A Messe É Grande E Os Operários São Poucos

Why your hiring pipeline is always dry and what to actually do about it

I've been running recruiting operations for mid-sized companies since before LinkedIn was a real thing, and the bottleneck never really changes. You have work to get done and not enough hands to do it. The pattern repeats across industries — software, manufacturing, logistics, healthcare support. The problem isn't that nobody wants to work. It's that your hiring process is filtering out good people before they ever reach you, and the ones who do make it through are often overpromised and underprepared. Here's the practical side of it that most people skip. When you're trying to fill roles and the candidate pool looks thin, the first instinct is to post more listings and hope something sticks. That's backwards. Start by figuring out which parts of your process are actually blocking qualified people from applying or accepting offers, then fix those first. I've seen time-to-hire drop from six weeks to eleven days just by removing a redundant technical screening round that nobody could explain the purpose of.

a messe é grande e os operários são poucos

This old line about the harvest and the workers isn't just a religious quote people paste into LinkedIn posts. It describes an actual structural problem in operations management. The demand for skilled labor outpaces the supply because companies keep asking for the same narrow set of qualifications instead of adapting what they're looking for to the reality of the available pool. You'll see this repeatedly in job descriptions that list seven years of experience for a role that barely existed five years ago, or require a specific tool someone probably learned on the job three months ago. The counter-intuitive part most hiring managers miss is that widening your criteria doesn't lower quality. It raises the signal-to-noise ratio. When you restrict candidates to an arbitrary checklist, you're not getting the best people. You're getting the people who happened to accumulate the right credentials, which is a different thing entirely. A developer who built three production systems in two years usually knows more than someone who sat at the same company for seven and never shipped anything complex. The data supports this if you actually look at performance metrics post-hire, not just hiring speed. I ran into a specific case last year where we needed five DevOps engineers for a rapid expansion. The standard posting got twelve qualified applicants out of three hundred total applications. We were about to extend offers to two people and leave three seats empty when I realized the job description was the bottleneck. We required AWS certifications, Terraform experience, Kubernetes mastery, and five years in production. The market for people with all four of those things plus five years of it was roughly two hundred individuals nationwide, and half of them were already employed at companies that weren't laying off.

The workaround was brutal but simple. We split the hiring into two tracks. Track one was for senior engineers who already had the full stack — those went straight to final rounds. Track two was for competent engineers who had three of the four requirements and a track record of learning the fourth quickly. We gave track two candidates a paid two-day practical assessment instead of the usual three-week interview gauntlet. Within three weeks we had five people onboarded. Three of them were from track two. One of them later became our lead on the Kubernetes migration. There are downsides to this approach that nobody talks about openly. Splitting into tracks means your hiring team needs to actually evaluate skills instead of checking boxes. That requires calibrated interviewers, which means investment in training people who are already understaffed. The paid assessment costs money upfront — we spent about four thousand dollars on contractor time for the two-day exercise across five candidates. It pays for itself when you compare it to the cost of a bad hire, which in our experience runs anywhere from eighteen to twenty-four months of salary in replacement and productivity loss. If you're in a smaller company without the bandwidth to redesign your entire hiring pipeline, there's a cheaper alternative. Partner with local training programs or bootcamps that specialize in the skills you need. I've used this successfully with a community college's cloud computing program and a few independent cybersecurity training outfits. They feed you pre-screened candidates who have the fundamentals and the hunger. The tradeoff is that these people will need more mentoring upfront, so factor that into your onboarding timeline. Budget six weeks of structured ramp-up instead of the two weeks you'd give someone who already knew the stack.

Another thing that quietly kills hiring effectiveness: compensation transparency. Post the salary range. Not a lowball number and not a vague "competitive." An actual range. I tracked this internally over eighteen months and found that postings with a real salary range saw a 40 percent increase in qualified applications and a 25 percent reduction in offer rejection rates. People are tired of playing guessing games. Give them the number and move on. The hardest part about solving the worker shortage isn't the recruitment tactics. It's admitting that your current expectations about what a candidate should look like or what they should already know are anchored to a labor market that no longer exists. The people who adapted fastest during recent hiring crunches did one thing differently: they started hiring for trajectory instead of history. That's a uncomfortable shift for managers who've built their careers on evaluating past achievements. But trajectory is what actually predicts future performance in roles that are still evolving, which is most of them now. If you want a concrete starting point, audit your last three open roles and identify the single requirement that eliminated the most qualified candidates. It's usually obvious once you look at the rejection data. Then remove it or replace it with a practical assessment. You'll be surprised how many people were blocked by something that had nothing to do with doing the actual job.