Inside the Greenhouse Recruiter Scorecard: How Tech Startups Rank Resumes

The Greenhouse recruiter scorecard isn't just a notepad; it's a structured data capture form with defined attributes that map directly to the job's core competencies. When a recruiter opens your application, they're presented with a custom panel—often built by the hiring manager—containing 3-5 scored fields (e.g., 'Technical Architecture,' 'Product Impact,' 'Ownership'). Each field is a dropdown or a 1-5 scale. Your resume text is parsed into a table format behind the scenes, and the recruiter's primary job is to scan for evidence to populate those scores. The final 'Overall Recommendation' (Strong No, No, Yes, Strong Yes) is often a weighted calculation of those individual scores, not a gut feeling.

This process is brutally efficient. In a startup environment where a single recruiter might screen 200 applications for a senior engineer role, they spend an average of 90 seconds on the initial pass. They are not reading; they are data mining. If your resume's narrative doesn't immediately surface quantifiable proof points aligned to the scorecard attributes, you default to a '3' or below—the zone of 'Maybe, if the pipeline is thin.' This is why templated, duty-based resumes fail: they provide no scannable data for the fields that matter.

Upon submission, Greenhouse extracts your resume text into an XML-like structure, segmenting it into tables for 'Experience,' 'Education,' and 'Skills.' Recruiters often use a split-screen view: your parsed resume on one side, the scorecard on the other. They perform keyword searches directly within the ATS (e.g., 'architected,' 'reduced,' 'scaled from X to Y') to rapidly validate claims. Greenhouse also allows 'knockout questions' configured at the application stage—like 'Do you have 5+ years of backend experience?'—which can auto-disqualify you before any human review.

The scoring rubric is rarely transparent. For a Staff Engineer role, a '5' in 'System Design' might require evidence of designing a system serving >10k QPS or improving p99 latency by >40%. A '1' is simply listing a known framework. Recruiters are instructed to demand production metrics. Vague statements like 'improved system performance' are ignored or scored low because they cannot be confidently tagged to a specific competency. This mechanistic tagging is what feeds downstream reports to hiring managers and determines which 10% of applicants get a phone screen.

Weak bullet: 'Responsible for improving API performance and reliability.' This is a duty, not evidence. It offers no metric for the Greenhouse scorecard attribute 'Technical Performance,' resulting in a default score of '2.'

Strong rewrite: 'Reduced p99 latency for Checkout API by 220ms (38%) through concurrent request orchestration and Go profiling, supporting 12,000 QPS during peak events.' This provides three scannable data points (220ms, 38%, 12k QPS) that a recruiter can immediately map to high scores in 'Performance Optimization' and 'Scale.' It answers the 'how' and the 'impact.'

Weak bullet: 'Managed a team to deliver new features.' This is unscoreable, giving no data for 'Leadership' or 'Product Impact.'

Strong rewrite: 'Led 4-engineer pod to launch 0-click returns feature, increasing customer repeat rate by 8% and processing 15% of all returns within 3 months of launch.' This ties leadership ('Led 4-engineer pod') to a concrete business outcome ('repeat rate +8%') and scale ('15% of all returns'), allowing high scores in multiple attributes.

Conventional wisdom says to pack your resume with every technology acronym. Inside hiring committees, we called this 'the buzzword bingo penalty.' Recruiters using Greenhouse are trained to be skeptical of keyword-stuffed resumes. If you list 30 technologies but your bullet points don't show applied depth with any of them, it lowers your 'Technical Depth' score. A concise resume demonstrating mastery of 5-6 core technologies through metrics-driven bullets always outperforms a dense list.

Another missed rule: seniority is judged by the complexity and ambiguity of the problems you solve, not your years of experience. A candidate with 8 years who only lists task execution will score lower on 'Ownership' than a candidate with 5 years who details navigating unclear product requirements, making trade-off decisions, and driving a cross-functional initiative. The scorecard rewards narrative proof of scope and decision-making, not tenure. This is why internal referrals often jump the queue—the referrer pre-populates the scorecard narrative for the recruiter.

1. Reverse-engineer the scorecard. Before applying, analyze the job description and hypothesize the 4-5 core competencies the hiring manager built into the Greenhouse scorecard (e.g., 'Scale,' 'Code Quality,' 'Product Sense'). Write these as headers on a notepad.

2. Audit every bullet. For each bullet point on your resume, ask: 'Which scorecard competency does this prove, and with what hard metric?' If a bullet doesn't map to a competency or lacks a quantifiable result, rewrite it or cut it.

3. Front-load the proof. Recruiters scan left-to-right and top-to-bottom. Place your strongest metric (%, $, latency reduction, headcount scale) at the beginning of the bullet. Example: 'Increased conversion 17% by redesigning...' not 'Redesigned the flow, which increased conversion by 17%.'

4. Prepare for knockout rules. If the application has mandatory questions (e.g., 'Salary expectations,' 'Visa sponsorship'), answer them precisely. An unrealistic salary or a 'Yes' on sponsorship where it's not offered can trigger an instant, irreversible reject in the system.

5. Test for parsing. Convert your resume to a plain .txt file. Does the formatting hold? Are your key metrics clearly separated on their own lines, or do they get jumbled? Greenhouse's parsing can mangle complex tables and text boxes.