Ranking · Candidate scoring software
Application scoring software as the applicant scoring system that scores every applicant against your criteria and ranks the pile
Candidate scoring breaks down the moment two reviewers grade the same resume differently. Without a shared, explicit set of criteria, a score reflects the reviewer's mood and instincts as much as the candidate, and the numbers cannot be compared or defended.
Confirm your work email, add up to 3 of your own resumes, see them ranked with reasons
Held for recruiter review · never auto-rejected
Screen the pile to watch Prescreener parse every resume, apply your criteria and knockouts, and rank the inbound applicants with a transparent reason on each.
Live, interactive · AI assists, you decide · no signup needed
AI assists · you decide · impact ratio reports for your bias audit
Human-in-the-loop you decide
Impact ratio reports for your audit
Prescreener brings consistent candidate scoring to the inbound pile. It reads every resume, applies your knockout and eligibility criteria, and scores role-fit against the same defined rubric for every candidate, then ranks the pile with a transparent reason behind each score. Recruiters review the ranked results and make every decision, since nothing is auto-rejected. The scoring comes with impact ratio reports for your EEOC and NYC Local Law 144 audit work and runs with candidate consent and AI disclosure, so the numbers hold up under scrutiny.
Why it works
What your team gets with candidate scoring software
One rubric for all
Every candidate is scored on the same defined criteria, so the scores are comparable across the pile rather than shaped by who reviewed whom.
Reason per score
Each score links to the criteria and resume evidence behind it, so a number is explainable rather than something you have to trust blindly.
Scores feed the rank
Role-fit scores drive the ranking, so the prioritized shortlist is a direct, auditable result of the criteria you set.
What it handles
Parsed, scored and ranked on autopilot
Prescreener reads every inbound resume, applies your knockout and eligibility criteria, asks the screening questions you set, scores role-fit against consistent criteria, and ranks the applicant pile so your recruiters review the top matches first.
- Scores every candidate on the same role-fit criteria
- Links each score to its criteria and resume evidence
- Ranks the pile from the scores with transparent reasons
- Lets recruiters review and decide on every candidate
- Impact ratio reports for your EEOC and NYC Local Law 144 audit work with candidate consent
Why Prescreener
One step that screens the whole inbound pile
Not a full ATS, not a six-figure assessment suite, and not a staffing agency. Parse, score, rank and hand off in one place, shaped to the criteria you already hire on.
Reads every resume
Prescreener parses every inbound application, applies your knockout and eligibility criteria, and asks the screening questions you set. Applicants consent and are told AI is screening their application, so no qualified person sits in a backlog for days.
Scores role-fit consistently
Every applicant is scored against the same criteria, with a transparent reason on each candidate, so the pile is compared consistently and the scoring comes with impact ratio reports for your EEOC and NYC Local Law 144 audit work.
Ranks the pile
The strongest matches rise to the top of a ranked pile your recruiters review first. Prescreener never auto-rejects, it ranks and flags candidates for review, and your team makes every decision.
In depth
Candidate scoring criteria, matrices and rubrics that hold up
What is a candidate scoring matrix?
A candidate scoring matrix is a grid that lists the criteria a role is actually hired on down one side and the candidates across the top, with a score in every cell and a weight against every criterion. The total orders the shortlist. It is the same instrument as a candidate scoring rubric or a candidate scoring sheet; the names change by industry, the mechanics do not.
What makes a matrix work is not the arithmetic, it is the discipline of writing the criteria down before anyone reads a resume. Criteria written after the fact tend to describe the candidate somebody already liked, which is how a hiring committee ends up with a documented process that produces exactly the decision it would have made without one.
Keep it to six to nine criteria. Fewer than six and everything scores the same, so the matrix stops separating anyone. More than nine and reviewers start guessing on the ones they have no evidence for, which adds noise dressed up as precision. Every criterion should be answerable from the application itself or from a specific question you ask every candidate.
How do you score candidates consistently?
Consistency comes from anchoring the scale, not from asking reviewers to try harder. A 1 to 5 scale where nobody defined what a 4 means will drift within a single afternoon, and it will drift further between two reviewers who have never compared notes.
