Matt Dixon on why AI should make customer service more effective, not just efficient

Companies have long treated customer service as an efficiency problem: How do we reduce handle time, deflect more calls, and lower the cost of serving customers?

AI makes those things easier. But Matt Dixon sees another opportunity emerging: using AI to make the customer experience better.

Matt Dixon, Founding Partner of DCM Insights and co-author of The Effortless Experience, The Challenger Sale, and The JOLT Effect, has spent years studying what makes customer and buying experiences work. On the latest episode of The Agentic Wave, he talked to Parloa’s CMO, Latané Conant, about how the ideas behind his research apply in today’s AI landscape.

Turn the contact center into a source of customer insight

When Matt and his coauthors developed the Customer Effort Score as part of the research behind The Effortless Experience in 2013, understanding friction depended largely on asking customers about it.

The problem? As Matt puts it, 

“Did it ever occur to you that asking customers to fill out a survey that asks them how hard it was to get their issue resolved after they just had a hard time getting their issue resolved is itself a source of friction and effort?” 

As Latané succinctly describes it: it’s annoying. 

Surveys also suffer from low response rates and tend to capture customers at the extremes: the people who complete surveys are either very happy, or to borrow Matt’s example, they’re the ones who “dropped 15 F-bombs on that call.”

AI offers another way. Instead of asking a small percentage of customers where friction exists, companies can analyze interactions at scale for the language, behaviors, and patterns that indicate high-effort moments with no added customer effort. That gives customer service leaders much stronger evidence to take to the rest of the business. 

Be proactive (but not creepy)

AI also creates new possibilities for reducing effort before customers even have to ask for help. If a flight is canceled, for example, an AI agent could proactively contact the customer and offer to rebook it. Or during a service interaction, an agent could recognize that customers with one problem often encounter another and help prevent the next call before it happens.

There’s a limit, though. Matt describes the principle as “proactive but not too creepy.” Anticipating a genuine need can make an experience easier; unnecessary or intrusive outreach can rub customers the wrong way.

And that principle extends beyond proactive service. While the original Customer Effort Score research focused on resolving customer problems, Matt points out that people value ease throughout the customer journey: products that are easy to use, pricing that's easy to understand, buying experiences that are easy to navigate, and service that gets them back on track quickly.

Effortless experience goes beyond customer service: it should guide the entire customer journey.

AI buyers don’t need more FOMO. They need less FOMU.

Organizations know they should invest in AI, and instilling more fear-of-missing-out (FOMO) in them won’t make them buy faster. Instead, guidance on where to start and how to avoid getting it wrong, will.

Matt’s research for The JOLT Effect found that customers often fail to make a purchase not because they prefer the status quo, but because they’re afraid of making the wrong decision. Matt calls that FOMU: fear of messing up.

And AI may be the perfect environment for it, with buyers facing an enormous number of use cases, vendors, configurations, and implementation decisions. They’re also trying to keep up with an overwhelming amount of information while hearing stories about failed pilots and AI investments that didn’t deliver.

Matt groups those sources of indecision into three categories: choice overload, information overload, and expectations overload, or uncertainty about the outcome. What buyers need now is confidence that they’re making the right decision.

Partners should narrow the choices

For Matt, overcoming FOMU means finding partners willing to narrow the options rather than simply presenting everything AI can do.

That might mean recommending a smaller, achievable use case instead of immediately pursuing the most ambitious vision. As Latané explains,

"[Companies] come to us and they're so excited about what's possible...then they come up with the hardest use case ever that no one at the company has ever been able to do."

She urges businesses to start with a simpler approach.

Matt agrees: get “some runs on the board,” then learn and grow together.

Strong partners should also help buyers navigate information overload by curating what matters, being candid about where their solution is and isn’t the right fit, and providing a clear path from purchase to value.

Smoke or Fire: Does customer service really drive loyalty?

Matt’s verdict: Fire…but perhaps not in the way most companies think.

His research found that a customer service interaction is four times more likely to create disloyalty than loyalty. Long waits, repeated information, unresolved issues, and rigid policies can quickly erode the goodwill a company has built.

That’s why Matt recommends thinking about loyalty as both offense and defense. Play offense with great products, pricing, branding, and buying experiences. When something goes wrong, play defense: make it easier than customers expect to resolve the problem.

Make effortless experiences the expectation

Matt notes that “being easy to do business with is ironically very hard.” But AI is bringing down the cost of ease to the point that companies have fewer excuses not to deliver it.

The opportunity isn’t simply to use AI to handle more interactions for less money. It’s to understand customers better, remove friction earlier, and make doing business with your company dramatically easier.

Watch the full episode of The Agentic Wave to hear Matt and Latané discuss customer effort in the age of AI, how to overcome FOMU when making AI investments, and why effortless experiences are finally possible at scale.