How to Hire Your First 5 People Without an HR Team (And Without Wasting 3 Months Doing It)

Start With One Decision That Matters More Than Any Job Description

Before writing a single word of a job posting, answer this: what does success look like at 90 days?

Not “they will be a great engineer” or “they will own marketing.” Something specific. “They will have shipped the API redesign and fixed the two critical bugs currently blocking enterprise onboarding.” “They will have run one full acquisition campaign and brought in at least 40 qualified leads with a CAC under 80 euros.”

This forces clarity you probably do not have yet. Most bad hires in early-stage startups trace back to a vague brief. The founder had a rough idea of what they needed, wrote a generic job description, hired someone who looked the part, and discovered 60 days in that the fit was completely off. You can avoid most of this by getting brutal about what you actually need before you start.

Once you have the 90-day outcome, writing the job description takes 30 minutes. Everything else follows from it.

The Job Description Is Doing More Work Than You Think

Your job posting is not just an announcement. It is your first filter. A well-written JD should automatically disqualify 60% of wrong-fit applicants without you having to read a single CV.

Bad JDs list requirements. Good JDs describe the problem.

Compare these two:

Version A: “3+ years of experience in growth marketing, strong analytical skills, familiarity with paid social and email automation.”

Version B: “We currently get 400 site visits a day and convert about 1.2% of them. We do not fully understand why. The person in this role will own figuring that out and building the acquisition system from scratch. No budget yet for a team or agency. You would be the first hire on this.”

Version B tells candidates what the actual job is. That scares off people who want a comfortable role with a clear playbook. It attracts people who find ambiguity motivating. That is exactly the filter you want at this stage.

Specificity in the JD also helps you screen faster. When someone applies who only has experience running large-team campaigns with agency support, you know immediately. You do not need to interview them to find out.

Screening 200 Resumes Is Not a Badge of Effort

If you are reading every application manually, something is wrong upstream. Either your sourcing is too broad, your JD is too vague, or you have not defined what “wrong” looks like clearly enough to filter quickly.

Here is a faster approach: define two or three automatic disqualifiers before the CVs arrive. These are conditions where, if present, you would never move forward, regardless of anything else in the profile. Write them down. When you are screening, check for those first. Most resumes can be assessed in 45 seconds this way.

Then define two or three must-haves. Not “would be nice” signals, but genuine requirements for the role to work. If those are not there, the CV goes in the no pile.

You will end up reviewing maybe 20–30 candidates carefully for every 150–200 who apply. That is manageable. Reviewing all 200 with equal attention is not.

Some teams now use AI scoring tools to do this first-pass automatically, ranking applicants against a defined rubric and surfacing only the top 15–20% for human review. That only works if your rubric is well-defined. Garbage in, garbage out. But if you have done the work above (90-day outcome, specific JD, clear disqualifiers), the rubric practically writes itself.

Interviews Without a Process Are Conversations. Conversations Do Not Predict Hiring Outcomes.

Most founder-led interviews go like this: you talk to someone for 45 minutes, cover a mix of their background and some things you found interesting, improvise a few hypothetical questions, and come away with a gut feeling.

That gut feeling correlates poorly with actual job performance. It correlates well with how similar the candidate is to you, how articulate they are in an interview setting, and whether the conversation was enjoyable. None of those are the job.

You do not need an elaborate interview process. But you need a consistent one. Here is a structure that works for small teams:

• Screening call (20 min): Confirm the basics. Are they actually interested in this specific role? Do they understand what the company does and where it is? Can they speak clearly about their relevant experience? This is not a culture vibe check. It is a logistics filter.
• Structured interview (60 min): Ask the same 4–5 questions to every candidate in the same order. At least two of those questions should be about a specific past situation, not a hypothetical. "Tell me about a time when a project you led went sideways. What happened, what did you do, how did it turn out?" is more useful than "How do you handle ambiguity?"
• Practical exercise (relevant to role): One focused task that takes 1–2 hours maximum. Not a week-long project for free. A task designed to show you exactly the skill you most need to see. If you are hiring an engineer, a debugging exercise on a real bug. If you are hiring a content person, rewrite one underperforming piece of your existing content.
• Final conversation with a reference (your call): Call one reference yourself, not a written one. Ask: "What does this person struggle with?" The quality of the answer tells you more than anything they say about strengths.

What to Skip

When you do not have an HR team, you cannot do all the things a company with one does. Here is what you can safely deprioritize at this stage:

• Formal onboarding programs. A good doc with the ten things they need to know in week one is enough. You will iterate.
• Competency frameworks. You do not need a leveling rubric for 8 people. You need clarity on what the hire needs to do in 90 days.
• Panel interviews. Every additional person in an interview adds coordination overhead and dilutes accountability for the decision. Two interviewers maximum. One is fine for early roles.
• Lengthy assessment centers. Multi-day processes are for companies that have enough brand equity to ask candidates to invest that time. You do not yet. Respect their time.

The Honest Trade-offs

Hiring fast and lean means accepting some things:

You will make hiring mistakes. Probably one in five or six, at this stage. That is not a failure of process. It is the reality of early-stage hiring, where the role often changes faster than any process can adapt to.

Skipping formal process saves time upfront but costs it on the back end if you do not document your rubric. Write down why you hired each person and what you were betting on. When it works or does not, you will have something to learn from.

Referral hiring (asking your network, asking early employees to recommend people) is underused by founders who overthink sourcing. Your first 5–10 hires should probably come mostly from people you or your team already know. The network has signal that a CV does not.

A 6-Week Hiring Sprint for One Role, Without an HR Team

• Week 1: Write the 90-day outcome. Write the JD. Post it in 3 places (LinkedIn, one job board relevant to your market, one community Slack or forum your target candidate would read).
• Week 2: Review applications. Apply your disqualifiers. Keep the top 15–20. Book screening calls.
• Week 3: Run screening calls. Move 4–6 forward to structured interviews.
• Week 4: Run structured interviews. Send the practical exercise to 2–3 finalists.
• Week 5: Review exercises. Call references. Make a decision.
• Week 6: Offer, negotiate, close.

Six weeks is realistic for most roles at this stage. If your process is taking longer, the bottleneck is usually in weeks 2–3 (slow application review) or week 5 (indecision after final interviews). Both are fixable.

Getting Started Today

Pick one open role. Write the 90-day outcome in two sentences. That alone will improve every other step.

If you want to go further and automate the screening and scoring pass, Kynto was built for exactly this situation: founders hiring without HR, who need a structured process without the overhead of setting one up from scratch.