In the Merit Score, quantified metrics carry 13% and action verbs 11% — combined, nearly a quarter of the total, and the only slice you can improve tonight without gaining a single new skill. Both axes are computed deterministically from your bullet text, so here is exactly what the engine (and every recruiter skimming the parsed result) is looking for.
13%
of the score: bullets with numbers, %, or $
11%
of the score: strong verbs, no passive voice
24%
combined — fixable tonight, no new skills needed
What counts as quantified
The metrics axis measures the share of your bullets containing a number, a percentage, or a currency amount. Not every bullet needs one — but a resume where almost none have them reads as unverifiable claims. The usual objection is my work isn't measurable. It almost always is, along one of four axes: scale, change, frequency, or money and time.
Unverifiable
“Maintained CI pipeline for the engineering team.”
No way to tell if this was a weekend script or core infrastructure.
Quantified by scale
“Maintained CI for a 14-engineer team, ~90 builds/day.”
Scale makes the same claim checkable — and memorable.
Vague change
“Improved page load speed across the site.”
Improved by how much? From what baseline?
Measured change
“Cut p95 page load from 4.1s to 1.9s across 12 core pages.”
Before/after deltas are the strongest form of evidence.
The four axes, in jobs that don't come with dashboards
Engineering and sales examples are easy and they're also why most people conclude this advice doesn't apply to them. It does. Here's the same four axes across roles that feel unmeasurable.
| Role | Scale | Change | Frequency | Money / time |
|---|---|---|---|---|
| Teacher | 5 classes, 140 students | Pass rate 71% → 88% over two years | Weekly parent updates to 140 families | Ran the department budget of $12k |
| Nurse | 24-bed unit, 6:1 patient ratio | Cut average discharge prep from 90 to 55 min | Precepted 4 new grads a year | Reduced supply waste ~$8k/yr |
| Customer support | 60–80 tickets/week | CSAT 3.9 → 4.6 | Handled the weekend on-call rota | Deflected ~15% of tickets with new help docs |
| Admin / EA | Calendars for 3 executives | Cut expense-report turnaround from 2 weeks to 3 days | Coordinated 40+ meetings/month | Renegotiated vendor contract, saved $4k |
| Marketing | Owned 6 channels | Organic signups +38% YoY | Shipped 2 campaigns/quarter | Managed $50k quarterly spend |
| Retail manager | Store of 22 staff | Shrink 2.1% → 1.3% | Built weekly rota for 22 people | Hit $1.8M annual revenue |
Notice how few of these require a formal reporting system. Team size, headcount you supported, tickets in a week, students in a room, dollars you were trusted with — you already know all of them.
Where to find numbers you've forgotten
- 1
Old performance reviews
Managers write down your results even when you don't. This is the single richest source.
- 2
Your own calendar and inbox
Count the recurring meetings you ran, the launches you shipped, the people you onboarded.
- 3
Tools you used daily
Ticket queues, CRM dashboards, build systems and analytics all keep history you can still read.
- 4
The person who replaced you
Handover docs you wrote usually list scope in exactly the terms a resume needs.
How many bullets need numbers
Roughly half, weighted toward your most recent role. A resume where every single line carries a metric reads like a pitch deck and flattens the ones that matter — the reader stops distinguishing between "grew revenue 40%" and "attended 12 meetings". Anchor the important claims, let the supporting ones breathe.
What the verb axis flags
The action-verb axis rewards bullets that start with a strong verb and penalizes three deterministic patterns:
- Weak openers — Responsible for, Helped with, Worked on, Assisted in. These describe proximity to work, not work.
- Passive voice — was tasked with, were delivered. The subject of your own bullet should be you.
- Duty-copying — bullets lifted from your job description rather than your results.
Weak opener + duty
“Responsible for onboarding new users to the platform.”
Flagged twice: weak opener, and it describes the job, not you.
Verb first + outcome
“Built the onboarding flow used by 2,000 new users/month.”
Swap the opener, keep the truth — verb first, evidence after.
Verbs that carry weight, by what you actually did
| If you… | Use | Avoid |
|---|---|---|
| Created something new | Built, Designed, Launched, Founded, Authored | Was involved in creating |
| Improved something existing | Cut, Reduced, Accelerated, Streamlined, Rebuilt | Helped improve |
| Owned a system or area | Owned, Ran, Led, Managed, Directed | Responsible for |
| Worked across teams | Partnered, Coordinated, Negotiated, Aligned | Assisted with |
| Analyzed or found something | Diagnosed, Identified, Modeled, Forecast | Looked into |
| Taught or grew people | Trained, Mentored, Onboarded, Coached | Supported the team |
The right-hand column isn't banned vocabulary — it's vocabulary that describes being near work instead of doing it. If a bullet genuinely describes support work, say what the support produced.
Three mistakes that undo the work
Do
- Pair the number with the thing it measures: '90 builds/day', not just '90'
- Use ranges and approximations when that's the honest form: '~15%', '60–80 tickets'
- Put the outcome before the method — recruiters skim the first half of a line
- Keep percentages next to their baseline so the reader can size the win
Don't
- Quantify the trivial — '12 meetings attended' actively weakens the page
- Stack three metrics into one bullet until nothing stands out
- Give a percentage with no base ('improved efficiency by 300%')
- Invent precision you never measured, ever
Why this pays double
These are quality axes, so they're relevance-gated — they can't rescue a resume aimed at the wrong job. But on a targeted application, they're the tiebreaker: two candidates with the same keyword coverage can differ by 10+ points purely on quantification and verb strength. Same skills, different evidence.
There's a second payoff that doesn't show up in any score. Numbers on a resume become the questions you get asked in the interview — and they're questions about work you actually did, on ground you already know. A page of vague duties invites a stranger to pick the topic instead.
Run your resume through the free checker — the metrics and verb axes come back with the exact bullets that were flagged, so you know which lines to rewrite first.
Frequently asked questions
- What if I don't have access to real numbers?
- Use scope instead of outcome: team size, ticket volume, number of systems owned, release cadence, budget range. "Owned deployment for 12 services across 4 teams" is quantified and verifiable. Never invent a percentage — a number you can't defend is worse than no number.
- How many bullets should have numbers in them?
- Aim for roughly half, weighted toward your most recent role. Every bullet carrying a metric reads like a sales page and buries the ones that matter. Recruiters scan for a few concrete anchors, not a wall of percentages.
Written by
MeritSlate Team
The team building MeritSlate's deterministic scoring engine — nine evidence-anchored axes, a nine-vendor ATS parse simulation, and every point backed by receipts.