A hiring manager with a regional bank has beenwith a regional bank has been searching fruitlessly for a software developer to create a better digital customer experience. Yet no one among the hundreds of applicants checks off more than a few of the items on the long list of required technical skills, including knowledge of the obscure programming language the bank uses. A hospital system needs to build a network that will seamlessly link patient records across its locations. When its human resources department finally identifies a seasoned engineer who has done similar work, he turns the offer down and takes a more senior position with a cloud-services firm.

With every company needing to harness the full power of technology to remain competitive, there is now a perpetual stampede to hire tech talent. Demand is growing exponentially for skills such as software engineering, , platform design, analytics-based automation, customer experience design, and . Eighty-seven percent of global senior executives surveyed by McKinsey said their companies were —and that was before the pandemic caused dramatic shifts toward remote work and e-commerce. The pressure is particularly acute for employers outside the tech sector.

One frequently repeated solution to the shortage of tech talent is for companies to hire candidates with more unconventional backgrounds. That sounds logical in theory, but it’s hard to put into practice. Hiring managers are skittish about choosing people with learning curves to fill mission-critical roles. It’s human nature to hold out for someone who feels like a safe choice because they already perform exactly the tasks you need.

Recent research from MGI and McKinsey’s People & Organizational Performance Practice offers some reassurance that could make it easier for companies to hire for potential rather than searching for an elusive perfect fit. In addition to showing how work experience enhances the value of human capital over time, the analysis quantifies the skill differentials associated with specific job moves. Zeroing in on the tech professionals in the data set shows that people routinely break into tech from other fields, and they make substantial shifts in skills and specialization when they do.

Our research shows that people are capable of mastering distinctly new skills and that unconventional tech hires are not so unconventional after all. But the willingness to hire them and the commitment to help them expand their capabilities require a shift in thinking.

The tech professionals in our data set are well paid and mobile—and 44 percent of them started in nontech occupations

Our data set of four million de-identified online work histories in four countries includes roughly 280,000 tech professionals, and it’s clear that they earn more and move more often than workers in other fields. Roughly 90 percent of the tech occupations we analyzed deliver above-average lifetime earnings. While workers across all professions changed roles every 3.2 years on average, tech professionals moved almost 20 percent more often, switching roles every 2.7 years.

We parsed millions of online job postings to quantify the “skill distance” associated with specific job moves (the share of new or nonoverlapping skills associated with the new job when someone makes a change). The size of the differential reflects someone’s opportunity to acquire or deploy additional skills when they assume a new role. People who start in tech typically overcome a skill distance of 27 percent every time they change roles.

More intriguing for hiring managers is the subset of tech professionals who started out in other types of occupations. These are not the experts who earned computer science degrees and never deviated from their chosen path. These are people who started out in entirely different lines of work and then reinvented themselves by adding new abilities along the way, perhaps learning to code, understand web architecture, or develop apps.

This is a common phenomenon in tech. Forty-four percent of the individuals who held tech roles at the end of the period we observed transitioned from non-IT occupations (Exhibit 1). To do so, they had to master a greater share of distinctly new skills—and their reward for doing so is upward mobility