Impact Score Explained: How nxtleap Rates Your Resume Achievements (0–100)

When nxtleap's Atomizer splits your brain-dump into individual achievement atoms, it assigns each one an Impact Score between 0 and 100. This score is not a vanity metric. It is the algorithm that decides which bullets rise to the top of your tailored resume and which get flagged for improvement before a recruiter ever sees them.

This guide explains exactly what the Impact Score measures, what the tiers mean, and the specific changes that move a bullet from a score of 34 to a score of 82.

The four signals the Impact Score measures

The Impact Score rates each achievement atom on four signals. The signals are weighted, but all four matter:

• Verb strength — Does the bullet open with a specific, active verb that names ownership? 'Led', 'Reduced', 'Shipped', 'Cut' score high. 'Responsible for', 'Helped', 'Participated in' score zero on this signal.
• Context specificity — Does the bullet name the scale, system, team size, or constraint that frames the achievement? 'An API' scores lower than 'a 3M-request/day API'. Context is what makes the same result mean different things at different companies.
• Result quantification — Is there a number attached to the outcome? A specific, credible number scores highest. A qualitative outcome ('enabling X for the first time') scores in the middle. No result at all scores lowest — and this signal has the highest weight of the four.
• Role relevance — During tailoring, how closely does this achievement match the target job description? High semantic overlap with the JD's key requirements scores higher, and the Matchmaker uses this signal to rank and rewrite the most relevant bullets first.

Quantification is the most influential signal. Bullets without any result almost never exceed 45/100, regardless of how strong the verb and context are. A bullet that opens with 'Led' and names a $2M product line will still score below 40 if the result is absent.

The three score tiers

0–40: The duty-statement tier

Bullets in this tier describe what the role required, not what you specifically accomplished. They contain weak or absent verbs, no context about scale, and no measurable result. Examples: 'Responsible for backend development.' 'Helped the team with the migration.' 'Worked on improving performance.'

These bullets may accurately describe your experience — but they give a recruiter nothing to act on. A resume built from 0–40 bullets looks identical to every other resume from someone with the same job title at a similar company. There is no signal of individual impact.

40–70: The activity tier

Bullets here have a specific verb and some context, but the result is missing or too vague to be convincing. Examples: 'Led the migration of the monolith to microservices.' (strong verb, context — but no outcome) or 'Cut API latency significantly.' (specific verb, vague result — what percentage? what was the before-state?)

These bullets are above average. They are reading correctly but leaving the recruiter with an unanswered question. The Power-Up question for a 40–70 bullet is almost always a variant of: 'What changed measurably, and can you attach a number to it?'

70–100: The evidence tier

Bullets at 70+ have a specific verb, clear context, and a credible, specific result. They are self-contained arguments. A recruiter reading one gets proof of impact in a single line — no follow-up question needed. These are the bullets nxtleap promotes to the top of your tailored resume.

Evidence-tier examples: 'Led the migration of a 6-year-old monolith to 9 microservices, cutting deploy time from 40 min to 6 and unblocking independent releases for 12 teams' (Impact Score: 88). 'Rebuilt lifecycle email into 6 behavioral tracks, lifting MQL-to-SQL conversion from 9% to 18% and adding $900K in influenced ARR' (Impact Score: 91).

A bullet moving from 28 to 84: the three stages

Here is the same achievement at three different stages of refinement, with the Impact Score at each stage.

Stage 1 — Raw dump input (Score: 28)

'Worked on making the checkout faster.'

Verb: passive ('worked on'). Context: none. Result: none. This is a raw brain-dump sentence — exactly what the Vomit Box is designed to receive. It needs all three components strengthened.

Stage 2 — After adding verb and context (Score: 53)

'Optimized the mobile checkout flow, which was the highest drop-off point in the funnel.'

Verb: specific ('Optimized'). Context: mobile, checkout, highest drop-off point. Result: still absent. nxtleap Power-Up: 'What was the cart abandonment or completion rate before you made these changes, and what did it become after?' This is the targeted question that unlocks stage 3.

Stage 3 — After Power-Up response (Score: 84)

'Redesigned the 5-step mobile checkout to 2 steps, cutting cart abandonment from 71% to 54% and lifting mobile conversion 23% over one quarter.'

Verb: owned ('Redesigned'). Context: 5-step → 2-step mobile checkout, one-quarter timeline. Result: abandonment rate with before/after numbers, conversion lift percentage. This bullet earns the top slot on a resume for any role touching product, growth, or engineering impact.

Impact Score vs. ATS keyword coverage: two different things

The Impact Score measures achievement quality — how strong the bullet is regardless of which role you are applying to. The ATS resume checker measures keyword coverage — how well your achievements map to a specific job description's language and requirements. They do different jobs, and both matter.

A bullet can score 90 on Impact Score but rank low in tailoring if it uses vocabulary the JD does not share. The Matchmaker combines both signals: Impact Score determines which bullets are strong enough to include; semantic relevance to the JD determines which ones appear first and how they are phrased.

Why nxtleap does not fabricate high scores

The Impact Score is descriptive, not prescriptive. It tells you how strong your bullet is based on what you have actually provided — it does not pad weak bullets with invented numbers to push them into the evidence tier.

Every claim on a resume has to survive an interview follow-up: 'Can you walk me through that 42% latency improvement?' A fabricated metric survives the resume screen and fails in the room. The Power-Up questions are designed to help you remember real facts, not to generate plausible-sounding ones.

To see how ACR structure and Impact Score work together, read the ACR framework guide. To learn the memory-capture method that feeds the Atomizer, see the Vomit Box method.