The Resume Proof-Type Reference, One Row per Kind of Evidence

This is a lookup table for the kinds of evidence a resume bullet can carry. It is for anyone rewriting a resume, profile, or portfolio who keeps hitting the same wall: a line that sounds like a claim the reader discounts rather than proof they trust. Read it to name the proof type behind any bullet, understand what that type proves and how it misleads, and swap in the strongest honest evidence the work actually supports.

Most resume advice stops at "add a number." That is half the job and the dangerous half. A number can inflate ownership, hide a baseline, or borrow credit, and every one of those cracks under a single interview follow-up. What you need is a taxonomy: which proof type a bullet uses, what it genuinely demonstrates, and whether it holds up when someone asks how you got it.

What a proof type is, and why a bullet needs one

A proof type is the category of evidence a bullet offers to back its claim, such as scope, a before/after change, or a named award. A bullet without a proof type is an assertion; a bullet with the right one is evidence the reader can check.

The underlying standard comes from two MIT sources. MIT CAPD frames the credible unit as the PAR statement: Project, Action, Result. Including all three turns a duty into an accomplishment. MIT's Communication Lab states the principle more bluntly as "show, don't tell": instead of saying you are a great candidate, show the reader by providing evidence from your experiences, awards, and impact. The proof type is the "show." It is the specific, checkable thing that converts "responsible for customer success" into something a recruiter does not have to take on faith.

The reason this matters is the audience. In Refolk's index, 118,617 recruiters and talent-acquisition professionals in the United States are the human readers a bullet must survive. Each runs a seconds-long first pass, and the probe that kills a bad number is cheap for them to run later, in the interview or at reference-check. So the game is two-sided: the proof type has to be skimmable enough to register in the first pass, and defensible enough to survive the drill.

The proof-type reference table

Here is the catalog. Each row names a proof type, states what it proves, and names how it misleads - the false positive that makes an interviewer reach for a follow-up. Jump to the row you need.

Two things anchor how to read this table. First, proof types are not ranked best to worst in the abstract; the strongest one is whichever is both true and defensible for that specific bullet. Second, the "how it misleads" column is the most useful column, because it tells you in advance which follow-up an interviewer will aim at the bullet.

What actually counts as a metric

A metric is any number that a reader and a parser can score as quantified, and that is a much wider set than percentages. Time, count, and scope numbers count just as much as percentages for the quantification category - only vague words fail to count at all.

This is the single biggest unlock for anyone who never worked off a KPI dashboard. You do not need a percent sign to be quantified. A before/after time, a ticket volume, a team size, a dollar budget - all of these parse and read as real proof.

A word on the ATS, the applicant-tracking system that parses your resume before a human sees it. The sources disagree on whether numbers raise the machine score. One parser-maker states plainly that numbers do not directly affect ATS scoring but make bullets more compelling during human review. Another describes a context multiplier of about 1.2x when a bullet carries a measurable outcome, and a third vendor claims quantified achievements are worth 10% of the score. Treat these as contested. The safe reading: numbers earn their keep with the human, and a bullet can be perfectly quantified and still skipped if the keyword context is absent.

Scope versus outcome, the two proof types people confuse

Scope proof describes what you were responsible for; outcome proof describes what changed because of you. They are not interchangeable, and swapping one for the other is the most common honest mistake on a resume.

A scope bullet reads "Managed a $3M budget across four product lines." It establishes authority. An outcome bullet reads "Cut onboarding time from 11 days to 4." It establishes impact. Scope is easier to reconstruct and harder to fake, but it carries a built-in risk: it can imply ownership you did not have. "Managed $3M budget" borrowed from a team dashboard you did not influence reads as authority you did not hold. Outcome is more persuasive when it is genuinely yours, but it tempts you to hide the baseline.

The practitioner move is to lead with outcome where the work and the baseline are honestly yours, and fall back to scope where the outcome was a team result you cannot claim alone. Then make the subjective lines the most reconstructable, because skills and responsibilities are the top two areas where resumes contain lies - which means those are exactly the lines an interviewer drills.

How to make a reconstructed number defensible

A defensible number is one you can explain and recalculate out loud, built from inputs you actually remember. You can quantify almost any achievement you never tracked by rebuilding the figure from four levers: how often you did the task, how much you handled, how long it took, and what changed after you got involved.

Frequency math is the workhorse. If you handled about 15 customer calls a day, five days a week, that is roughly 75 interactions weekly or about 3,900 annually. The figure is honest because it is reconstructed from a cadence you know, not invented. The risk, covered in the failure modes below, is drift: multiply a peak day by every working day and the number overstates. Quote the conservative daily rate.

Two rules govern honesty once you have a figure. Use a number you are roughly 80% confident in, and round down rather than up. If your real pace was 30 to 50, quote 30 - an underclaim you can prove beats an overclaim you cannot. Then soften it. Phrases like "approximately," "over," "up to," and a tilde signal that the figure is an estimate while still giving the reader a concrete picture. These qualifiers raise trust, they do not lower it, because they tell the reader you know exactly what kind of number this is.

