P.V. Kannan, Co-founder and CEO
Over the past two decades, large companies that sell directly to consumers have spent billions of dollars annually on sales, customer service and support. Surprisingly, though, I've seen that many of those brands fail to wow their larger customer base. But with all this investment, shouldn’t consumers be thrilled by their experiences with these brands? I call this “the billion-dollar mystery.”
There are two main reasons consumers reach out to brands: The first is to get information (e.g., “When will I get my order?” or “What is your order cancellation policy?”); the second is to complete a task (e.g., “I want to upgrade my phone plan,” or “I need to return an item"). Consumers expect both of these things to be smooth and easy, like ordering a Lyft or an Uber. When using those services, it only takes them a couple of clicks to get what they need.
Conversely, when interacting with a bank, phone company or utility, for example, they are confronted with a confounding array of channels to choose from, most of which cannot resolve even the most basic queries.
As the founder and CEO of a software and services company that focuses on improving customer experience through the use of artificial intelligence, I have worked with large companies in every industry. I have also explored how AI, when combined with human intelligence, can impact the customer experience. Below are four things to keep in mind when investing in new channels:
In the past few years, consumers have embraced new channels, such as messaging, as a convenient way to interact with one another. As such, brands have added those channels as new ways to interact with a company. However, unless those new channels can result in resolving the consumer’s issue, they are causing damage to a company’s brand reputation.
For example, if a messaging channel can only resolve the consumer’s intent 40% of the time, that means 60% of the time the consumer is forced to go to another channel for a resolution. This happens when a channel is set up without thoroughly understanding consumer intents. To better understand consumer intent, you can analyze large numbers of call and chat transcripts so you can build out conversational models. Because of the volume of data, however, many companies opt to use AI to help streamline these processes. (That's the idea behind my company's model.)
The biggest challenge in implementing AI is understanding the return on investment. It is important to analyze where automation can immediately save money or generate revenue. If the ROI cannot be immediately demonstrated, then it doesn’t make sense to deploy AI in that way.
As your company builds solutions for each channel, you must also consider how much time your consumers have spent on the previous channels. For example, if a consumer has already spent 10 minutes trying to resolve an issue on the web, by the time it escalates, the company’s net promoter score has gone down ten points because of the wasted time.
A couple of years ago, during the height of chatbot hype, several companies rushed to create bots that did not have the ability to escalate to a live human agent. Essentially, these were just the chatbot version of frequently asked questions. If a consumer asked a question the right way, they would get the answer they were looking for, but if they asked it in a different way, they might “break the bot.”
In those cases, companies must have a way for a human to take over the conversation. Companies that choose to use chatbots should start by automating the most common journeys so that the consumer can accomplish them without having to rely on a human. For less common journeys, the conversation should be routed straight to a human, with the goal of automating over time as the system gets smarter.
Having channels that can only do one thing is fine, as long as consumers know that. Consumers need to know which channel will get them the result they are looking for, and if a channel can only handle a limited set of intents, that should be stated upfront. For example, everyone knows that 911 serves one purpose: to reach someone in the event of an emergency.
Smart speakers, such as Google Home and Amazon Alexa, are great at answering basic questions and providing information, but they can’t conduct transactions. Consumers seem to be generally fine with that because they know what those devices are for. I suggest that companies brand each channel so that it is clear which intents it can handle — and which it can’t — so that consumers don’t have the wrong expectations.
Companies can also consider using tools that help them visualize customer journeys across multiple channels, such as phone, email, text and the web. That will help them understand where their customers are getting resolutions and where they are hitting obstacles. Using customer journey analytics, companies can look at the “last mile” of the consumer’s journey, including listening to calls or looking at chat transcripts. Using these tools, companies can understand what leaves customers with a negative sentiment toward the brand and is ultimately leading to churn.
Trying to analyze customer journeys without having tools is like trying to dig a trench with your bare hands. It can be done, but it takes several times longer. The tool, however, is only as good as the human analyst who is using it. Companies need an analyst who understands customer journeys and can ask the right questions to help them gain valuable insights.
Gaining insight into customer journeys can be challenging, but there are a few steps you can take to see where breakdowns are taking place to ensure you're providing the most positive customer service experience possible.