You don't have a budget for HackerRank, Codility, or a battery of personality tests. You probably don't have a recruiter either. What you have is a stack of applications, a few hours a week, and a hire that needs to go right the first time because a bad one will cost you far more than the salary alone.
The good news: most of what separates a strong candidate from a weak one shows up in plain sight, if you know where to look. You don't need paid tools to catch it. You need a consistent process and the discipline to trust specific evidence over a good vibe in the room.
Look at how they handle the application itself
Before you evaluate a single skill, evaluate behavior. How someone applies tells you how they'll work.
- Did they tailor anything? A generic, mass-blasted application usually signals a generic effort on the job itself. A candidate who references your product, your stage, or a specific problem in your job post is telling you they read past the title.
- Did they follow instructions? If you asked for a one-paragraph note on why they're interested and they sent a template cover letter, that's data, not an accident.
- Response time and clarity in early emails. You're not looking for instant replies. You're looking for whether their messages are clear, direct, and answer the actual question you asked.
None of this requires software. It requires reading closely instead of skimming for keywords.
Replace the resume with a small, real task
Resumes reward good writing, not good work. A 45 to 90 minute task tied to the actual job will tell you more than an hour of unstructured conversation.
Keep it small and specific:
- For an engineer, a scoped bug fix or a short feature spec based on a real (simplified) problem you're facing.
- For a PM, a one-page plan for how they'd prioritize a messy backlog you describe to them.
- For a designer, a critique of your actual product, not a generic portfolio review.
Pay for the time if you can, even a small stipend. It signals you respect their hours, and it filters out candidates who aren't serious. If you genuinely can't pay, keep the task under an hour and say so upfront. For a deeper walkthrough of vetting engineers specifically when you're not technical yourself, see how non-technical founders can vet engineering candidates.
Ask about decisions, not accomplishments
Most interview answers are rehearsed summaries of a resume line. You want the version underneath it.
Instead of "tell me about a time you led a project," ask:
- "Walk me through a decision on that project you'd make differently now."
- "What was the actual failure mode, and how did you find out about it?"
- "Who disagreed with you, and what happened?"
Strong candidates get more specific under this kind of pressure. Weak candidates get vaguer, retreat to generalities, or pivot to a different, safer story. That shift is one of the most reliable signals you can get for free.
Check references yourself, informally
You don't need a formal reference-checking service. A 15-minute call with a former manager or peer, arranged directly, gets you further than a checklist.
Ask open questions rather than yes/no ones:
- "What would you brief their next manager on before day one?"
- "Where did they struggle, and did it improve?"
- "Would you hire them again, and for what kind of role specifically?"
Hesitation, qualifiers, or a reference who "has to check with HR before saying anything" are signals in themselves. A genuinely strong former manager usually wants to talk.
Look for evidence of real, shipped work
For product and tech roles specifically, a portfolio of real output beats a polished interview every time.
- Engineers: actual commit history, a shipped side project, or a technical blog post that shows how they think, not just what they built.
- PMs: a case study with real metrics, even directionally, not just a slide about "driving growth."
- Designers: shipped work in production, not just concept mockups.
If a candidate can't point to anything real after several years of claimed experience, that's worth a direct, non-accusatory question: "What's something you built that's live right now?"
Watch for the gaps between confidence and specificity
The single most useful pattern-match on a startup budget: confident-but-vague answers are a red flag, and hesitant-but-specific answers are usually a green one. Strong candidates who genuinely did the work can go three levels deep on any claim they make. Weak candidates run out of detail fast and default to buzzwords.
This applies whether you're technical or not. If you're interviewing engineers without an engineering background yourself, structure your questions around depth rather than jargon; our guide on interviewing engineers when you're not an engineer walks through how to do that without faking expertise you don't have.
Don't skip this step: make sure the candidate is real
All of the above only works if you're actually spending your limited time on real applicants, not expired listings, recycled resumes, or profiles that never respond. A surprising amount of founder hiring time gets burned on candidates who ghost after the first email, or on job boards where the pipeline looks full but half the "applicants" are stale.
This is one reason WellHired verifies employers and listings on both sides of the marketplace: the goal is that the time you spend evaluating a candidate profile, task, or reference isn't wasted on someone who was never really in the market. If you're building a lean process and want to spend your limited hours on signal instead of noise, browsing verified roles and candidates on WellHired is worth a look before you sink more hours into a platform full of dead listings.
Put it together into a repeatable process
You don't need all six signals for every hire. Pick three or four that fit the role and run them the same way for every candidate in the pipeline. Consistency is what turns "gut feeling" into something you can actually compare across candidates, which matters more on a small team where one wrong hire has an outsized impact. For more on building that repeatable process without an HR team, see how to build a hiring process with no HR team, and if you want the numbers behind why this discipline pays off, the real cost of a bad hire is worth reading before your next req goes out.
FAQ
What's the single best free signal for spotting a strong candidate?
Specificity under follow-up questions. Ask a candidate to go one level deeper on any claim they make; strong candidates get more detailed, weak ones get vaguer or change the subject.
Do I need a paid assessment tool to screen technical candidates?
No. A small, scoped, real-world task tied to the actual job, reviewed by you or a trusted technical friend, catches most of what a paid assessment platform catches, at no cost beyond your time.
How long should a take-home task be for an early-stage hire?
Keep it to 45 to 90 minutes. Anything longer starts filtering out strong candidates who have other offers and won't do unpaid work for hours.
Should I still check references if I'm hiring fast?
Yes, but keep it informal and short. A 15-minute direct call with a former manager is faster than a formal reference-check service and often more revealing.
How do I avoid wasting time on candidates who aren't serious?
Use a platform with verified employers and real, current listings rather than boards full of stale postings, and look for early signals like tailored applications and clear, prompt communication before you invest interview time.