Tips for Driving Customers to Self-Serve
A single customer self-service interaction through a chatbot or FAQ portal can cost as little as 10 cents, while a live agent interaction can cost up to $8 or more. So companies have a clear financial incentive to push customers to self-service channels whenever possible, but the benefits go far beyond a few lines in the ledger.
“Promoting self-service is not just about reducing costs; it’s about empowering customers to use the easiest and most efficient solution,” says Keith McIntosh, senior principal of research in the Gartner Customer Service & Support Practice.
“Businesses should implement self-service with the strategic intent of not only deflecting phone calls but also delivering delightful self-service experiences for customers and building the brand,” adds Anand Subramaniam, senior vice president of global marketing at eGain. “Poor self-service, purely focused on reducing live customer contact, can turn off customers and crater the brand.”
Across the customer service world, the advantages of self-service for companies and customers are clear. But the best way to drive customers to self-service is a lot murkier.
For some, the thought was that the best path toward customer self-service adoption was to make it the only option, but that might not be the best course.
It was a move that Frontier Airlines took in November 2022 when it eliminated live customer support entirely. The low-cost airline instead directed customers to interact with it through online channels, including a chatbot and social media. And while Frontier claimed positive results with high customer satisfaction for the chat interactions, some customers told a different story, especially when dealing with complex issues like cancellations. Others complained that the lack of human interaction was frustrating or that the system was difficult to navigate.
Frontier has since gone back on its digital-only mandate, bringing back live phone support for customers within 24 hours of their flights and for elite frequent flyer members. It is also introducing a callback service for other inquiries.
Frontier is not the only company that has promoted self-service in this way. In fact, recent research from CCW Digital found that as many as 91 percent of consumers have felt that companies are forcing them to use AI-based self-service, whether they want to or not.
Some companies have tried to push customers to self-service by making the live agent experience as painful or inconvenient as possible. Customers looking to reach live agents encounter convoluted or unhelpful automation tools, long wait times, repetitive questioning, and worse.
Sure, customers want the accuracy and speed that automation offers, but they don’t want to be forced into a self-service system that doesn’t meet their needs, says Jason Valdina, senior director of go-to-market strategy for engagement channels at Verint.
At the same time, though, a wide berth of research has found that customers would sooner navigate a good self-service system rather than waiting an inordinate amount of time to speak to a live agent. But the system has to be good.
“If you haven’t built your self-service well, any push to it is too strong,” Valdina stresses. “A good design allows you to be a little more pushy with self-service.”
Yvonne Daugherty, global head of industries at Ushur.ai, a provider of customer experience automation for regulated industries, adds that when people choose not to use self-service, it’s usually because the experience is unhelpful, inaccurate, or incomplete and doesn’t support their immediate needs. They try, get stuck, then require a live agent chat or call.
If, for example, a customer asks a common question that the automated system doesn’t understand, it indicates that the natural language understanding is flawed, the chatbot knowledge base is inadequate, or a little bit of both, Valdina says, noting that problems like these do more harm than good when it comes to self-service promotion.
So how should companies promote their self-service channels? It depends.
How to Advance Automation
“There are a lot of different tactics that work very well to drive people to self-service,” Valdina says. “One that is often overlooked is promoting self-service at the point at which people try to contact you.”
In many cases, that first point of contact might very well be a contact center agent, but very few agents are doing a good job of promoting self-service. A recent study by Gartner found that 60 percent of customer service agents fail to promote self-service options.
But it goes beyond that. Despite the potential benefits of self-service, when agents do mention self-service in customer interactions, 25 percent make neutral comments and 12 percent make explicitly negative remarks, often to the detriment of their operations.
“Agents play a crucial role in this process, and their ability to positively endorse self-service options can really matter,” McIntosh says.
Further highlighting this point, Gartner’s research found that when agents promote self-service, twice as many customers say they will adopt self-service when they encounter future service issues with the company.
Companies, therefore, should focus on training agents to effectively communicate the benefits of self-service options to customers, Gartner suggests.
eGain’s Subramaniam also recommends offering a co-browsing feature that allows agents to show customers how to use self-service for future interactions.
There’s another factor that is also working against self-service, according to Gartner’s research: As leaders continue to invest heavily in generative artificial intelligence to automate self-service, great phone experiences can actually disincentivize customer self-service. While 55 percent of service leaders are exploring customer-facing genAI chatbots, only 35 percent of customers who last interacted via phone are willing to use genAI digital assistants. This reluctance stems from the success and comfort customers experience when resolving issues directly through phone interactions. After all, why would they shift to a genAI digital assistant when their current phone method is effective?
The Role of GenAI
“Service leaders should focus on integrating genAI solutions that complement existing phone interactions rather than replacing them. By positioning genAI as an enhancement to the phone experience, organizations can reassure customers that the digital assistant is designed to streamline their journey, offering both self-service solutions and seamless transitions to human agents when needed,” McIntosh said. “Meeting customers where they are while offering innovative solutions is key to driving adoption and satisfaction.”
That also means making self-service available as early as possible in other interactions. Most companies make it easy to find self-service options on their websites or lead with the self-service option via a chatbot online or for call-ins, Valdina adds. Gartner estimates that more than 85 percent of customer service interactions start with self-service.
“If you start with some service up front, you can provide very fast and accurate service for your customers,” says Dimitris Vassos, founder and CEO of Omilia, a customer interaction and conversational AI solutions provider. For example, a bank or credit card company can offer balance information at the start of a call, with a quick opt-out for customers who are contacting the business for another reason.
“This is proactive, it’s very accurate, and it will satisfy a lot of the callers up front,” Vassos says. “You want to be proactive and personalize self-service as much as you can.”
