Resume Best Practices (2025/2026)

The complete, no-BS guide to writing resumes that get interviews.
Plus a Claude Code skill that optimizes your resume with AI.

---

Who Is This For?

• Software engineers updating their resume for a job search
• Career changers breaking into tech
• PMs, TPMs, and startup generalists targeting $150K+ roles
• New grads writing their first professional resume
• Anyone who wants AI-assisted resume optimization via Claude Code

What's Inside

The Guide -- Everything below covers structure, formatting, ATS optimization, bullet writing (Google's XYZ formula), quantification tricks, print-to-PDF CSS specs, role-specific keyword lists, and before/after examples. Sourced from recruiter advice, eye-tracking research, Cracking the PM Interview, and ATS vendor documentation.

The Skill -- A Claude Code slash command that reads your resume, scores it, rewrites weak bullets, and matches keywords against a job posting. Jump to setup.

---

Claude Code Skill: /optimize-resume

This repo includes a Claude Code skill that automatically analyzes and optimizes your resume. It scores ATS compatibility, rewrites weak bullets using the XYZ formula, checks quantification, and matches keywords against a job posting.

Setup

Clone this repo and the skill is automatically available in Claude Code:

git clone https://github.com/StephanieKoehl/resume-best-practices.git
cd resume-best-practices

Or copy just the skill into any existing project:

cp -r resume-best-practices/.claude/skills/optimize-resume your-project/.claude/skills/

Usage

Inside Claude Code, run:

/optimize-resume path/to/resume.md

Or with a job posting to tailor against:

/optimize-resume path/to/resume.md path/to/job-posting.txt

Claude will return:

• Scorecard (ATS compatibility, bullet impact, quantification, keyword match)
• Top 3 priority fixes with before/after examples
• Line-by-line rewrite suggestions using the XYZ formula
• Missing keywords from the job description

---

Table of Contents

1. The 7-Second Rule
2. Resume Structure
3. ATS Optimization
4. Writing Bullets That Hit
5. Quantifying Impact
6. Before / After Examples
7. Print-to-PDF Specs
8. Handling Tricky Situations
9. Role-Specific Advice
10. What's In and What's Out in 2025/2026
11. Resources
12. Contributing

---

The 7-Second Rule

Eye-tracking research (Ladders study) shows recruiters spend 7.4 seconds on an initial resume scan. In that scan, they look at:

1. Name
2. Current title and company
3. Current role dates
4. Previous title and company
5. Previous role dates
6. Education

That's it. They are not reading your bullets on the first pass. Your resume's visual hierarchy needs to make these six things instantly scannable.

What gets you into the "yes" pile (second pass)

• Most recent role bullets (read most carefully)
• Summary/headline (if concise)
• Skills section (quick keyword scan)
• Career trajectory (moving up, lateral, or down?)

What gets you instantly rejected

1. Typos and grammatical errors (#1 reason)
2. Irrelevant experience dominating the page
3. No quantified achievements -- bullets that read like job descriptions
4. Unexplained gaps
5. Wall of text -- no white space
6. Generic/unfocused resume (not tailored to the role)
7. More than 2 pages for a PM role
8. Objective statements (use a summary or nothing)
9. "References available upon request" (wastes space)
10. Personal pronouns ("I managed..." should be "Managed...")

---

Resume Structure

Length

One page is the gold standard. Cracking the PM Interview and Google's own recruiters agree: one page per ten years of experience. Even senior candidates benefit from the constraint -- it forces prioritization, which is itself a PM skill.

Layout

Single column is strongly recommended. Two-column layouts cause problems with:

• ATS parsing (many systems read left-to-right across rows and scramble content)
• Screen readers and accessibility
• Printing fidelity

Section Order

For PM/TPM/startup roles with 5+ years of experience:

1. Header -- Name, location, LinkedIn, website, email (phone in print only)
2. Summary -- 2-3 lines, tailored per application
3. Experience -- reverse chronological, most recent first
4. Skills -- compact, grouped by category
5. Education -- degree, institution, GPA (if > 3.5 and recent)
6. Highlights / Leadership -- notable achievements, fellowships, publications

Bullets Per Role

Role recency	Bullets
Most recent / most relevant	3-4
Mid-tier roles	2-3
Older or less relevant	1-2

Never more than 5-6 bullets on any single role. If you need more detail, save it for the interview.

