Leadership at Lithuania-based marketing automation firm Omnisend tried many of the usual tactics to get employees to use AI. They created games and videos to help the team get more familiar with AI tools and capabilities. They also hosted monthly “AI days,” asking employees to clear their calendars to spend time experimenting with the technology.

“It worked for a while until we realized we’re hitting a wall, because now we have activity and productivity, but we don’t necessarily have impact,” Bernard Meyer, head of AI operations at Omnisend, told Fortune.

To start moving toward the measurable business impact they felt would be necessary to truly transform the company for the AI era, they added a new initiative: offering quarterly salary increases for employees who use AI in ways that create measurable business impact. They’re not looking for causal usage of AI tools, but specifically for employees to discover and use AI in ways that move the needle for the company at large. For Q1, the first quarter with the initiative in place, just over half of the company’s 246 employees qualified and received raises between 2% and 4% so far.

Speaking with Fortune, Omnisend cofounder and CEO Rytis Lauris expressed wanting “to pay higher salaries for highly effective people.” He also emphasized his belief that the labor market will reward those who learn how to use AI well—at the expense of those who don’t.

“We think that AI is not threatening your job, but at the same time, we very loudly, vocally said that you will lose your jobs, the jobs as they are today,” said Lauris, describing the general shift he anticipates in the job market. “So it’s either you as individuals transform for the future needs, or you will lose your jobs.”

Evaluating business impact

Omnisend has historically offered salary increases on a quarterly basis, so this cadence isn’t new. But this new approach—an additional way to earn raises on top of the typical review process—required leadership to determine specifically how to correlate AI usage to business outcomes.

They decided that to qualify, employees would have to show measurable value in at least one of three areas: efficiency (meaningful time or cost savings); impact (KPI movement, improved quality, better business outcomes); or scalable adoption (workflows adopted by teammates or across teams).

In one example of work that earned the compensation bump for Q1, the finance team automated a major part of invoice processing, surpassing the original goal to automate 36% of the process and ultimately reached up to 85% automation. Since the workflow was built and implemented through a shared team effort, all members of the finance team involved received the pay increase. Additionally, members of the legal team earned the raise for using AI for vendor assessment and legal intelligence workflows, reducing manual review time. Several engineers also received raises for finding ways to use AI to improve workflows that were adopted across teams.

Lauris admits it’s “honestly really difficult” to measure the business impact of AI adoption, and it’s more challenging for some job functions than others. For product managers, for example, their workflows aren’t consistent, and while they create a road map, it’s really the engineers who execute it. They had to agree on an “objective subjective measurement” to evaluate product managers on their impact.

Each manager is responsible for conducting the reviews for their direct reports, determining who meets the goal for the salary increase, and what percentage raise to give them. AI is not the sole factor informing compensation; employees still go through the regular salary review process as usual, and the AI salary bump does not affect any other part of that evaluation. But demonstrating AI impact does give employees a chance to earn more than they would otherwise, and Lauris said managers should use the AI-based raises as a tool to push AI adoption.

“We’re trying to motivate people on many levels, and one of those is money. It is also recognition as an AI champion within the company,” said Meyer. “They get the chance to share their knowledge across the organization, to be seen as AI enthusiasts.”

Comparison culture

Overall, Meyer said he sees the AI-based quarterly raises as “a bar raiser.”

“You are looking around at the team, and you see that of six people on the team, three of them got it. And then next quarter, an additional one or two people get it,” he said, offering a hypothetical of how an employee would compare themselves to their peers.

While Lauris said they’re not publicly announcing who gets the raises across the company, they also “don’t restrict sharing it within teams.” He and Meyer imagine—and seem to be counting on—teammates talking about it among themselves. Individual managers can also speak openly about who’s earning AI-based raises and who isn’t within their teams.

“[Managers] can use it as a tool and publish it within the team, like, ‘Look, these colleagues got salary raises, and you did not, so would you like to catch up?’” said Lauris.

In a larger sense, the initiative is putting into action what Omnisend leadership believes will dictate not only the future of the company, but also what’s expected from workers in the new era of AI-led business. The company needs to transform, Lauris said, and leadership is also willing to help people transform on the individual level.

“It’s a very, very clear message strengthening the main message that some people’s salaries are already increasing more than others, and it’s absolutely on you if you want to earn well in the upcoming five or 10 years or you want to stay where you are,” he said.