Most resumes read like job descriptions. “Responsible for backend services.” “Managed marketing campaigns.” “Handled customer accounts.” These are duty statements. They describe the shape of the role rather than the work the person did. Duty bullets are the single most common reason a strong candidate gets passed over by an applicant tracking system, and the second most common reason a recruiter skips them on the seven-second scan.
Impact bullets are the fix. Twenty rewrites are below, and the structure is reusable across every industry.
Why duties do not get interviews
Duty bullets fail at every layer of modern resume screening. They fail at the ATS layer because they are interchangeable across every candidate who has held the same job title; they carry no differentiating signal. They fail at the recruiter layer because eye-tracking research from Ladders, Inc. shows recruiters spend an average of 7.4 seconds on initial resume review, and quantified results catch the eye while duties blur. They fail at the hiring-manager layer because the manager is trying to predict whether the candidate can deliver, and a duty statement provides no evidence either way. The Harvard Business School research on hidden workers estimated 27 million qualified Americans get rejected by automated screening for reasons unrelated to their actual ability, and resumes written as duty lists are disproportionately represented in that rejected pool. The fix is not louder writing or more buzzwords. It is to convert every bullet into evidence of impact.
The impact formula
Every bullet should follow the same three-part structure: action verb + what you did + measurable result. The action verb signals ownership. The middle clause names the work. The measurable result anchors the bullet in something a reader can evaluate. This formula works because it satisfies all three layers of screening at once. ATS keyword extraction picks up the verbs and skills. Recruiters’ seven-second scan locks onto the numbers. Hiring managers get the evidence they need to predict performance.
If you cannot produce a measurable result for a bullet, write the bullet differently. Common substitutes when a hard number is unavailable: scope (“45-account portfolio worth $3.2M ARR”), comparison (“shortest cycle time on the team”), durability (“system still running three years later, no rewrites”), recognition (“became the team’s reference implementation”), or speed (“shipped in six weeks against a twelve-week plan”). Any of these is stronger than “responsible for.”
Twenty before-and-after rewrites
The pattern below covers the most common role categories. Each example shows the same person describing the same work, written first as a duty and then as an impact bullet.
Software engineering
- Before: Responsible for backend development and code reviews.
After: Rebuilt the payments API, reducing transaction failures 42% and saving $180K annually. - Before: Worked on the data pipeline.
After: Shipped a streaming pipeline that processes 2M events/sec at sub-second p99 latency, replacing a batch system that ran 4 hours behind. - Before: Contributed to mobile app features.
After: Owned the iOS checkout rewrite that lifted conversion 18% and cut crash rate from 0.9% to 0.05%.
Marketing
- Before: Managed social media accounts and created content.
After: Grew Instagram from 8K to 47K followers in 6 months, generating 340 qualified leads/month. - Before: Ran email campaigns.
After: Designed a lifecycle email program that lifted MQL-to-SQL conversion 31% and added $640K in pipeline per quarter. - Before: Worked on SEO.
After: Took the blog from 12K to 180K monthly organic visits in 18 months by rewriting 40 high-intent pages and shipping internal-link automation.
Sales
- Before: Responsible for meeting sales targets.
After: Closed $2.4M in new ARR across 18 enterprise accounts, exceeding quota by 135%. - Before: Managed customer relationships.
After: Owned 22-account book worth $4.1M ARR, retained at 118% NRR, churned zero in two years. - Before: Used Salesforce.
After: Rebuilt the territory’s pipeline hygiene in Salesforce, lifting forecast accuracy from 62% to 91% in two quarters.
Project management
- Before: Led cross-functional teams on multiple projects.
After: Delivered three product launches on time and under budget, coordinating 24 engineers across 4 timezones. - Before: Tracked project status.
After: Reduced average project cycle time from 14 weeks to 9 by introducing weekly checkpoint reviews and a single source-of-truth status board.
Customer success
- Before: Handled customer accounts and resolved issues.
After: Managed a 45-account portfolio worth $3.2M ARR with 96% retention and 1.2-day median time-to-resolve. - Before: Conducted onboarding sessions.
After: Designed an onboarding curriculum that cut time-to-first-value from 22 days to 6 days and lifted 90-day activation by 41%.
Operations / Finance
- Before: Reviewed monthly reports.
