Two years ago, postage equipment giant Pitney Bowes waded into the virgin waters of the Internet of Things (IoT) and put sensors on some of its metering machines. By pulling diagnostic data and pushing updates, Pitney Bowes planned to make the machines more efficient. It could also sell consumables such as ink whenever sensors showed them running low.

Sounds like everyone wins, right? Not quite.

At GE Minds+Machines in San Francisco last year, a Pitney Bowes executive hinted at internal discord. Salespeople didn’t want to sell IoT-enabled machines to existing customers out of fear of losing future commissions, the executive said. They had forged decade-long customer relationships and counted on replacing worn-out machines every few years, but IoT tends to prolong the product lifecycle.

Related: Zero One: Betting on Big Data

Dan Adams, vice president of data product management at Pitney Bowes, confirmed the issue in an interview this month. Adams says he tries to convince salespeople that customer expectations have changed. Customers today won’t put up with a company that sells products that break down regularly.

In the brave new world of digital, customer experience matters most.

“How do we have additional products and services for the client to buy that are beneficial to them while they’re experiencing the value of the original purchase?” Adams says. “It’s really through innovation, not selling a product because the old one broke.”

Adams is in a difficult spot. He’s in charge of setting the strategy for delivering Pitney Bowes’ high-growth data products, in part, to existing customers. If the customer experience is stuck in the stone age, then the chances of selling new-fangled data products will be slim.

The trouble started when Pitney Bowes built a device, called SmartLink, for connecting metering machines to the cloud. SmartLink monitors an array of features, such as ink and prepaid postage levels. Adams wants to productize data culled from smart metering machines. For instance, Pitney Bowes could aggregate usage data by, say, zip code to unearth macro trends and lookalike customers for small businesses to target.

“More interestingly from my standpoint is looking into the core business of Pitney Bowes itself, at the data assets in the conducting of our business, to see what we might turn into products,” he says.

Related: 2017 Salary Guide: 16 Hottest Digital Transformation Skills

To do this, though, a customer would probably need to be part of the aggregated data set. More importantly, the customer must trust Pitney Bowes to handle data in a way that doesn’t expose the identity of any single business. Trust is critical, especially when it comes to collecting and selling customer data.

Customers will have to take a leap of faith that Pitney Bowes won’t do anything unsavory, such as selling an inferior product just to reap a future commission. Rather, today’s empowered customers will buy data products from Pitney Bowes, Adams says, “because the product they do have is still working.”

Based in Silicon Valley, Tom Kaneshige writes the Zero One blog covering digital transformation, AI, marketing tech and the Internet of Things for line-of-business executives. He is eager to hear how digital transformation is impacting your business. You can reach him at tom.kaneshige@penton.com.