The Distance Between a Seventy and an Eighty Is Smaller Than Most Candidates Think.

ATS scoring produces a number. That number crosses a threshold or it does not. The outcome is binary: advance or reject. But the inputs that determine which side of the threshold a resume lands on are not binary at all. They are continuous, incremental, and often surprisingly small. A resume scoring 68 and a resume scoring 74 might differ by a handful of keyword matches and a slightly better-articulated skills section. The candidate experience of these two outcomes could not be more different.

We examined the score distribution across our production dataset to understand how resumes cluster around common screening thresholds, what separates near-misses from passes, and how much change is actually required to cross from one side to the other. The findings suggest that most rejected resumes are closer to passing than their candidates realize, and the gap is closable with targeted adjustments rather than wholesale rewrites.

Where resumes actually cluster in the scoring range

The first thing that stands out in the data is how tightly pre-optimization scores cluster. The majority of base resumes, before any optimization, land in a relatively narrow band. They are not uniformly distributed across the 0-100 range. They concentrate in the middle, with thin tails at the extremes.

The bulk of resumes fall between 30 and 60. No pre-optimization resume in our dataset scored above 72. The mean sits at 46, and the median at 44. This means more than half of all base resumes start below 50, well below typical ATS screening thresholds.

This clustering matters because it means the difference between a "typical" resume and a "good" resume is not a 30-point gap. The resumes near the top of the pre-optimization range (60-72) are not dramatically better than those in the 40-55 range. They tend to have better keyword alignment with specific job descriptions and slightly more structured skills sections. The gap is incremental, not categorical.

Why a continuous score creates a binary outcome

ATS screening thresholds convert a nuanced score into a yes-or-no decision. Whether the threshold is set at 70, 75, or 80 depends on the employer, the role, and the volume of applicants. But the mechanism is the same everywhere: resumes above the line get reviewed by a human, resumes below it do not.

The problem is that thresholds create an artificial cliff. A resume scoring 69 and a resume scoring 71 are nearly identical in quality, but one gets rejected and the other advances. From the scoring engine's perspective, there is almost no meaningful difference. From the candidate's perspective, the difference is everything.

At a threshold of 70, which is on the permissive end, only 10% of pre-optimization resumes in our dataset would pass. At 75, it drops to 3%. At 80, none pass. This is not because the resumes are terrible. It is because most resumes are written for human readers, not scoring engines, and the baseline keyword alignment between a generic resume and a specific job description is inherently low.

What actually separates a near-miss from a pass

We compared resume pairs where one version scored just below a threshold and the post-optimization version crossed it. The component-level differences reveal where the points come from, and how few changes are needed to bridge the gap.

The pattern is striking. Of the 19-point average gap between near-misses and passes, 13 points come from keywords alone. Skills contribute 4 more. Together, these two components account for 17 of the 19 points, or roughly 89% of the difference. Experience, education, and contextual fit contribute almost nothing to the gap.

This aligns with what we documented in our analysis of how skills and experience trade off in scoring: the components that candidates can change (keywords, skills articulation) carry 65% of the total weight, while the components tied to their actual history (experience, education) carry only 25%. The scoring system is, in effect, testing resume writing quality as much as candidate quality.

The actual edits that close the gap

If keywords are the primary differentiator, then closing the rejection gap is fundamentally about keyword alignment. But what does that look like in practice? We examined the specific content changes in optimization pairs where the pre-optimization score was within 15 points of a passing threshold.

Seven to twelve modified bullet points out of twenty or more. That is the scale of change we are talking about. This is not a resume rewrite. It is targeted adjustment of specific phrases to improve alignment with the job description's language. The candidate's actual experience, achievements, and career narrative remain intact. What changes is the vocabulary used to describe them.

Consider a concrete example. A project management resume described a past accomplishment as "led cross-functional initiative to streamline operations." The job description used the phrase "managed cross-departmental projects to improve operational efficiency." The meaning is identical. The vocabulary overlap is minimal. The optimized version adopted the job description's phrasing, and that single bullet point contributed measurably to the keyword score increase. Multiply that pattern across seven or eight bullet points, and you have enough movement to cross a threshold.

Starting score determines how far optimization can take you

Not all resumes benefit equally from optimization. Where a resume starts in the scoring range significantly affects how much it can gain. Resumes starting from very low scores have more room to improve on keywords but face harder constraints on experience and education. Resumes starting closer to thresholds need fewer changes but have less room for large gains.

The practical implication is important: resumes starting in the 50-69 range are in the optimization sweet spot. They have enough baseline alignment with the target job that keyword optimization can reliably push them past common thresholds. Resumes below 50 may need more fundamental changes, such as targeting jobs that better match their actual experience, before optimization can help. As we found in our study of the same resume scored against different jobs, the starting score often reflects job-resume alignment more than resume quality.

Why near-misses feel like complete failures

Candidates never see their ATS score. They submit a resume. They wait. They receive a rejection or they receive a callback. There is no feedback mechanism that says "you scored 68 and the threshold was 70." The experience of a near-miss is indistinguishable from the experience of a complete mismatch.

This lack of feedback creates a distorted mental model. A candidate who keeps getting rejected assumes their resume is fundamentally inadequate, or that they are not qualified for the jobs they are applying to. In reality, many of those rejections are near-misses. The resume was competitive but did not cross the line. The fix might be as simple as incorporating a few phrases from the job description.

The information asymmetry is the core problem. Candidates are making career decisions based on rejection signals that carry no diagnostic information. A candidate who understands that they scored 68 against a 70 threshold would make very different choices than a candidate who just knows they were rejected. The first candidate tweaks their resume and reapplies. The second questions their career trajectory.

What this means for how you approach applications

The data points toward a few concrete shifts in how candidates should think about the application process.

Where this fits in our broader research

This article closes a loop that started with our earliest scoring research. In The Anatomy of ATS Optimization, we showed where score gains come from mechanically. In our before-and-after analysis, we cataloged what changes and what does not. In our format analysis, we showed how structural choices affect parsing before scoring even begins.

This article adds the distribution lens: where resumes actually cluster, how thresholds create binary outcomes from continuous scores, and why the gap between rejection and advancement is smaller than it appears. The common thread across all of this research is that ATS scoring is more mechanical, more predictable, and more addressable than most candidates believe. The system is not inscrutable. It follows rules. Understanding those rules does not guarantee success, but it removes the biggest source of unnecessary failure: applying with a resume that does not speak the scoring engine's language.