Hired someone for $8K/month last year because they "had experience."
Great resume. Said all the right things. Confident in interviews.
3 months later: nothing shipped. Excuses. Meetings about meetings. Always "almost done."
$24K and a quarter of momentum—gone.
Editorial update, October 11, 2026: $8K a month for three months equals $24K before other costs. Payroll, deliverable and evaluation records are unavailable here; the original three team success examples remain personal accounts. Repeated experience taxonomies, red/green-flag lists and universal firing/action rules were condensed. Original section links remain: The Resume Delusion; What Experience Actually Means; Red Flags I Now Recognize; Green Flags That Actually Matter; What Experience Actually Predicts; The 3-Week Rule; What I Look for Now; The Uncomfortable Truth About Experience; The Resume vs Reality Test; What This Means for Founders; Action Steps for Better Hiring; Why This Matters.
I now want a bounded, paid work sample as well as interviews and experience. The task needs to reflect the job, with clear expectations and feedback.
When work is not meeting expectations, I need to address that promptly and follow the applicable employment and contract obligations. A calendar deadline alone does not justify ending someone’s employment.
That hire made me want direct evidence of relevant work, alongside experience and references.
The Interview Theater
Their interview performance:
Technical Questions: Answered everything perfectly. Used the right buzzwords. Explained complex concepts clearly.
Experience Stories: Had impressive examples from previous companies. Talked about "leading initiatives" and "driving outcomes."
Cultural Fit: Said exactly what I wanted to hear about working style, communication, and autonomy.
References: Glowing reviews from former managers and colleagues.
My conclusion: This is exactly what we need.
The reality: The interview did not predict the work output I needed in that role.
The 3-Month Nightmare
Week 1-2: The Honeymoon
- Lots of questions (seemed engaged)
- Setting up development environment
- "Getting familiar with the codebase"
- Weekly check-ins going well
Week 3-6: The Red Flags
- First deliverable pushed back "for quality"
- Meetings to discuss meetings
- "Almost ready" became the default status
- Simple tasks taking weeks, not days
Week 7-12: The Excuses
- "The codebase is more complex than expected"
- "Need to refactor before adding features"
- "Previous team made questionable decisions"
- "Setting up proper testing frameworks first"
The output after 3 months:
- Zero features shipped to customers
- One internal tool half-finished
- Extensive documentation about what needed to be done
- Multiple proposals for "better approaches"
What our $3K/month contractor in India shipped in the same period:
- 4 major features
- 12 bug fixes
- 2 integration improvements
- Consistent weekly progress
The math was humiliating.
The Sunk Cost Trap
Why I kept them for 3 months:
Month 1: "They're still ramping up. This is normal."
Month 2: "We've invested so much in onboarding. They'll turn it around."
Month 3: "They're almost done with that big feature. Let's see it through."
The psychology: I didn't want to admit I'd made a $24K mistake.
The reality: The quoted $8K monthly rate is about $1,846 a week if annualized as $8,000×12÷52. That is a scenario calculation, not an actual payroll-period measurement, and it excludes other costs.
The lesson: I needed to evaluate the situation using current evidence, rather than justify delay with money already spent.
The New Hiring Framework
Old approach: Interview and Hope
- Post job description
- Review resumes
- Conduct interviews
- Check references
- Make offer
- Pray it works out
Paid, job-relevant work sample
Editorial update, October 11, 2026: The $500–$1,000, one-to-two-week project was the author’s reported practice, not a current recommended wage or legal safe harbour. Pay, worker classification, confidentiality and termination duties depend on the work and jurisdiction. Ontario’s employee-status guidance explains that work during a trial may be employment; the OHRC hiring guidance addresses accommodation in testing. A useful sample needs job-related criteria, comparable conditions, appropriate accommodation and authorized staging data—not a covert social-affinity test.
- Post simple project description
- Ask for work samples (actual code, not talks about code)
- Give paid trial project (1-2 weeks, $500-1000)
- Evaluate based on output, not words
- Hire only if trial exceeds expectations
The difference: A work sample can show how someone handles that particular task. It does not prove how they will perform in every future role or project.
The 2-Week Trial System
How it works:
Week 1: Simple Feature
- Give them a well-defined, small feature
- Provide access to staging environment
- Set clear expectations and deadline
- Minimal guidance (see how they work independently)
Week 2: Customer Feedback Integration
- Share real customer feedback
- Ask them to propose and implement solution
- See how they prioritize and communicate
- Evaluate their customer empathy
What I'm evaluating:
- Output quality: Does it work? Is code clean?
- Communication: Do they ask good questions? Update proactively?
- Speed: Do they ship on time or make excuses?
- Initiative: Do they suggest improvements or just follow orders?
- Customer focus: Do they understand why features matter?
Pass rate: I estimated that about 30% passed my trials. The candidate count, rubric and evaluation records needed to verify that rate are unavailable here.
The India Contractor Revelation
While the $8K/month "experienced" developer produced nothing:
My $3K/month contractor in Bangalore:
- Shipped 4 features in 12 weeks
- Communicated daily via Slack
- Fixed bugs within hours of reporting
- Suggested useful product improvements
- Never missed a deadline
The skill difference: Both could code, but the work I observed in this period differed. I did not run a controlled skill comparison.
The mindset difference: I was more satisfied with the contractor’s communication and shipping during that period. I cannot establish either person’s motives from those outputs.
The cost difference: The monthly quoted rates differed by $5K. Four features shipped versus zero does not produce a finite 10× output ratio, and feature counts alone do not measure equal scope or quality.
The lesson: For me, a paid sample provided useful evidence that the resume had not. That does not establish that global hiring or trial projects beat every other process.
Real Examples: Trial Success Stories
Current team members who passed trials:
Developer #1 (2 years experience):
- Trial: Built customer dashboard feature in 1 week
- Red flag: Asked a lot of questions initially
- Green flag: Shipped working solution ahead of schedule
- 6 months later: Our most productive team member
Developer #2 (4 years experience):
- Trial: Integrated payment system in 2 weeks
- Red flag: No experience with our tech stack
- Green flag: Learned it during trial and delivered
- 8 months later: Promoted to lead developer
Developer #3 (1 year experience):
- Trial: Built mobile-responsive feature
- Red flag: Fresh graduate, minimal experience
- Green flag: Code quality exceeded expectations
- 4 months later: Handling complex projects independently
Common traits: These are my reported team examples, not a controlled study showing that experience has no predictive value.
The $24K Lesson
What I learned from wasting $24,000:
This hire’s experience did not translate into the output I needed.
A resume describes experience; I also wanted evidence of relevant work.
An interview can assess some skills, but ours did not reveal the mismatch.
Meeting confidence was not enough for me to assess this role.
Recent relevant work is useful evidence, not a guarantee of future performance.
I need to revisit a poor fit promptly, with fair feedback and the appropriate process.
Most importantly: I want to compare what people say with what they do on work relevant to the role.