How Hiring Tech Backfired - and Made Everything Worse
Labor market economics for the lay person
Title: The Efficiency Illusion: How Hiring Tech Backfired on Everyone
You apply to 200 jobs. You hear back from three. Maybe.
Meanwhile, employers insist they can't find anyone to hire.
That isn’t a coincidence. It’s a system failure.
Hiring software was supposed to solve this. But instead, it quietly made everything worse.
The Promise of Hiring Tech
Applicant Tracking Systems (ATS) were introduced to make hiring easier. They promised:
Faster resume filtering
Objective, bias-free evaluation
Streamlined processes for overwhelmed HR teams
They sounded like magic. But the more we used them, the more the system broke down.
3 Ways the System Backfired
1. Inflexibility
ATS systems sort resumes using rigid yes/no filters. But human beings are not binary.
It’s not just about whether or not they understand the difference between yes and no for things. Its about when are they answering it, between work meetings, or right before dropping off a kid? How long is the jargon? What is the background of a person with that jargon? Does the plumber recognize what that all means?
How do doctors offices handle jargon like this when you walk in. Do they go over to make sure you got the right answer or do they just make a diagnosis right from what you input while in the waiting room? Why is that relevant?
Is a simple error like a skimmed phrase what you want to rule out good candidates?
A Harvard Business School study found that 88% of employers admit qualified high-skill candidates are screened out because their resumes don’t match job descriptions exactly. That number jumps to 94% for middle-skill roles. (Harvard Business School, 2021)
Instead of broadening the pipeline, ATS narrowed it. We replaced human nuance with brittle black/white logic.
2. Language Mismatch
If your résumé doesn’t use the “right” keywords, file types, fonts, or layouts, it may be instantly rejected by applicant-tracking systems—even if you’re highly qualified. A 2023 Wall Street Journal article reports that outdated or non‑ATS‑friendly formatting can knock resumes out of the pile before humans ever see them (WSJ, 2025) Career-tech specialists like Jobscan and LinkedIn consistently warn that complex formats, unusual fonts, headers/footers, tables, images, or creative sections can “confuse ATS” and cause important content to be missed (JobScan, 2025)
No major ATS vendor has publicized a fix or provided transparency on whether these issues are fully resolved. Instead, there’s a silent trust in “perfect” resume design, even though millions of qualified candidates may still be filtered out for trivial reasons. That lack of public-facing data or accountability means there’s no meaningful debate about who’s responsible or how we’d know if ATS actually works as advertised.
True story: When it comes to recruiting companies their ability to handle it isn’t overly better. One company filtered applicants for “plant experience” and got a flood of botanists. They meant manufacturing. (LinkedIn anecdote, Problems with ATS, 2023)
ATS systems can't correct for human miscommunication. And they reward applicants with money and time to game the system—amplifying socioeconomic bias. (Wired, 2022; Zippia, 2023)
3. The Efficiency Illusion
The goal was faster hiring. But the result? Longer unemployment, more ghosting, and higher churn. Which meant increased use of the ATS systems that CAUSED this. It’s a self re-enforcing system of inefficiency.
Unemployment durations have grown from 12 weeks in 1990 to over 22 weeks in 2022. (Statista, 2023) Job seekers send more applications to break through filters. Employers rely more on automation to handle the flood. The cycle feeds itself.
Amazon scrapped its own ATS after it found baked-in gender bias. (Reuters, 2018) Interestingly part of the issue there is that when an employer KNOWS that bias is baked in - because its their system, they are liable for hiring discrimination. When they don’t know because its a black box of another company - they are exempt.
In reality - isn’t it interesting on that light - that so many companies conceal which hiring system they are using? Job seekers rarely know if they are engaging with Greenhouse vs Workday vs ICIMS vs others. They think its the company website because they use company names for the recruiters and websites they are moving from.
Wouldn’t job seekers be less likely to call out what seems discriminatory if they thought they were talking to the company itself? Less likely to ask for help and seem vulnerable because they didn’t realize the website wasn’t the company website and it was ok to?