Formatting Specs

Element	Spec
Name	20-28pt, semi-bold or bold
Section headings	11-13pt, bold, uppercase, letter-spacing 1-2px
Job titles	10-11pt, bold
Company / dates	10-11pt, regular weight, dates right-aligned
Bullet text	9.5-10.5pt (never below 9pt), line-height 1.3-1.5
Margins	0.5in to 0.75in on all sides
Section spacing	16-24px between sections
Item spacing	8-12px between job entries
Bullet spacing	2-4px between bullets

Font Choices (ATS-safe and print-friendly)

Best: Inter, Calibri, Source Sans Pro, Garamond, Cambria

Safe serif: Georgia, Palatino, Charter

Safe sans-serif: Helvetica Neue, Arial (fallback)

Inter has emerged as the go-to modern sans-serif for resumes -- designed for screens but prints beautifully.

Color

• Monochrome is safest. Use #1a1a1a for text (not pure black #000)
• One accent color maximum -- for name or section headings only
• Never use color to convey meaning (ATS can't see it, B&W prints lose it)

---

ATS Optimization

How ATS Works

Major platforms (Greenhouse, Lever, Workday, iCIMS, Ashby) parse resumes by:

1. Extracting text from the document
2. Identifying sections (experience, education, skills)
3. Matching keywords against the job description
4. Scoring and ranking candidates

File Format

PDF is widely accepted and safest for preserving formatting. The old "DOCX only" advice is largely outdated. Just ensure the text is selectable (not a scanned image).

What Breaks ATS Parsing

Don't	Why
Tables	Content read out of order or dropped
Multi-column layouts	Text from columns gets interleaved
Headers and footers	Often ignored entirely
Text boxes	Floating boxes frequently missed
Images, logos, icons	Invisible to ATS
Skill bars / progress circles	Convey zero information to parsers
Fancy characters (em-dashes, smart quotes)	Can render as garbage
Non-standard section names	ATS looks for "Experience," "Education," "Skills"

Keyword Strategy

This is the single highest-leverage ATS optimization:

1. Mirror the job description language exactly. If the JD says "cross-functional collaboration," use that exact phrase.
2. Include both acronyms and spelled-out versions: "Technical Program Manager (TPM)"
3. Place keywords in context, not in a keyword dump. ATS increasingly penalizes stuffing.
4. Hard skills matter more than soft skills for ATS scoring.
5. Match the seniority language. If the JD says "led" and "managed," use those verbs.

The Variant Resume Strategy

The highest-impact ATS strategy: maintain a master resume with all your experience and bullets, then create tailored variants for each application that:

• Select the most relevant summary
• Choose bullets tagged with matching skills
• Reorder skills to front-load JD keywords

---

Writing Bullets That Hit

Google's XYZ Formula (Gold Standard)

"Accomplished [X] as measured by [Y], by doing [Z]"

This mirrors how PMs are evaluated: by outcomes, not activities. It forces you to lead with impact.

The Principle

Lead with the result, not the activity.

• Bad: "Managed the product roadmap and worked with engineers"
• Good: "Shipped [outcome] by [action], resulting in [metric]"

Power Verbs (Ranked by Impact for PM/TPM Roles)

Highest (leadership + results): Led, Directed, Built, Designed, Launched, Drove, Scaled, Delivered, Established, Shipped

High (execution + initiative): Architected, Developed, Implemented, Automated, Deployed, Created, Orchestrated

Medium (analysis + influence): Identified, Analyzed, Synthesized, Discovered, Reframed, Facilitated

Kill these words: Helped, Assisted, Participated in, Was responsible for, Worked on, Supported

Common Mistakes

1. Leading with responsibilities, not results -- "Responsible for product roadmap" vs "Defined product roadmap that delivered 4 key features, increasing retention by 15%"
2. Vague impact -- "Improved processes" (which processes? by how much?)
3. Activity without outcome -- "Conducted user research" (and then what?)
4. Listing tools instead of outcomes -- "Used Jira, Confluence, Figma" vs "Built analytics dashboard that replaced manual weekly reporting"
5. Too long -- Each bullet should be 1-2 lines max. If it takes 3 lines, split or cut it.
6. Not tailoring to the role -- Your PM resume should emphasize different bullets than your TPM resume.
7. Using first person -- No "I" statements. Start with strong action verbs.

---

Quantifying Impact

Almost everything is quantifiable if you reframe it. This is critical for ops, support, research, and early-stage roles.

