The Tech Hiring Market Smells Like Lemon

Over the years, one thing changed the way I think about hiring more than anything else: interviewing at scale.

When I was building teams, I noticed a pattern that became harder and harder to ignore. The engineers in my team and I spent a tremendous amount of time in interviews that shouldn't have been scheduled. Candidates looked strong on paper, but once we spoke, it often turned out they had only a vague understanding of the role, struggled to explain experience they had listed on their own CV, or showed little real motivation for the opportunity.

At some point, we tried avoiding such interviews by adding a simple 30-minute conversation early in the process. We added it just to answer three basic questions: Does this person understand the opportunity? Are they actually qualified for it (not just on paper)? And do they genuinely want to do it?

What shocked me was how effective that simple step became. It didn't filter out weak candidates in some dramatic way, but it showed us how much noise had entered the market. Very often, the problem was not deep technical nuance. It was that people had not prepared at all, could not clearly speak about work they claimed to have done, or were applying so broadly that this role was just one more tab open in the browser. It felt like most people were just spam applying everywhere.

That experience made me much more skeptical of surface-level signals in hiring. And it made me think of lemons. There is this famous economic idea: the market for lemons.

In 1970, economist George Akerlof described a market failure that became one of the most important ideas in information economics: the market for lemons.

Even though the paper is complex and supports the claims with a lot of math, his core insight was simple. When buyers cannot reliably tell high quality from low quality before making a decision, they stop paying for quality. Over time, that pushes the best options out of the market and leaves behind more of the low-signal ones.

That idea came from used cars, but I think it maps surprisingly well to parts of the tech hiring market today.

Employers are overwhelmed. Application volume is up. Noise is up a lot. Trust is down. Candidates are competing in crowded funnels where attention is scarce, filtering is aggressive, and it is increasingly difficult to stand out through genuine quality alone.

This is what a lemons dynamic looks like in hiring. When employers cannot easily distinguish between truly strong candidates and highly optimized candidate profiles, they become more defensive. I've noticed it myself. They lean harder on proxies: brand names, perfect title matches, referrals, years of experience, and increasingly rigid filters.

When candidates cannot distinguish between serious employers and black-hole processes, vague job posts, fake listings, or endless interview loops, they also adapt. They mass-apply, over-polish, automate, tailor aggressively, and try to beat the system before the system filters them out. Many AI-companies have popped up that help candidates to spam the market even more.

I believe that most candidates are not “lemons.” The problem is that the market starts treating everyone as potentially indistinguishable. And once that happens, surface-level signals lose value.

10 years ago, a polished CV used to mean something. People prepared for their interview, or even put on some nicer clothes. Today, many of those things are easier than ever to generate. When synthetic polish becomes cheap, polish stops being a trustworthy proxy for quality. That creates a serious market distortion.

High-quality candidates dissapear because they are buried in volume. Junior candidates get locked out because employers no longer want to “take the risk” of potential and instead demand proof. Strong but unconventional candidates get missed because they do not fit the narrow pattern-matching rules the market falls back on when trust is low. Just getting invited to an interview is an achievement today.

In economic terms, the issue is information asymmetry.

Candidates know far more about their real ability, work ethic, judgment, and communication than an employer can infer from a resume. Employers know far more about the real quality of a role, manager, team, and process than a candidate can infer from a job description. Both sides are making decisions with incomplete information. Both sides know the other side may be optimizing presentation. So both sides trust less.

The answer to a lemons market is better signaling. In hiring, better signaling means giving both sides access to evidence that is harder to fake and easier to understand: real context behind experience, clear role requirements, great screening tools, structured evaluation, proof of competencies, transparent process design, and human judgment where it matters most.

That is also why we believe the future of hiring is not a bigger database. Many tools promise access to 100s of millions, if not "billions of candidates".

The platforms that win in the next decade will not be the ones that flood employers with more candidates. As a hiring manager I don't want more candidates, that's just more work. I want the right candidates.

Companies that reduce uncertainty. They will help companies understand not just who mass-applied, but who can actually do the work, in what environment, at what level, and why. They will help candidates be seen for real capability, not just formatting, buzzwords, or who was best at gaming the funnel.

That is part of the problem we care about at Talentsearch. We do not think hiring is broken because there is no talent. We think it is broken because the market has become extremely inefficient, too noisy to recognize talent reliably.

The result is exactly what Akerlof warned about decades ago: when quality becomes hard to observe, markets drift toward mistrust, defensive behavior, and worse outcomes for everyone.