to process a return for a product, Amazon makes it really easy for the customer to do so. The customer simply needs to view the online order for the product, click the clearly labeled return option, and Amazon's self-service capabilities will walk the customer through the multi-step process. It's an incredibly simple and seamless experience: letting users independently help themselves.
If the return proves to be more complicated, the customer can contact a human agent and get it resolved. But more often than not, customers handle the return on their own without ever needing to go through an agent, heavily reducing the amount of issues that agents need to resolve. Considering Amazon's massive growth and potential for taking over the world, it's safe to say the online retail giant is doing something right.
Those more complicated issues will always require a human touch. Bots just aren't technologically advanced enough to handle such nuanced issues. So the question is really whether to have a bot as a first line of defense, as opposed to self-service. Amazon says no...but many companies promote their chat functionality as the primary option for support, with a bot serving as the first user touchpoint.
Concerning the latter, once the customer enters the chat and provides a question, the bot has to categorize it and then provide a solution. If it's not the right solution, the customer has to say so and then try to alter the question to get a different response. This process can be wearisome and annoying, as opposed to simply letting customers navigate potential solutions for themselves.
And by going one step further, beyond something as common as "returns," and providing a full database of resources for customers to easily and quickly self-serve for a variety of issues, companies will make the experience better for themselves and their customers and still allow that same potential for cost-savings as a bot. It's the same information that is necessary to help customers, just presented in a way that is more convenient for them to use. Self-service, when executed properly, can have predictive search functionality that requires even less effort from the customer. This is basically like how Google auto-fills in possible inquiries as you are typing.
Customer-facing bots still have a long way to go before they can adequately rival what customers can do for themselves. Maybe bots will be preferable once they can accurately predict what the problem is before customers even have to initiate conversations.
In the meantime, they can play a very important role in the background.
Bots can be used to make workflow management more efficient, for example. When a customer does have a more advanced problem, bots can help the agent in proposing suggested solutions. The agent can then use different pieces of the bot-offered solutions and find something nuanced that truly works for the customer. This way the customer doesn't have to go through trial-and-error with the bot, as the agent can use a combination of individual expertise and a wealth of automatically provided product knowledge to best address the issue.
This level of agent efficiency will also increase over time. Thanks to machine learning, bots will continue to learn from what agents are finding more useful and effective and improve the experience for both the agent and the customer in the process.
This is the best bot for both worlds.
Tracy Oppenheimer is senior content producer for in-app customer service startup, Helpshift. She produces both written and video stories about customer service tech today and where the industry is headed tomorrow.