Time-Based Metrics

• "Reduced [process] from X hours to Y hours"
• "Accelerated [timeline] by X weeks"
• "Saved team ~X hours/week by automating [process]"

Efficiency / Process Metrics

• "Cut [error rate / rework] by X%"
• "Reduced [ticket volume / escalations] by X%"
• "Improved [response time] from X to Y"

Scale / Scope Metrics

• Use ranges: "Managed program across 3-5 teams"
• Use relative terms: "Largest client engagement in company history"
• Use "X+" format: "Conducted 30+ user interviews"
• Use frequency: "Shipped bi-weekly releases for 8 months straight"

Adoption / Satisfaction

• "Achieved X% adoption rate within [timeframe]"
• "Maintained X+ NPS / CSAT score"
• "Reduced churn by X% in [segment]"

Research / Discovery

• "Synthesized insights from X interviews into Y actionable recommendations"
• "Identified X critical usability issues, all resolved before launch"

The Estimation Hack

Estimate conservatively and use "~" or "approximately." Recruiters know early-stage companies don't have perfect analytics.

• "~30% reduction" is credible
• "Grew from a handful of users to 500+" is fine
• "Saved approximately 20 hours/week" works

A reasonable estimate is far better than no number at all.

---

Before / After Examples

Weak vs. Strong Bullets

Support / Ops:

BEFORE: Managed customer support inbox and resolved tickets
AFTER:  Reduced support response time from 10-30 days to same-day by triaging 
        a 112-conversation backlog, maintaining inbox zero across 22 sessions

Product:

BEFORE: Conducted user research and shared findings with the team  
AFTER:  Surfaced a critical launch blocker through 13 user research calls — 0/3 
        external users could complete the core workflow — shifting product 
        priorities before public launch

Engineering / Automation:

BEFORE: Built an automation tool for the support team  
AFTER:  Eliminated manual support triage by building an AI-powered system 
        (TypeScript + Intercom API + Claude) that classifies and drafts 
        responses autonomously 3x daily

Leadership / Scaling:

BEFORE: Helped grow the company from a small team to a larger one
AFTER:  Scaled operational infrastructure from 15 to 80+ staff, building hiring 
        processes from scratch and resolving supply chain disruptions that 
        threatened production timelines

Founder:

BEFORE: Started a company and raised funding  
AFTER:  Raised $1M+ as a solo founder at 23, recruited a cross-functional team 
        of engineers, pilots, and designers, and negotiated partnerships with 
        Tesla, Siemens, and Rolls Royce before exiting in 2021

Weak vs. Strong Summary

BEFORE: Experienced professional with a background in technology and leadership. 
        Passionate about building products and working with teams.

AFTER:  Technical founder turned product leader who ships in complex, ambiguous 
        environments. Raised $1M+ solo, scaled teams from 1 to 80+, and built 
        partnerships with Tesla, Siemens, and Rolls Royce. MSc in Computer 
        Science with a track record across AI, biotech, aerospace, and neurotech.

---

Print-to-PDF Specs

If you're building a web-based resume (HTML/CSS) that prints to PDF:

Page Setup

@page {
  size: letter; /* or A4 for international */
  margin: 0.5in 0.6in;
}

@media print {
  body {
    font-size: 10pt;
    line-height: 1.35;
    color: #1a1a1a;
    -webkit-print-color-adjust: exact;
    print-color-adjust: exact;
  }
}

Page Break Control

/* Never break inside a job entry */
article {
  break-inside: avoid;
  page-break-inside: avoid;
}

/* Prevent orphaned headings */
h2, h3 {
  break-after: avoid;
  page-break-after: avoid;
}

What Prints Well vs. What Doesn't

Prints well: Solid borders, background colors (with print-color-adjust), SVG icons, system fonts, pt units, high-contrast text

Avoid for print: Box shadows, gradients, very thin fonts (< 300 weight), transparency, complex CSS Grid (use Flexbox), rem/em units, background images, web fonts without embedding

Semantic HTML for ATS + Accessibility

<main aria-label="Resume of Your Name">
  <header>
    <h1>Your Name</h1>
    <address aria-label="Contact information">
      email | phone | location | linkedin
    </address>
  </header>
  <section aria-label="Professional Experience">
    <h2>Experience</h2>
    <article aria-label="Role at Company">
      <h3>Role Title</h3>
      <time datetime="2023-01">Jan 2023</time>
      <ul>
        <li>Achievement bullet</li>
      </ul>
    </article>
  </section>
</main>

---

Handling Tricky Situations

Short-Tenure Roles (< 6 months)

Label it as a contract or engagement. The word "contract" immediately signals it was time-bounded by design, not a quit/fire.

• Use "Contract" or "Engagement" in the title
• Use month format ("Feb 2026 -- Mar 2026"), not "5 weeks"
• Make the bullet so impressive the short duration becomes a feature
• Don't omit it -- it fills gaps and the work may be relevant

Career Gaps

• Brief parenthetical if helpful: "(Career break for [reason])"
• Or just let the dates speak -- most recruiters understand gaps now
• Never over-explain on the resume. Save it for the interview if asked.