MIT CAPD confirms the softer path: estimations and approximations are fine, especially early in a career, as long as you move toward real metrics over time. The goal is not false precision. It is a figure you can stand behind in a room.

Duty (as written): "Responsible for handling customer support tickets."
Frequency: about ___ per day, ___ days a week
Volume (frequency x cadence, rounded down): ~___ per year
Before/after: resolution time went from ___ to ___
Scope: covered ___ products / ___ regions / a team of ___
Rebuilt bullet: "Resolved ~[volume] support tickets a year across [scope], cutting average resolution time from [before] to [after]."

The procedure, bullet by bullet

Run every bullet through this sequence. It moves from inventory to a defended, anchored resume in about ninety minutes of focused work, plus verification time that varies.

One ordering note: the sources disagree on sequence. Some put quantification first and the honesty check last; others argue specificity and truth come before any number is forced onto a line. I favour the second. Tag and reconstruct before you polish, because forcing a number onto a bullet that does not need one is how padding creeps in.

How this goes wrong

Every proof type has a characteristic failure - the false positive that reads fine on the page and collapses in the room. This is the section to study hardest, because avoiding these is most of the value.

Scope inflates ownership. "Managed $3M budget" reads as authority, but a number borrowed from a team dashboard you did not influence implies ownership you did not have. The probe is "what was your specific decision on that budget?" Fix it by naming the slice you actually controlled.

Before/after hides the baseline. "Cut response time 50%" from an absurd starting point is technically true and genuinely misleading. The interviewer asks "what was it before, and why was it that high?" Defend with the raw numbers, not the percentage, so the baseline is visible.

Frequency math drifts. Multiplying a good day's pace by every working day overstates volume. The false positive is a peak rate treated as the average. Round down and quote the conservative daily figure.

Round percentages read as invented. A suspiciously clean 30% or 50% invites "how did you calculate that?" Honest ranges and softened figures beat fake precision every time. A multi-million-dollar impact stated as a range is more credible than a 23.7% that implies a rigor you did not have.

Named-brand proof borrows credit. "Worked with a Fortune 500 client" implies impact you may not have driven. The interviewer asks what you specifically delivered. Keep the brand, but attach your own verb and result to it.

Vague recognition is unfalsifiable. "Award-winning" without the award's name, year, and awarding body cracks on contact. If you cannot name all three, drop the claim.

Effort masquerades as impact. "Worked 50+ hours per week" is a number, but it is not an achievement. Numbers should describe impact, not effort. Cut any figure that measures how hard you worked rather than what changed.

ATS false comfort. Numbers may not lift the machine score at all, and a bullet can be fully quantified yet still skipped if the keyword and context are absent. Do not let a number stand in for the relevant skill the posting asks for.

How long you actually have, and what it means

Design the top third of each role for a fast skim, but do not treat any single second-count as law. Recruiter first-pass time spans roughly 6 to 46 seconds across studies, so the famous "7-second rule" is the low end of a wide range, not a consensus.

The spread is about 7.7x from low to high. The 7.4-second figure comes from a single firm's eye-tracking study of 30 recruiters, and it is contested; it was measured when applicant volume was high, which pushed recruiters to spend less time per resume. The honest reading: your strongest proof has to land in the top third for the fast case, but a relevant resume can earn a full minute. Both facts argue for the same move - put your two to four strongest anchors where a skim lands, and keep the rest sharp.

This is also where mirroring the reader pays off. The people who skim your resume use a specific vocabulary for the proof types in your target role, and matching it costs nothing. I built Refolk to write a resume from your own history and tailor it to each posting, which removes the manual pass of re-labelling every bullet against a new job description - the exact friction the procedure above describes.

Keeping the resume honest as the work changes

The resume is not finished when it ships; it drifts as your memory of the numbers fades and as you take on new work. Re-run the probe step before every application so no figure outlives your ability to defend it.

A note on verification sources, so you can keep this current locally. The numbers in this guide about recruiter attention and ATS scoring come from a mix of eye-tracking studies and vendor claims that disagree with each other, and I have flagged the contested ones. Re-check them the same way I did: treat any single firm's figure as one data point, look for the range across independent studies, and discount methodology that is not published. The 30-to-40% callback claim for quantified bullets, for instance, has no published methodology and should be treated as unverified.

One honest gap worth stating plainly. In Refolk's index, a keyword comparison of resume-bullet phrases returned near-zero matches - not because the phrasing does not matter, but because the index holds short professional headlines, not resume-bullet prose. So the index is excellent for seeing how peers in your target role describe themselves and for sizing the recruiter audience, and it is not a measurement of which bullet phrasing wins. Use it for the first, not the second.

Work this reference row by row and the resume stops reading as a list of claims. It becomes a set of checkable facts, each with a proof type you chose on purpose and can defend in the room.