Applicant tracking systems (ATS): what they do, and where they fall short

Almost every company above a handful of employees runs hiring through an applicant tracking system, and almost nobody outside recruiting quite understands what it does. That gap is how a small mythology grew up around it, most of it about resumes being silently devoured by keyword-hunting robots. It is worth separating what an ATS actually is from the folklore, because the real limitations are different from the imagined ones, and they change how you should build the rest of your process.

The short version: an ATS is a system of record and workflow. It is excellent at logistics and neutral on quality, which means it will organize a biased process exactly as efficiently as a fair one. This guide explains what it genuinely does, debunks the keyword myth, names where ATS-centric hiring actually goes wrong, and covers what to pair it with so the workflow tool is not quietly making your selection decisions for you.

What an ATS actually does

At its core, an ATS does four mundane but valuable things. It collects applications into one place, stores a structured record for each candidate, moves people through defined pipeline stages, and coordinates the communication and scheduling around all of that. For a team handling hundreds or thousands of applicants this is genuinely essential. Without it, hiring dissolves into lost emails and forgotten candidates. It is, in other words, infrastructure.

What it is not is an evaluator. An ATS has no opinion about who is good; it presents a list and tracks what humans do with it. That distinction is the whole point of this article, because the moment people treat the ordering of that list as a judgment of merit, a logistics tool starts driving selection decisions it was never designed to make.

The keyword myth

The most durable piece of ATS folklore is that it auto-rejects any resume missing the exact keywords from the job description, which has spawned an entire industry of advice on how to beat it. The reality is more boring. While modern systems can filter and rank, the idea that they silently bin qualified candidates over a missing synonym is largely overstated. Most rejection still comes from a human skimming the list the ATS presents, which means the fairness problem lives in the skim, not the software.

This matters because it redirects your attention to the real risk. Optimizing resumes for keywords is mostly wasted effort; fixing how the list gets reviewed is where the leverage is. That is the same insight behind reducing hiring bias: the damage happens in fast human judgment, so that is where the structure needs to go.

Where ATS-centric hiring falls short

The limitation is not a flaw in any particular ATS, it is a category boundary. Because the ATS surfaces resumes and tracks stages, teams that lean on it as their primary decision aid end up making the same resume-driven, proxy-heavy judgments it was never meant to improve. It will faithfully run a process built on pedigree and keywords, complete with tidy dashboards, and give you no signal at all about who can actually do the work. The danger is precisely that it makes a weak process look organized and data-driven.

The fix is not to replace the ATS, because you still need the workflow. The fix is to put a real evaluation layer in front of the human skim. A skills-based assessment or a structured AI interview changes what the list is ordered by, so the ATS ends up tracking a merit-based pipeline rather than a pedigree-based one.

Choosing and sizing an ATS

If you are shopping for one, resist buying on feature count. The questions that actually matter are smaller and more practical:

• Volume: how many applicants and roles are you really handling? Heavy configurability is wasted at low volume and essential at high volume.
• Integrations: will it connect to your job boards, calendar and HRIS without constant maintenance?
• Workflow fit: does it match how your team actually hires, or will you bend your process to its assumptions?
• Total cost of ownership: setup, training and the integrations you will maintain, not just the licence.
• What it does for decision quality: usually nothing, which is the point. Plan to add that layer separately.

What to pair it with

The healthiest stack treats the ATS as one half of a pair. The ATS owns logistics: where every candidate is, what was said, what happens next. A dedicated evaluation step owns quality: who is actually good, judged consistently on the skills the role needs. Keep those two jobs distinct and each does its own well. Blur them, and you get a tidy pipeline making poorly-informed decisions.

How Spoon Hire complements your ATS

Spoon Hire is not trying to replace your system of record; it is the evaluation layer that makes the record worth tracking. Candidates are assessed through a structured, skills-first interview and surfaced as an anonymized, merit-ranked shortlist, so what flows into or alongside your ATS is ordered by demonstrated ability rather than by who had the most keyword-optimized resume. For a larger team it sits in front of the ATS; for a small team it can remove the need for a heavyweight one. See how it fits the wider tool stack or how it works for companies.

Frequently asked

Keep reading