And wouldn’t the company be able to create more opportunities for unofficially official discriminatory practices by having access to an ATS system that they don’t officially “know” how it works? Or I mean what kind of questions do they ask when deciding between Workday vs Greenhouse vs otherwise. Is it about the price point? Is it about the depth of the language model? Is it about the work family associations and the lexical phrase coordination? How do we know and how does that knowing impact reality?
This is not about knowing an answer to these questions. Its about that not knowing a clear answer is a bit of a red flag all by itself. Shouldn’t we know? Why wouldn’t job seekers know? Why wouldn’t they deserve to be notified so they can make informed decisions?
What IS the goal of letting so many people wander around unable to make informed decisions about processes that affect their access to work aka financial safety and socio-economic mobility?
Recruiters now spend more time managing software than engaging humans and it shows too.
4. Questions as to Tech Support
When a candidate can’t tell that they aren’t on the company website, they are on a 3rd party website - how often do you think that affects their ability to ask for tech support?
While candidates use of ATS systems makes up the ATS systems business, how often are candidates supported in their use of that tech? Do they receive FAQ pages, and a tech help line? Are they trained in updates to the system and notified that its aok to ask questions because we aren’t the employer? Or is that restricted?
Are they properly trained on those systems? Or is that training provided to the people who do it 24/7?
And if the people who use it 24/7 require training, require tech support and require FAQ pages - despite their extensive experience in these systems - how often does that make ATS systems “break” for candidates, and how profitable might those breaks be for ATS companies?
For example if the system existed to process the candidates, and candidates were well over 95% of the users according to BLS, how much money might you save by not offering full time tech support for the 95% of the users that drove your revenue. By not creating the pages.
And how profitable would those miniscule breaks in functionality be when spread over 100 million people?
These aren’t answers, these are questions, but its interesting that this information isn’t widely available to job seekers - the people it would affect.
The Systemic Problem: Jevons Paradox Hits Hiring
The Jevons Paradox says that when a tool becomes more efficient, we use it more—sometimes so much that we erase the gains. (Two stakeholder version I call the Micah Paradox described more fully in my SSRN)
That’s what happened here. Hiring tech made it easy to apply to hundreds of jobs. So people did. But the more we relied on automation to manage the flood, the worse the outcomes became.
Who Loses
Job Seekers
Millions of "Hidden Workers" — caregivers, veterans, immigrants, career switchers — get silently filtered out. Qualified applicants never even reach human eyes. (Harvard Business School, 2021)
Employers
You don’t just face a talent shortage. You face a signal breakdown.
You can't tell if a flood of resumes means demand or spam.
You can't tell if applicants actually want the job or are shot-gunning apps to get past filters.
You can't tell if you’re acting in good faith or if job seekers assume you're faking the role.
This breeds churn. Roles stay open. Trust erodes. And your internal culture suffers—because when people don't believe hiring is real, they stop acting like it's real.
Recruiters
You're stuck in the middle. Told to meet quotas while using systems that actively undermine your results.
What Needs to Change
We need transparency. Feedback loops. Audits.
The 2020 Sandvig v. Barr decision affirmed the right to test algorithms for discrimination. (ACLU, 2020) That means researchers and watchdogs can now legally examine hiring tech the way cybersecurity firms audit networks.
We need a market redesign—one that includes:
Third-party hiring audits
Verified candidate feedback systems
Two-sided accountability, like Yelp or Airbnb
Hiring isn’t a one-sided game. Without trust on both ends, the market collapses.
Let’s Make Hiring Work Again
Tech didn’t ruin hiring on purpose. But it did make the market more fragile, opaque, and wasteful.
If we want better results, we need better systems—systems rooted in transparency, feedback, and shared trust. Its what we are creating at The Job Applicant Perspective.
Want to go deeper? Read my full scholarly papers here:
👉 https://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=7663493
I also have a LinkedIn white paper written a few years ago describing many of these challenges
How ATS Harms US: 3 Critiques + Solution and References
It’s a bit older and before I had created my business and the solution as I see it - but it very much documents in depth a narrative look at this phenomenon.
Stay sane out there all. The gaslighting runs hard and I would wipe it with a Jedi mind wipe if I could. For now, clarity is my tool via words.
And if you’d like to be a part of the first user reviews of the hiring process - a way to start rebuilding trust in the labor market - stop by The Job Applicant Perspective