Anchor each point with an observable statement. For a criterion like "high-volume AP experience", a 5 might read "has owned AP end to end at 1,000 invoices a month or more", a 3 "has processed AP as part of a broader finance role", a 1 "no evidence of AP work". Now two people scoring the same resume land in the same place, and a third person reading the scores six weeks later can still tell what they meant.
The scoring model below is the shape most teams end up with once they separate what gates from what ranks. Knockouts are pass or fail and never get a weight, because a candidate without the required license is not 30 percent qualified, they are ineligible.
| Criterion type | How it is treated | Example | Typical weight |
|---|---|---|---|
| Knockout | Pass or fail gate, never scored | Work authorization, required license, shift availability | Not weighted |
| Core skill | Scored on an anchored scale | Years running the specific process the role owns | 30 to 40 percent |
| Domain experience | Scored on an anchored scale | Worked in the same industry or regulatory environment | 15 to 25 percent |
| Scope and scale | Scored on an anchored scale | Team size managed, transaction volume, budget owned | 15 to 20 percent |
| Credential | Scored, rarely a gate unless legally required | Certification that is preferred but not mandatory | 5 to 15 percent |
| Preference | Scored last, low weight | Nice-to-have tool familiarity | Under 10 percent |
Swipe to see the full comparison →
Why manual candidate scoring drifts
A scoring sheet solves the criteria problem and leaves the capacity problem untouched, which is why so many carefully designed matrices quietly stop being used by week three.
Scoring one candidate properly against eight anchored criteria takes six to ten minutes. On a requisition with 300 applicants that is 30 to 50 hours, so the matrix gets applied to the twelve people a recruiter already shortlisted by skimming. The scoring then documents a shortlist that was assembled without it. The output looks rigorous and the selection was not.
There is a second drift that is harder to see. Scores assigned on Monday with a full hour differ from scores assigned on Thursday between two intake calls, even from the same reviewer using the same sheet. Nobody is being careless. Human judgment on a repetitive comparison task is simply not stable across a week, and the effect is largest on the middle of the pile, which is exactly where the marginal hire comes from.
Software fixes the capacity half and inherits the criteria half. A tool that scores every application against your anchored criteria removes the sampling and the day-to-day drift, but it will apply a badly specified criterion just as evenly as a good one. The hour spent writing the matrix is still the hour that decides whether any of this works.
What makes a candidate score defensible?
Defensible means a person who disagrees with the ranking can see what produced it, and a person who has to explain the process later can do so without reconstructing anybody's memory.
In practice that needs four things. Criteria written before review and unchanged during it. An anchored scale, so a score means the same thing in every cell. Evidence attached to each score, meaning the specific line in the application that justified it. And a record that survives, so the reasoning is still readable when someone asks about a requisition you closed two quarters ago.
This is also the practical compliance position. Title VII, the ADA and the ADEA apply to any selection procedure, whether it runs in a spreadsheet or in software, and New York City Local Law 144 adds an annual bias audit and candidate notice where automated employment decision tools are used. A consistently applied, evidence-backed score is easier to stand behind than an undocumented skim, which is what it is usually being compared against.
One rule holds regardless of the tooling: scoring ranks and flags, it does not decide. Nothing should be auto-rejected on a score. A recruiter reviews the ranked list and makes every call, which is both the safer position and the one that produces better hires, because the score has no idea which candidate just relocated to your city.
Application scoring system that ranks every applicant with reasons
An application scoring system reads each application against the criteria you set for the job, gives it a score, and sorts the pile so the strongest applicants sit at the top. The useful ones also write down why: which requirement the resume met, which it missed, and the line of the resume that shows it. Without that evidence a score is just a number a recruiter has to take on trust.
Most applicant tracking systems stop short of this. Some filter on knockout answers, some return a check or a cross per criterion, and some rank only the candidates a recruiter searches for. Prescreener scores every inbound application as it arrives, applies the knockout rules the same way to everyone, and attaches the evidence to each score, so an applicant ranking system and an audit trail come from the same pass. A recruiter still makes every decision.
| What the system does | Knockout filter only | Prescreener application scoring |
|---|---|---|
| Reads every application | Reads the screening answers, not the resume | Yes, resume and answers |
| Score per applicant | Pass or fail | Role-fit score against your criteria |
| Evidence behind the result | The answer given | The resume lines behind each criterion |
| Ranks the pile | No | Yes, strongest first |
Swipe to see the full comparison →
What types of systems let recruiters attach evidence links to each scoring criterion automatically?
