Our interview practices are still filtering out innovative candidates
- Michael Amenta
- Jul 9
- 4 min read
Knowledge workers don't all excel in the same ways. It's time to update our approach.
Seven years after David Epstein's hit book Range made a compelling case for breadth over specialization, little seems to have changed in hiring practices. Outside of a few areas in executive search and venture capital, candidate selection for so-called 'wicked' roles — where innovative thinking is required — has become even more focused on specializations via the growth of ATS and AI screening on skills and domain fit. And subsequent interviews often continue this pattern by failing to examine breadth.
The lack of change may stem from how we categorize our jobs, especially since our classification of knowledge workers hasn't been formally updated in decades. Much has changed in the past 70 years — with revolutions in computing, the internet, and AI — but we still largely examine the same two buckets of work that were established in the 1960s: "knowledge work" occupying white-collar jobs and "manual work" in blue-collar ones.
Technical work informally occupies a distinction against non-technical work that is soft-skill dominant, but this categorization is incomplete. For example, we don't commonly update interview and selection criteria despite some workers operating in Epstein's so-called 'kind' environments, where excellence is derived from predictable tasks, while others operate in 'wicked' environments that are more chaotic.
Instead, a more useful taxonomy could have four groups. Routine cognitive work can include specialists in operations, support, or compliance. And non-routine cognitive work can be broken down into analytical (e.g. performance marketing, engineering, and finance), interpersonal (people management, sales, and PR), and innovation (architects, scientists, and product managers) workers.
These groups are flexible by role, seniority, and situation. For example, medical professionals span groups: surgeons are closest to routine cognitive workers, general practitioners to analytical, and research doctors to innovation workers. However, all three are senior-level positions that require interpersonal and analytical attributes; and innovation skills are also required of top leaders in any field.
The consequence of not using this more specific classification is acute with the innovation group because they most clearly fit into Epstein's 'wicked' environment, where he finds that generalists outperform specialists. There, we're still not screening for pivots over pedigree, analogous thinking, and aptitude over exact fit.
Innovation skills revealed
Research indicates that inventiveness is most commonly associated with cross-functional experience and thinking. A 2007 Harvard Business School study of innovative problem solving led by Karim Lakhani found that the "further the focal problem was from the solvers’ field of expertise, the more likely they were to solve it."
Lakhani's paper aligns with other research findings that demonstrate having breadth beyond one's specialization is the sweet spot for innovation. This is known as having a "T-shaped" profile, where breadth of experience is indicated by the cross bar and depth in a specialization is indicated by the vertical line. T-shaped skills were originally attributed to McKinsey's selection of consultants in the 1980s, but were later made famous by Tim Brown, former CEO of the prolific design firm IDEO, who found these individuals were best suited for design thinking, brainstorming, and ingenuity.
Where domain specialists might fall into familiar patterns of thinking, cross-domain solvers can more easily draw upon similarities to analogous problems that other fields have already solved. This was reinforced in a study of over 17 million research articles across five decades led by Brian Uzzi that found the highest-impact papers were the ones combining an atypical component from a separate field with conventional patterns in the target field.
Character traits and environment matter too. Lakhani's research also found that intrinsic motivations like enjoying problem solving are "stronger and more significant" factors than a desire to win substantial award money. This finding echoes what Viktor Frankl identified about meaning as a primary motivator. In addition, having enough free time to explore the problem was also positively correlated with success, while career and social motivations were negatively correlated.
All these findings show that innovation worker interviews need to stop prioritizing precise fit and start actively searching for ingenuity. This might map to a question like: "tell me about a time you borrowed an idea or method from a completely different field to solve a problem in your domain?". What distinguishes a strong answer is whether the candidate can name the conventional pattern, identify the atypical element, and articulate the combination. In addition, a candidate who has published case studies, articles, or otherwise documented their decision-making is more verifiable and signals knowledge sharing that's linked to increases in innovation across teammates. Finally, a personality assessment can also be considered since the Big Five's "openness to experience" dimension is consistently identified as having the largest positive correlation with innovation across domains.
The product example
These findings sync with my personal experience in product management, where breadth hasn't been actively targeted and success stories are rarely examined for the cross-domain thinking that produced them.
For example, our hackathon-winning QA system drew upon investment risk management concepts, primarily z-scores, to identify data entries that were likely errors. Later, in our consumer financial app, we increased account linking rates by 10x from leveraging gamification methods used by consumer apps like PlayStation, Duolingo, and Peloton. We then doubled the top of our prospect funnel, while simultaneously boosting revenue conversion rates, by establishing a "freemium" website that had been popularized by SaaS applications. And finally, we significantly increased financial advisor contact rates by leveraging asynchronous messaging portals that had been successful at driving high-efficiency contact with medical doctors.
In each example, adding an atypical component from another field (breadth) to conventional norms (depth) was the recipe for impactful growth.
Advancing hiring practices in innovation roles
Unfortunately, the same situational questions applied to a salesperson or performance marketer are routinely used to evaluate product roles that require a different assessment. With ATS screening for skills and domain fit at peak levels, and with AI providing high-quality capabilities within domains, there's more urgency than ever to recenter talent selection on what's more rare and precious: the ability to identify and apply novel combinations across domains.
Epstein showed us that range matters in wicked environments. It's time to deliberately hire for it in the roles where range matters most.