A hiring plan for startups

A hiring plan is not a headcount chart with optimistic numbers on it. It is a short, honest answer to one question: which roles does the business need next, in what order, and why. Get that right and hiring becomes a sequence of deliberate decisions. Get it wrong and you hire to a vague sense of growth, then wonder why the new people are busy but the constraint has not moved.

This guide covers how to tie roles to real needs, how to sequence and define them, the cost and time reality nobody warns you about, and why candidates and new hires belong in one system.

Start from the constraint, not the org chart

Founders often plan hiring by drawing the company they want to be and filling in the boxes. That produces a tidy chart and a lot of premature hires. The better starting point is the constraint: the one thing that, if it stays broken, stops the business getting better.

For each role you are considering, write one sentence that finishes “if we do not fill this in the next quarter, then…”. If the sentence names something concrete (deals we cannot close, a product area with no owner, support that is drowning), the role is real. If the best you can manage is “we would have more capacity”, the role is a want, not a need, and it can wait.

This keeps the plan honest. A startup has a fixed amount of money and attention, and every seat you open spends both. The job of the plan is to make sure you spend them on the thing that is actually holding you back.

Sequence the hires

Order matters more than count. Two hires in the right sequence beat four in the wrong one.

A few rules that hold up:

• Hire the person who unblocks other hires first. A first engineering lead makes the next three engineers productive. Hiring the three first and the lead later usually costs you the difference twice.
• Do not hire a manager before there is anything to manage. Layers added ahead of the work create overhead, not output. Add the manager when the team exists and the founder can no longer hold it directly.
• Front-load the roles with long lead times. Senior and specialist seats take months to fill. If you know you need one in two quarters, open it now, because the alternative is a gap exactly when the work arrives.
• Leave slack. Plan the next few hires firmly and the rest loosely. The far end of the plan will change, and pretending otherwise just makes you wrong on a schedule.

Define the role before you post it

The most expensive hiring mistake is opening a role you have not defined, because you find out it was underdefined three months in, after the person has started.

Before anything goes public, write down four things:

1. The outcome the role owns. Not a list of tasks, the result. “Own the release process so shipping stops being a founder job.” A role defined by its outcome attracts people who want to own an outcome.
2. The real must-haves. The two or three things a person genuinely cannot do the job without. Be ruthless, because every must-have you add shrinks the pool and lengthens the search.
3. The level and shape. Is this a first hire who builds from nothing, or an operator who runs an existing function. Getting the level wrong wastes everyone’s time on both sides.
4. How you will know it is working. What the first ninety days should produce. If you cannot describe success, you cannot interview for it and you cannot manage to it.

With those four written down, the job description writes itself, and the interview process has something concrete to test against.

The cost and time reality

Two numbers founders underestimate, every time.

Cost. The salary is the small part. Add employer taxes, equipment, software seats, the recruiter or job-board spend if you use them, and the real cost: the founder and team hours spent sourcing, screening, and interviewing. A single considered hire eats days of senior time that would otherwise go into the product. This is a reason to hire deliberately, not often.

Time. From opening a role to someone actually doing the work, plan on months, not weeks. Sourcing takes time, scheduling a fair loop takes time, making a clean decision takes time, and then the person you want has a notice period. A hire you need in Q3 is a role you open in Q1. Build the lag into the plan so people arrive when the work does.

Keep candidates and hires in one system

Here is the part that quietly wastes the most time. Many startups run hiring in one tool and their HR records in another. Every time an offer is signed, someone re-types the name, the start date, the salary, and the documents from the first system into the second. It is slow, it is error-prone, and it drops exactly the data you most need to get right.

The cleaner model keeps candidates and employees in the same place, so a signed offer becomes an employee record in one step. In Capstan the Recruitment module does exactly this: an accepted offer converts the candidate through the same onboarding machinery a directly added hire uses, so their details and documents carry straight through and the new record keeps a link back to the application it came from. No AI ranks or screens the candidates along the way; a human reads every application on the merits.

Two facts about the shape of it are worth having before you plan around it. Each open role is a requisition with its own publish lifecycle, and publishing snapshots the job description as it stood at that moment, so a post you edit next quarter does not silently rewrite what a candidate applied to. And the module is charged against active recruiter seats rather than your headcount, which means a hiring pause costs you less than it would on a per-employee price. Like every module it needs a paid plan under it; the pricing page has the figures.

The practical payoff is that your plan, your pipeline, and your team stay one continuous record instead of three that disagree.

Where to go next

Once the plan names a role, the next moves are concrete. Write the job description from the outcome and the must-haves, run a fair interview process that respects the candidate’s time, and have the onboarding checklist ready so the signed offer turns into a productive start. You can also look at the product directly: the core is free up to twenty active employees, and the module prices are public.