That was the takeaway from a panel at Fortune’s Brainstorm Tech conference in Park City, Utah, on Tuesday. Speakers talked about the modest progress in data crunching and analysis, but added there’s much more to be done.

Real estate listing site Zillow has grown more attuned to rising consumer expectations, said Jeremy Wacksman, the company’s chief operating officer, who spoke on the panel. For example, just a few years ago, the company routinely let users post photos of homes for sale or rent. But with the rise of generative AI, it’s now able to provide users with more immersive 3D walkthroughs of listings, Wacksman said. The result is “magical,” he said.

Another company, business software maker Salesforce, has used machine learning and AI for years, said Susan Emerson, senior vice president of go-to market strategy for AI products. A decade ago, the company could understand some rudimentary details about how customers used its products: Did they buy licenses? Did they ultimately deploy the software? Today, Salesforce can collect far more kinds of customer data, with more velocity and specificity, to help drive more adoption and “unlock trapped value,” Emerson said.

Panelists also discussed their internal efforts to get employees to use Microsoft Copilot, an AI assistant that helps users create content, automate repetitive tasks, and produce code while using Microsoft apps. The idea is that the tool makes workers more productive by having AI quickly handle some of their chores. “We have a record of engagement in Salesforce. [Copilot is] not optional. It’s on the screen,” Emerson said, referring to how her company encourages workers to use the technology and tracks that use.  

Zillow’s Wacksman said his company is also pushing the use of Microsoft Copilot and making it easier for employees to find those tools that they use more often. “We encourage people to get their hands dirty,” he said. 

Companies collecting data about customers to better personalize products for them raises questions about privacy. It’s a moving bar based on customer expectations that change over time, Wacksman said. He pointed to Uber and Airbnb, two now popular services that premiered in the late 2000s. At the time, “We all thought how creepy it was,” he said, but now users accept without much thought. For consumers, the tradeoff is between value and privacy, said Raj Seshadri, chief commercial payments officer with Mastercard. Consumers own their data and decide who to share it with, while companies must be transparent about how they use it. But consumers have to get some value from the company otherwise “they won’t share it,” she said.