The Customer Journey is More Than a Single Conversation
Contributed by Sierra
For years, the goal in customer service was reasonably straightforward: answer quickly, resolve the issue, and leave the customer satisfied. AI is making that easier. At Chime, for example, AI agents have increased resolution rates from 50% to more than 70%, helping members get fast answers around the clock.
But customers rarely accomplish something meaningful in a single conversation. Buying a car takes time. Finding a home can take months. Getting a mortgage involves documents, decisions, and follow-up. Filing an insurance claim can stretch across days or weeks. Customers do not think of those as separate interactions. They are simply trying to get something done.
That is where the next generation of AI agents gets more interesting. Instead of treating every conversation as a new event, an agent can retain context, take action, and keep working toward an outcome over time. It can answer the question in front of the customer while also making progress on the bigger job: finding the home, finishing the application, booking the appointment, or resolving the claim.
Customers are trying to do something
Because many customer experience processes and platforms have been built around discrete interactions, that is also how the industry has been conditioned to measure success: handle times, response times, and resolution rates. Those measures matter, but they don’t necessarily tell you whether the customer actually got where they were trying to go.
Consider someone shopping for a car. They may browse inventory online, call to see whether a vehicle is available, schedule a test drive, change the appointment, ask about financing, and eventually visit a store. The customer’s goal is not to have six successfully resolved interactions. They want to find the right car.
At CarMax, AI voice agents handle common questions, including store hours and vehicle availability, and help get customers to the right associate when a person is needed. The company is also expanding the experience toward actions like scheduling appraisals, browsing appointments, and test drives. The agent handles the initial question while also keeping the buying process moving.
The real measure is what happens next
The growing capabilities of AI agents also suggest a different way to think about success. Resolution still matters, but so does what flows from it.
At Redfin, customers can search for homes conversationally instead of translating what they want into a rigid set of filters. They can say they want natural light, room for a dog or a neighborhood they can walk around, then refine the search from there. Customers who use the experience view nearly twice as many listings and are 47% more likely to request a tour.
That creates a better customer experience and a better business outcome. Customers can express what they actually want, get more relevant results and move more naturally from browsing toward buying.
Similarly, Rocket Mortgage’s AI-powered Digital Assistant does more than answer questions about mortgages. It can collect information, pull credit, present personalized rates and help customers move toward preapproval. Customers who start with the agent close at three times the rate of those who do not. Customers who use both the agent and a banker convert at four times the rate.
That begins to make the case for AI as more than a cost saver. At each of these companies, the agent is doing more of the work around the human relationship so the customer keeps moving, while people spend more time on the parts of the process where judgment and expertise matter most.
The work does not end when the conversation does
Companies are generally good at helping customers while they are actively engaged, but much less consistent in the days and weeks in between. Many of the outcomes that matter take time. A customer gets an insurance quote and wants to think about it. A patient gets a referral but has not scheduled an appointment. A homebuyer sees a few interesting listings but is not ready to tour them yet.
A good employee knows how to handle that gap. They remember where the customer left off, follow up, answer the next question without making the customer start over, and notice when something has stalled. The challenge is not knowing this kind of attention matters. It is providing it to everyone.
Long-horizon agents make that possible in a new way. An agent can keep working toward an outcome after the first conversation ends, retaining context, moving across channels and picking up where the customer left off.
In insurance, that could mean staying with a prospective customer from quote through purchase, or helping a policyholder through First Notice of Loss and the many steps that follow. In healthcare, it could be helping a patient move from referral to scheduled care, including paperwork, insurance questions, and follow-up along the way. For a subscription business, it could look like understanding why a customer is considering leaving and working to resolve the issue over time rather than waiting until they reach the cancellation page.
Choosing an agent for the experience you actually want
No one is looking to integrate AI for AI’s sake. CX leaders are looking for an experience customers can trust and one that meets the same standards they expect from their own teams.
I’m not willing to have an AI agent that can’t perform at the level that our people can.
Jonathan Marek
COO, Guild
When education benefits company Guild created its evaluation rubric, it looked at accuracy, but also at judgment, tone, and how well an agent knew when to hand a member to a person, according to Danielle Mendel, Guild’s Director of Operations Excellence. With the company’s member support specialists achieving an exceptional 4.8 out of 5 CSAT score, COO Jonathan Marek was adamant that any AI solution would need to match—not merely approach—that standard. “I’m not willing to have an AI agent that can’t perform at the level that our people can,” he explained.
She recalled a member who was waiting on an enrollment confirmation and was clearly anxious. The Sierra agent gave the right answer, but it also handled the situation the way Guild hoped a strong support specialist would: with empathy, context, and a useful next step.
Your CX team becomes more important, not less
The introduction of long-horizon agents is naturally also changing the role of the CX team. The knowledge that used to live only in individual support interactions — where customers get confused, which situations require more empathy, where policies create friction, when a person should step in — can now shape the agent itself.
Instead of answering the same question hundreds of times, CX teams can shape how that experience works at scale. Operational ownership of its AI agents sits with the operations excellence team, which handles much of the day-to-day journey building and no-code updates. More technical colleagues step in where code is required. They can see where an agent is struggling, improve the journey, and decide where human involvement matters most. The people closest to customers become central to how the system itself gets better.
From cost savings to growth
For years, CX leaders have been asked to improve the customer experience while also taking cost out of the system. AI agents make it possible to do both, while also helping more customers get where they were trying to go in the first place.
The first wave of AI in customer experience was largely framed around efficiency: answer more questions, reduce handle time, and lower support costs. The companies that get this right will not simply answer more questions with AI, but will use agents to help more customers by acting on every signal and deciding what’s next, across days or months — while building teams that are increasingly responsible for designing how that happens.
If an agent can help more people finish an application, request a tour, complete a purchase, renew a service, or book an appointment, its value goes beyond reactive customer support. It shows up in better customer experiences and stronger business outcomes.