tech
Recruitment / Hiring
Finding, evaluating, and hiring the right people. The highest-leverage decision in any business.
Definition
Recruitment is the process of finding and hiring people. For service businesses, hiring is the #1 multiplier or constraint - good hires compound, bad hires destroy. The standard sequence: define the role (with specific outcomes, not vague tasks), source candidates (referrals first, networks second, job boards third), screen (resume + short call), interview (3-5 people, structured questions), test (paid project mirroring real work), reference check, offer. Skipping the paid test step is the most common (and expensive) mistake.
The true cost of a bad hire in the US
US Department of Labor and SHRM data put the cost of a bad hire at 30 to 200 percent of annual salary depending on role seniority. For a 100K knowledge worker who underperforms for 9 months before exit, direct costs alone (salary, payroll tax, benefits) hit 90 to 110K. Indirect costs (manager time, lost productivity on dependent projects, team morale damage, customer impact, recruiting replacement) typically double that number. For founder-led US service businesses under 5M revenue, a single bad hire can wipe out a quarter of net profit. The math of investing 40 to 60 hours in hiring per role is overwhelmingly positive when the alternative is a 200K mistake.
Sourcing channels ranked by US hiring data
Channels ranked by quality of hire and cost. One, employee referrals: highest quality, lowest cost, 5x conversion to hire compared to job boards (LinkedIn data). Pay 1K to 5K referral bonuses for successful 90-day hires. Two, founder network and LinkedIn outbound: high quality, moderate effort. Three, specialized communities (Slack groups, Discord servers, Substack newsletter audiences). Four, recruiters (contingency for senior roles, 20 to 30 percent of first-year salary; retained for executive search, 33 percent plus). Five, general job boards (LinkedIn, Indeed, ZipRecruiter): high volume, lower signal, expect 1 to 3 percent conversion from applicant to hire. Six, AngelList / Wellfound for US startup talent, Lever / Greenhouse ATS for managing the process.
The paid trial as the highest-signal interview
Resumes predict performance at roughly 0.18 correlation (essentially noise). Unstructured interviews predict at 0.20. Structured interviews 0.51. Work-sample tests 0.54. A paid trial project (10 to 40 hours of real work, paid at fair market rate) outperforms every other signal because it tests the actual job. Format: scope the trial as a real project the company genuinely needs done, pay fairly (50 to 200 dollars per hour depending on role), set a 1 to 2 week timeline, evaluate against a written rubric. US labor law allows paid trials but be careful about classification: contractor versus employee distinction matters under IRS rules and state wage law (especially California AB5).
Compensation benchmarking and offer construction
US compensation benchmarks come from Levels.fyi (tech roles), Glassdoor, Salary.com, Payscale, and BLS Occupational Employment Statistics. Compensation has three layers: base salary (cash), variable (commission or bonus), equity (stock options or RSUs for startups). For US startups under 50 employees, total compensation should hit market median to attract solid candidates and top quartile for high performers. Equity grants for early employees: 0.1 to 1 percent depending on stage and role. Construct offers with a 24 to 48 hour decision window, not pressure tactics. Slow yes is better than fast maybe; founders who chase candidates with desperation often get the candidates other companies passed on.
FAQ
How long should the hiring process take?
Optimal US hiring timeline: 3 to 5 weeks from posting to offer. Faster than 2 weeks usually means under-screening; slower than 6 weeks loses top candidates to competing offers. Standard sequence: week 1 (sourcing and resume screen), week 2 (30-minute initial calls with 8 to 12 candidates), week 3 (90-minute structured interviews with 4 to 6 finalists), week 4 (paid trial project for 2 to 3 finalists), week 5 (reference checks and offer). LinkedIn data shows that the best candidates have an average of 10 days in market before accepting an offer somewhere; slow processes simply lose them.
When should I hire my first employee versus a contractor?
Hire a US W-2 employee when: the role is core to delivery, you need 30+ hours per week, you want control over how and when work happens, the role is long-term. Hire a 1099 contractor when: scope is project-based, the person serves multiple clients, you need specific expertise short-term, IRS classification rules support contractor status (control of work, financial control, relationship type). Misclassifying a worker as 1099 when they should be W-2 is a serious US federal and state tax exposure. California AB5 and similar state laws further restrict contractor classification. Consult a US employment attorney before scaling contractor use.
Should small businesses use an ATS (Applicant Tracking System)?
Once you hire more than 4 to 6 people per year, yes. Below that volume, a structured Google Sheet or Notion database is enough. Above that, ATS tools save real time and protect compliance. Best US options for SMB: Workable (99 to 599 dollars per month), Greenhouse (mid-market, custom pricing), Lever (mid-market). Solo founders hiring 1 to 2 people per year do not need this; series-A startups hiring 10 to 30 people per year absolutely do. The biggest ATS benefit beyond efficiency is compliance: structured records of why each candidate was rejected, which protects against US EEOC discrimination claims.
How do I do reference checks well?
Three rules. One, talk to managers, not peers - peers say nice things about everyone. Two, ask behavioral questions, not yes/no: 'tell me about a time X struggled,' 'on a scale of 1 to 10 how would you rate X compared to others you have managed at this level.' The 1 to 10 scale forces specificity. Three, ask the same questions you asked the candidate so you can compare narratives. Many US managers refuse to give substantive references due to legal concerns; in those cases, ask for off-record conversation (will not be quoted). A reference check that takes 15 minutes and produces nothing is a strong negative signal.
What are the most common US hiring mistakes founders make?
Five recurring mistakes. One, hiring for skills you can teach instead of judgment and ownership you cannot. Two, hiring a clone of yourself instead of someone who fills a gap. Centerpiece bias makes founders gravitate to candidates who interview well by being similar. Three, skipping the paid trial because it feels slow. Four, hiring under deadline pressure ('I need someone next week') and lowering the bar. Five, not negotiating compensation up front (US candidates expect negotiation; founders who avoid it overpay or underpay both signals). Slow down the process: a 30-day delay to find the right person is dramatically cheaper than a 9-month bad hire.
In your business
- →Define the role by outcomes (what you want them to deliver in 6 months), not by tasks
- →Use a paid test project for finalists - resume and interviews predict performance poorly
- →Reference-check before offering, not after