Evidence-linked scoring comes from AI screening and scoring systems that read the full application, score each criterion separately and cite the passage that justified each score, rather than from a classic applicant tracking system. Most ATS scorecards store a number per criterion and leave the evidence field for the reviewer to type in by hand.
The distinction to test in a demo is whether the evidence is generated or merely allowed. Plenty of systems have a notes box next to each criterion, which means a reviewer can attach evidence if they have time. On a requisition with 300 applicants they will not, so the box stays empty on everyone below the top twelve. A system that attaches evidence automatically writes the justification for every criterion on every applicant, including the ones nobody opened.
Three questions separate the two in a vendor call. Does every criterion score come with a quoted line from the resume or an answer the candidate gave, not a generic sentence? Is that evidence produced for the whole pile or only for candidates a recruiter has already advanced? And does it survive export, so the reasoning is still readable when someone asks about a requisition you closed months ago? Prescreener answers yes to all three: each criterion score carries the resume evidence behind it, for every applicant, and a recruiter still makes every decision.
| System type | Scores each criterion | Evidence attached automatically | Covers the whole inbound pile |
|---|---|---|---|
| ATS scorecard | Yes, entered by a reviewer | No, free-text notes typed by hand | Only candidates someone opens |
| ATS with keyword or match score | One overall match figure | Rarely, usually matched keywords only | Yes |
| Criteria-based AI screening that refuses to rank | Met or not met per criterion | Short reasoning per criterion | Yes, often capped per month |
| AI scoring and ranking layer (Prescreener) | Yes, per criterion, weighted | Yes, resume evidence per criterion | Yes, every applicant |
Swipe to see the full comparison →
Candidate scoring systems compared by how they score
Candidate scoring systems fall into four groups, and the group matters more than the brand. Rules-based systems score knockout answers only. ATS-native AI puts one number or band on each applicant. Criteria-count tools say how many of your requirements a candidate meets but refuse to rank. A dedicated scoring layer scores every criterion, weights them, ranks the pile and shows the evidence.
The published examples make the difference concrete. Bullhorn gives a Screening Score from 0 to 100, Oracle Recruiting Cloud rates applicants 0 to 5 once an admin switches it on, and Eightfold rates 0 to 5 stars in half-star steps. Greenhouse Talent Matching sorts applicants into five match bands on one pipeline stage. Recruitee returns a count such as 5/5 criteria met and states that it will not give rankings or suitability scores. Each is a legitimate design; they answer different questions.
Price follows the same split. Rules-based scoring usually comes with the ATS. AI scores inside an ATS are often gated to a higher tier or a separate license, as the ATS pricing comparison shows vendor by vendor. A standalone layer is priced on its own and runs beside whichever ATS you keep.
| Type of candidate scoring system | What the score looks like | Examples | Ranks the whole inbound pile |
|---|---|---|---|
| Rules-based knockout scoring | Pass or fail on screening answers | Knockout questions in most ATSs | No, filters only |
| ATS-native AI score | One number or band per applicant | Bullhorn 0 to 100, Oracle 0 to 5, Eightfold 0 to 5 stars, Greenhouse five bands | Often, on the tier or stage where it is enabled |
| Criteria-count screening | Criteria met, per requirement | Recruitee, Teamtailor | No, by design |
| Dedicated scoring layer (Prescreener) | Weighted score per criterion with resume evidence | Prescreener | Yes, every applicant, alongside your ATS |
Swipe to see the full comparison →
Good questions
Questions about candidate scoring software
Explore more
More ways hiring teams screen with Prescreener
Resume ranking software
Resume ranking software with a transparent reason behind every position.
Learn moreApplicant tracking system screening
The AI screening layer that ranks your inbound pile alongside your ATS.
Learn moreResume parser
A resume parser that does not stop at fields. It scores and ranks too.
Learn moreStop reading the inbound pile by hand. Put screening on autopilot.
Set your role and criteria and Prescreener parses every inbound resume, scores role-fit, and ranks the pile for your team. Prescreener ranks and flags candidates, your recruiters make every hiring decision.
Consistent criteria · impact ratio reports for your bias audit · human-in-the-loop