Similarly, the more depth self-service can provide up front, the more companies can push that option.
If, for example, a customer calls in for help with information on the company’s website, a robust self-service engine can ask the caller where he is on the site and direct him to the correct area, Valdina explains.
Companies can also use QR codes to facilitate self-service for mobile phone users when they are at physical points of contact, such as a store. QR coders can also be used at a train station, for example, so commuters can purchase tickets or find schedules on their own without involving human agents.
“You could drive [customers] to a web chat, messaging channel, or social media. It’s a great way to put your best foot forward for a self-service option,” Valdina says.
New York’s Metropolitan Transportation Authority added such an option a few years ago, providing commuters with quick access to electronic tickets, scheduling, etc., while reducing the need for agents to handle these types of calls.
Regardless of how they get there, though, customers will only use self-service initially and consider it as a preferred channel for subsequent interactions if it is efficient and customer-friendly, experts agree.
To provide efficient, customer-friendly self-service, companies need to use a customer experience automation platform that can leverage key contextual data at any point in the customer’s journey to personalize responses, Valdina says. Any systems they use, he explains, should be able to quickly identify customers based on their phone numbers and other identifying information (if needed).
Automate with Empathy
Though self-service by its very nature is automated, infusing responses with empathy can help with customer acceptance and use, according to Ushur.ai’s Daugherty. “Responding with intelligence and empathy matters. If self-service can recognize what the customer is asking, remember context, and respond clearly and empathetically, that’s a game changer. That feels less like a machine and more like a helpful interaction. Accessibility for different languages and abilities matters, too.”
The ability to scale, especially when self-service needs spike suddenly (like when an airline has to cancel a large number of flights due to an unexpected storm), is important as well, which is where many self-built self-service systems fail, according to Vassos.
“Having static, out-of-date self-service that has inflexible workflows and has to follow an exact order doesn’t help,” he states.
Many of those older systems aren’t internet-based and can’t produce answers on the fly, Valdina says. “They can only do what they are trained to do. If the customer needs to do something else, they can’t do it and the customer gets very frustrated.”
For that reason, Daugherty also recommends using systems that have a very specific form of generative AI. “There is a difference between ChatGPT, which is built on all kinds of data sources, and intelligent generative AI that is made to support specific industries,” she says.
AI agents should also be built on very specific industry knowledge bases that understand the unique terminologies used in specific industries, experts recommend.
Modern self-service solutions also need modern technology stacks that include retrieval-augmented generation (RAG) to locate and deliver accurate, up-to-date, and relevant responses, Valdina says.
“Companies that are ahead of the curve in terms of self-service are already leveraging the content that they have,” he continues. “They might have a video and a script along with it for people who are hearing-impaired, or they have other content sources that they can feed into virtual assistants to drive self-service. RAG makes it easier to fine-tune and enhance self-service.”
Valdina adds that the self-service efforts of many companies fall short due to substandard or nonexistent RAG capabilities. “They build a bot, they put self-service on different channels, but [the technology] is too cumbersome and doesn’t allow them to iterate. They can’t enhance it.”
Companies in some industries, like healthcare or financial services, might also need to authenticate customers before allowing them to access information or perform certain tasks through self-service. When such measures are needed, they should be simple and straightforward.
And while the consensus is that self-service should be as robust as possible, enabling customers to conduct a wide variety of transactions and access a large library of information to minimize the number of interactions that eventually go to a human agent, it’s important to not go overboard early on. Daugherty says it’s better to start with a self-service system with minimum capabilities and then add depth later. Companies could miss out on cost savings by waiting too long for a much more robust system, she says.
Automate Intelligenty
While it’s usually in the company’s and in the customer’s interest to use self-service, there will inevitably be calls that need to be escalated to live agents for one reason or another. When that happens, the underlying technology should make such escalations fast and simple, experts agree.
“That would seem to be table stakes, but there are many brands that are missing out on that,” Valdina says. “The handoff to a human needs to include the information that has previously been communicated so the customer doesn’t have to repeat everything. Even though the basic technology to do that—the screen pop—has been around for decades, some systems haven’t incorporated newer channels, so the repeated details are necessary when moving from one system to another.”
And when a customer asks to escalate to a human agent, companies still have an opportunity to handle the interaction via self-service, according to Subramaniam. “When customers are on hold for human-assisted service, organizations can send a contextual self-service link to the customer where they might find the answer, while assuring them they won’t lose their place on the call (or live chat) queue if they decide to try out self-service. Since most calls today originate from smartphones, this approach is an effective way to gently nudge customers to digital self-service.”
But in the end, one caveat remains true: “Your self-service better be contextual, correct, and consumable for this approach to work,” Subramaniam states emphatically.
Phillip Britt is a freelance writer based in the Chicago area. He can be reached at spenterprises1@comcast.net.
The Chatbot Paradox Still Lingers
While 90 percent of companies today are investing in artificial intelligence-driven customer experience solutions and 87 percent are increasing their self-service investments, customers are increasingly frustrated by chatbot loops, impersonal interactions, and the lack of human fallback, CMP Research found in a recent customer survey.
The report spotlights what CMP calls “the chatbox paradox” in which more automation doesn’t always equal better customer experience.
Additional insights from CMP’s data include the following:
- 41 percent of CX professionals cite conversational interactive voice response systems/voicebots as a top tech priority in the next two years.
- New advancements, including empathetic voicebots, voice cloning, and seamless multimodal switching, are reshaping what smart self-service looks like.
- Poor implementation and cost-first thinking continue to erode customer trust in self-service automation.
According to CMP, quite a few companies are now rethinking their self-service strategies, asking not just what AI can do, but how it will impact customer trust, loyalty, and retention.