I once became so frustrated trying to get help with a hotel booking through Priceline that I cancelled my card. I couldn't get anyone to respond, and I felt stuck in a support loop with no useful way out.
I don't know what technology was behind the interaction. What stayed with me was the frustration of needing someone to help and being unable to reach them.
As a business owner, I think that experience raises a question we need to answer before we automate more of the customer relationship. Where do we want technology to help, and where do we want a person to remain directly involved?
For me, critical decisions need human judgment. And when it comes to the customer experience, I want people responsible for the relationship. There is real value in someone understanding what the customer needs and taking responsibility for what happens next.
---
Let the customer choose a person
My preferred handoff rule is simple: as soon as they ask.
If a customer wants to speak to a person, I don't think they should have to complete more automated troubleshooting to earn that conversation. They have already told us what they need next. We should make that route clear and start the handoff.
That doesn't mean every business can have someone instantly available at every hour. It means being honest about the wait or callback, keeping the request with a responsible person, and following through. A promise to connect someone needs an owner just like any other customer commitment.
Research gives us a reason to take that choice seriously. In a survey of 3,566 business and consumer customers conducted in February and March 2026, Gartner found that 87% considered access to a person essential when companies use generative AI in customer service. Gartner's September 2026 survey findings.
There is also room to recognize where customers appreciate automation. A Capgemini survey of 9,500 consumers, conducted in late 2024, found that respondents favored people for empathy while slightly more than half favored chatbots or virtual agents for speed and convenience. Those reported preferences challenge the idea that everyone wants a person for every interaction. Capgemini Research Institute, 2025 report, Figure 24.
I can appreciate that convenience and still want the customer relationship held by people. The part I would protect is access to someone who can listen, use judgment and take responsibility when the customer needs them.
---
Give the person better information
The AI opportunity that interests me most is what happens internally, from marketing to sales and from sales to onboarding and operations. Information can get dropped at those transitions. I think AI can help carry it forward so the next team member is better prepared to serve the customer.
There is a useful example of AI supporting people in a study published in the Quarterly Journal of Economics. Researchers examined an AI assistant used by 5,172 support workers at one business-software company. Employees remained responsible for the conversation and could edit or ignore its suggestions. The study found productivity gains, particularly for less experienced workers, with results varying across employees. It did not find an economically significant average change in satisfaction measured by net promoter scores. Brynjolfsson, Li and Raymond, “Generative AI at Work”.
What I find useful is the arrangement: technology helps prepare the person who is serving the customer. The person still has a job to do, including deciding whether the suggested response fits.
For our own handoffs, there are three things I want the next person to understand:
- What was promised. What did we actually commit to, and where is that commitment recorded?
- What the customer is trying to accomplish. What matters to them beyond completing our process?
- What they've already told us. What should the next person know before starting another conversation?
I don't want customers having to repeat themselves every time they move to the next person. Those three questions give the team a practical place to start.
---
Make the handoff useful to the customer
Imagine a service company bringing on a new client. This is a hypothetical example. During sales, the client explains a deadline, describes a previous problem and asks for a particular outcome. The salesperson records the discussion, and onboarding receives the account.
An AI assistant could help organize those notes around the promise, the goal and the information already shared. The receiving employee would then check the summary against the original conversation and agreed terms. A customer's requested date needs to remain distinct from a date the company actually promised.
That review matters. A polished summary isn't a reason to assume every detail is correct. Keep the source available, identify unanswered questions and give the next person enough authority to resolve the ordinary ones. When a consequential decision is needed, make clear who can make it.
The next conversation can then begin with what the team already understands and what still needs confirming. That is the experience I want to build: a person who arrives prepared, respects what the customer has already shared and owns the next step. The AI supports that preparation behind the scenes.
The team should also check whether the handoff is useful in practice. Ask the receiving employee what was missing and whether the customer had to explain it again. That gives you something specific to improve.
---
Start with one real handoff
Choose one recent sales-to-onboarding handoff and review it with the person who received the customer. Can they find what was promised, explain the customer's goal and see what the customer already told the team? Identify one missing piece and correct how it gets passed along.
The answer shouldn't depend on the founder remembering the whole conversation. Give the receiving person the context and decision boundaries they need, so a human customer experience can continue without every question coming back to the owner.
That's where I see the value in AI: helping our people show up better prepared while we preserve human access and responsibility. When the next customer reaches your team, will they feel that someone already understands what matters to them?