Data transformation company dbt Labs has used AI to drive a variety of measurable impacts. The engineering team shipped a feature in 24 hours using an end-to-end AI development loop. One employee used Claude to build a script that assembles and creates onboarding materials, saving over 20 minutes per new customer. Another created a daily AI summary of the company’s more than 30 Slack channels, replacing a half-hour of employees’ manual scanning with a quick read.
These types of efficiency numbers are real and easy to regard as progress, but CEO and founder Tristan Handy believes many leaders are thinking about them all wrong. Inside his organization, he’s found that the biggest benefit from AI isn’t captured in efficiency measures or time savings, but rather in the creativity and iterative power they unlock.
As a growth-stage startup company whose technology powers the very AI transformation companies are seeking, the opportunity for dbt Labs as Handy sees it isn’t to use AI to reduce costs or slim down—it’s to seize market opportunities. AI efficiencies are the vehicle enabling the organization to take more shots, solve more problems for customers, and use those learnings to improve quality, not just increase quantity.
“In a business context, you’re not creative because you want to sell novels,” said Handy. “You’re creative because you want to come up with better ideas to drive the metrics that you’re accountable to.”
Top line versus bottom line
The pressure on business leaders to demonstrate measurable success and ROI on AI initiatives continues to grow. In a recent survey of 660 technology executives conducted by Deloitte, 79% said driving measurable business outcomes is a top priority in 2026. There are two ways to go about it, Handy noted.
“I think it’s about, are you trying to measure the upside that you get from AI as a cost-cutting measure, or as a revenue enhancing measure?” he said.
The companies finding the most success are doing both. A study by McKinsey that analyzed almost 2,000 companies’ AI transformation efforts found that while 80% of respondents set efficiency as their objective for AI, it was those that also set goals around growth and innovation that have seen the most value from the technology. Specifically, the report notes organizations with growth-led AI goals were more likely to report achieving a range of qualitative enterprise-level benefits from their AI efforts.
The right way to measure success for these different facets of AI transformation goals is not one and the same, however.
“The first question you should be asking for every single AI initiative is: Do you want this to, at the end of the day, grow top line or bottom line?” said Handy.
Efficiency as a creative catalyst
Measuring cost-cutting AI efforts is fairly straightforward, which Handy believes is why many leaders are focused on it. Driving gains on the revenue side, however, requires creativity.
On the revenue side, if the question is how to refactor the work to make lines go up and to the right, “that actually does require creative thinking,” Handy said. “It requires new products. It requires operating in different creative ways.”
For example, in a recent meeting where dbt Lab executives were discussing the company’s data lake product, Handy asked if they knew the reasons why the company’s top 20 customers that use the product chose it. Nobody knew, so they prompted an AI agent that plugs into the company’s internal data—including all the quantitative data and sales call transcripts—to look into it right then as they continued the meeting. The agent went off and did the research, and came back a few minutes later with all the reasons and relevant information. Handy doesn’t see the time saved on answering the question as the win. The real benefit, he says, is that the AI allowed the group to continue pursuing a path that the conversation had suddenly taken.
“In the room, that conversation now becomes more creative, because we’re not stuck putting that in the bucket of things to follow up on later,” he said. “You can actually fully explore the idea and get into it without being blocked.”
Put another way, shrinking the time it takes to accomplish tasks allows companies to get more work in front of customers, which allows them to learn more in the process. Handy, however, emphasizes that it shouldn’t be about simply increasing output. Inside dbt Labs, he is aiming for a steady increase of quantity and quality together, rather than a burst of production.
“You have to think of efficiency and creativity as the same thing,” he said.