After: Identified $1.1M in recurring vendor overbilling by rebuilding the monthly close audit; recovered 12 months of overpayment in one quarter. - Before: Worked on process improvement.
After: Cut weekly close from 7 days to 2 by automating reconciliations across three payment processors.
Data / Analytics
- Before: Built dashboards.
After: Replaced 14 manual reporting workflows with one self-service Looker model used by 90+ stakeholders weekly. - Before: Analyzed customer data.
After: Identified a churn signal four weeks earlier than existing alerts; the new model retained $780K in ARR in its first six months.
Design
- Before: Designed product features.
After: Led the redesign of onboarding, lifting day-7 activation 27% and cutting support tickets per new user by 38%. - Before: Worked on the design system.
After: Shipped a component library that reduced engineer-design rework time per ticket by 40% and unlocked two designers from production work.
HR / Recruiting
- Before: Managed the hiring pipeline.
After: Filled 38 roles in 12 months at a 4.2-week median time-to-fill, with 89% offer acceptance and 94% one-year retention.
Why outcome bullets matter even more under AI screening
The shift from keyword-based ATS to outcome-based screening makes impact bullets matter more, not less. A keyword-matching ATS only cared whether “Python” appeared on your resume. An outcome-based model reads the bullet itself and ranks on what it finds. That means impact bullets are now the input the ranker uses to decide if you should be in the top ten.
This is exactly the problem CurriculoATS was built to solve. The model reads resumes for measurable outcomes, revenue generated, teams scaled, systems shipped, problems solved, and produces a 0–100 composite score with a written reasoning paragraph that cites the specific impacts it found. A duty-list resume scores poorly because there is nothing to find. An impact-bullet resume scores well because the work is legible to the model. See the impact scoring page for the full mechanism.
Before CurriculoATS, our founder Dev spent years at Amazon working on search and recommendation systems. The single principle that drove the design here is the same: rank on signals the input cannot fabricate cheaply. Keywords are cheap; anyone can paste them. Outcomes are not; the candidate has to have actually shipped the work.
How to rewrite your bullets in one sitting
Most candidates can convert their entire resume in 60–90 minutes if they follow the same loop on every bullet.
- Read the bullet aloud. If it could appear on the resume of any other person who held the same title, it is a duty. Mark it.
- Ask: what changed because I did this? Revenue, time, cost, scope, quality, retention, conversion, error rate. The change is the impact.
- Add a number. If you do not have an exact number, use scope (“$3M portfolio”), comparison (“vs. team average”), or speed (“in six weeks vs. the planned twelve”).
- Cut everything else. The bullet does not need adjectives. Action verb, the work, the result. Stop.
- Test against the seven-second scan. Read the bullet for one second and check whether the number is the thing you remember. If not, the bullet is structured wrong.
Founder questions
What if I do not have hard numbers from my last role?
Use scope and comparison instead. “Owned the company’s first mobile app” is a strong systems signal. “Cut feature delivery cycle from 9 weeks to 4 against the team baseline” works without absolute numbers. Hard numbers are best, but credible relative measures are far better than duty statements.
How many bullets should each role have?
Three to six per role for the last two roles, two to four for older roles, and one or two for anything more than a decade back. Most resumes are too dense; the seven-second scan rewards space and hierarchy, not volume.
Will impact bullets actually help me get past the ATS?
It depends on which ATS you are facing. A keyword-matching system rewards token presence regardless of bullet quality. An outcome-based system reads what the bullet actually says, which means impact bullets are the difference between a top-ten rank and a bottom-half rank.
Should I use the same bullets for every job application?
The bullets should not change much; the work is what it is. The framing can change. Lead with the bullets most relevant to the role you are applying for, and sometimes rewrite the verb to mirror the language of the JD if it is more accurate. Do not pad with keyword stuffing.
What does CurriculoATS read on a resume?
The model reads bullets, not just keywords. It extracts revenue, teams, systems, and problems signals and produces a written reasoning paragraph for every candidate. Read the resume screening page for the full mechanism.
What to do next
If your resume is mostly duties right now, set aside an hour and run the five-step rewrite on every bullet. Then read the CurriculoATS glossary for the canonical definitions of impact signals, and the impact scoring page for what an outcome-based ATS reads when your resume comes through. The bullets you write tonight are the input every screening system uses tomorrow.