Diverse / Non-Linear Background

This is a strength, not a weakness. But it needs explicit framing.

In the summary: State the throughline. Don't make the reader connect the dots. Example: "Product leader who has shipped across enterprise HR, biotech, AI developer tools, and neurotech -- specializing in translating complex technical domains into products users love."

In each role: Use consistent language that reinforces the thread.

Find your throughline:

• "Taking complex, technical products to market"
• "Building 0-to-1 in ambiguous environments"
• "Bridging technical and business stakeholders"

---

Role-Specific Advice

Technical Product Manager (TPM)

Emphasize: Technical depth + product judgment + cross-functional leadership

Keywords: Technical requirements, system design, API design, engineering trade-offs, technical roadmap, architecture decisions, developer experience

Differentiator: Show you can build AND lead. Most TPMs can't code -- if you can, that's a major signal.

Product Manager (PM)

Emphasize: User empathy + business impact + stakeholder management

Keywords: User research, product-market fit, OKRs, roadmap prioritization, go-to-market, customer discovery, A/B testing

Bullet pattern: "Defined and launched [feature] that drove [X]% increase in [metric], collaborating with engineering, design, and data teams"

AI Product Roles

Emphasize: Understanding of AI-specific challenges (data quality, model evaluation, user trust, iteration cycles)

Keywords: LLM, AI agents, prompt engineering, RAG, model evaluation, human-in-the-loop, responsible AI, generative AI

Bullet pattern: "Shipped AI [feature] to [X] users, designing human-in-the-loop workflows that maintained 95%+ accuracy while reducing manual review by 70%"

Startup Generalist

Emphasize: Breadth, speed, ownership of ambiguous problems

Keywords: 0-to-1, founded, scaled, full-stack, end-to-end ownership, wore multiple hats

Bullet pattern: "Wore 5 hats (product, ops, hiring, support, QA) as employee #3, scaling from 0 to [X] customers in [Y] months"

Critical Keywords for 2025/2026

AI/ML: LLM, Large Language Models, AI agents, prompt engineering, RAG, generative AI, model evaluation, AI safety, human-in-the-loop

Technical PM: System design, API, microservices, data modeling, developer experience (DX), CI/CD, observability

Product: Product-led growth (PLG), activation, retention, product analytics, discovery, dual-track agile, Jobs-to-be-Done (JTBD)

---

What's In and What's Out

In (2025/2026)

• Minimalist single-column layout
• Subtle typographic hierarchy (weight and size, not color)
• Tailored resumes per role (master + variants)
• Brief 2-3 line summary as value proposition
• Links to portfolio, GitHub, LinkedIn (as plain text URLs)
• AI-optimized keyword density
• "Key Achievements" highlight section

Out

• Skill bars / proficiency charts (universally mocked by recruiters)
• Two-column layouts (ATS problems)
• Infographic resumes (fine for portfolios, not applications)
• Photos (US/Canada/UK -- still standard in parts of Europe/Asia)
• QR codes (gimmicky for digital submissions)
• "Creative" formatting for non-design roles
• Objective statements
• "References available upon request"

---

Resources

Books

• Cracking the PM Interview -- Gayle Laakmann McDowell & Jackie Bavaro
• Cracking the Coding Interview -- Gayle Laakmann McDowell (resume chapter)
• Decode and Conquer -- Lewis Lin (PM interview prep)

Open-Source Resume Templates

• jsonresume/resume-schema -- JSON-based resume standard with dozens of themes
• posquit0/Awesome-CV -- LaTeX-based, beautiful spacing and typography (44k+ stars)
• WebPraktikos/universal-resume -- HTML/CSS with Tailwind, print-optimized
• sneas/cv-template -- Minimalist HTML/CSS designed for Chrome print-to-PDF

Key Principles

1. One page. No exceptions for < 10 years.
2. Lead every bullet with a quantified result.
3. Mirror the job description's exact language.
4. Single column, clean formatting, ATS-safe.
5. Your resume should be scannable in 7 seconds.
6. Tailor for every application. A master + variant system is the way.

---

Contributing

See CONTRIBUTING.md for guidelines. PRs welcome for:

• New resume advice or corrections
• Before/after examples
• Role-specific keyword lists
• Improvements to the Claude Code skill
• Translations

---

Compiled April 2026. Sources include Google recruiter guidance, FAANG hiring manager surveys, Ladders eye-tracking research, ATS vendor documentation, and career coaching literature.