Who should a seed-stage founder hire first? The defensible default sequence is: a founding engineer first, then a commercial hire (sales if your deals are large, marketing if your product is self-serve), then a second engineer, then a product or customer-facing generalist, and fifth, whatever role is currently your biggest personal bottleneck. The first two positions are well supported by data and investor consensus. Positions three through five depend on your go-to-market motion, and anyone who tells you otherwise is selling a template, not a strategy.

This playbook walks through each hire in order, explains the reasoning so you know when to deviate, and covers the two questions founders ask most: how the founding engineer bar changed in an AI-assisted world, and whether hire #2 should be sales or marketing.

## Why your first five hires are really your whole company

At the seed stage, every hire is a meaningful percentage of your entire company. According to [Carta's State of Startup Compensation H2 2025](https://carta.com/data/startup-compensation-h2-2025/), the median seed-stage startup employs just **four people**. That means your next hire is 25% of the team on the day they sign, and 20% of the team they help create.

This isn't a cute statistic. At four or five people, a single hire reshapes how decisions get made, how fast you ship, and which habits calcify into culture. There is no HR layer to absorb a mistake, no team to route around a bad fit, and no budget to carry someone who isn't pulling weight. A mis-hire at a 400-person company is a write-off; a mis-hire at a 4-person company is a quarter of your payroll and half your runway burned on a do-over.

The trend is also moving toward smaller, not larger. Axios reported in 2024 that seed startups raising in the first half of that year averaged 5.3 employees, down from 6.9 three years earlier. (Averages run higher than Carta's median of four because a few outlier teams pull the mean up.) Founders are doing more with fewer people, which raises the stakes on each individual hire and the order you make them in.

So before you write a single job description, get the sequence right.

## The default sequence (and why there's no perfect one)

Here is the honest version: **no credible source publishes a benchmarked ranking of all five roles.** Even articles titled "your first five hires" typically specify hires one and two, then go vague. That's because hires one and two have strong cross-source consensus, while three through five genuinely depend on your product and motion.

With that caveat stated, here is the defensible default:

1. **Founding engineer**, if you're building a technical product and don't have a technical co-founder who can carry the build alone.
2. **First commercial hire**: a founding account executive if your deals are large and high-touch (roughly $50K+ contracts), or a first marketer if your product is low-price and self-serve.
3. **Second engineer**, a generalist who broadens what the product team can ship.
4. **A product/design generalist or first customer-success hire**, depending on whether your gap is "build the right thing" or "keep the customers we have."
5. **Whatever is currently the founder's biggest personal bottleneck.** Be brutally honest about where your own hours go.

Hires 1 and 2 are well supported. Hires 3 through 5 are a synthesis, not an industry benchmark, and the rest of this article gives you the reasoning to adapt them to your situation. A sequence you understand beats a ranking you copied.

## Hire #1: the founding engineer (and how the bar changed)

The founding engineer is still the default first hire for technical products, but what you should screen for has shifted hard in the last two years. The short version: **AI commoditized coding speed, so the scarce skill is now judgment.**

Two data points anchor this shift:

- [CodeRabbit's State of AI vs Human Code Generation](https://www.coderabbit.ai/blog/state-of-ai-vs-human-code-generation-report) (December 2025) analyzed 470 GitHub pull requests and found AI-co-authored PRs contained roughly **1.7x more issues** than human-only PRs (10.83 vs 6.45 issues per PR). One caveat the report itself raises: AI authorship was inferred from signals rather than confirmed, so treat the figure as directional.
- [GitClear's analysis of 211 million lines of code](https://www.gitclear.com/press_mentions) found code churn, meaning code rewritten or deleted within two weeks of being written, **nearly doubled from 3.1% to 5.7%** between 2020 and 2024, with AI-assisted coding named as a key driver.

Put those together and the picture is clear. Raw output is cheap. Anyone with a coding assistant can produce a lot of code quickly. What AI cannot supply is the judgment to decide what to build, the discipline to set architectural boundaries, and the critical eye to review machine-generated code before it ships. As CodeRabbit's own framing puts it: 2025 was the year of AI speed; 2026 is the year of AI quality.

Your founding engineer's value is now the review-and-judgment layer over AI output, not keystrokes per day.

### What to test for now

Screen for three things, in this order:

1. **Product judgment.** Given an ambiguous problem and incomplete information, what do they choose to build first, and why? The founding engineer makes architecture calls with 30% of the context a big-company engineer would demand.
2. **Review discipline.** Hand them a plausible-looking AI-generated diff with two subtle problems buried in it. Do they catch them? Do they ask where the code came from? An engineer who rubber-stamps AI output will compound the churn problem instead of containing it.
3. **Shipping speed with judgment intact.** Days, not weeks, but with the self-awareness to know which corners are safe to cut.

The practical way to test this is a take-home assignment that is *deliberately AI-assistable*, then an evaluation focused on what they chose to build and how they reasoned, not raw velocity. We've written a full guide on [structuring code assignments candidates don't hate](/blog/how-to-structure-code-assignments), and our [backend engineer hiring guide](/blog/how-to-hire-backend-engineer) covers the role-specific details.

One more note on scope: if your product demands an engineer who can also sit with customers and shape requirements, you may want the [forward deployed engineer profile](/blog/how-to-hire-forward-deployed-engineer) rather than a pure product engineer. Same slot in the sequence, different archetype.

## Hire #2: sales or marketing? Let your deal size decide

The second hire is commercial, and the choice between sales and marketing should be decided by your go-to-market motion, not your personal comfort. The fork is simple:

| Your motion | Signal | Hire |
|---|---|---|
| Sales-led B2B | Contracts roughly $50K+, demos, negotiation, procurement | **Founding account executive** |
| Product-led / self-serve | Sub-$10K price point, credit-card signup, no sales call needed | **First marketer** |
| In between | $10K–$50K deals, some touch required | Usually the AE, but founder-led sales should continue longer |

If your deals are large and high-touch, hire a **founding AE**. The consensus profile from investors like [Sierra Ventures](https://www.sierraventures.com/ascend/hiring-a-startups-first-sales-hire) and executive search firm SPMB: a senior account executive with 3 to 7 years of B2B SaaS experience, who has personally carried a quota, ideally closing deals in the $50K to $300K range. Sierra's framing is worth internalizing: the first AE is part seller, part researcher, part builder. They don't just close deals; they help define your ideal customer profile and build the sales motion from scratch.

If your product is low-price and self-serve, a salesperson has nothing to sell that the product page can't. Hire a **first marketer** instead, usually once you have early signal of product-market fit, to turn whatever channel is already working slightly into a channel that works repeatedly.

Whichever side of the fork you take, one rule holds: **hire the doer, not the leader.** A VP of Sales with a 20-year resume will spend three months building a hiring plan for a team you can't afford. There is no playbook to manage yet; you need the person who writes the first draft of the playbook by doing the work.

## Hires #3 through 5: generalists, bottlenecks, and T-shaped people

For the remaining three hires, the most defensible principle is: **hire T-shaped generalists, and hire toward your bottleneck.** At zero to ten employees you are still in discovery mode. Your product, your customer, and your business model are all moving. Adaptability beats depth, because the role you hire a specialist for today may not exist in eight months, and a specialist is too large a fraction of a five-person team to risk on a guess.

A working default:

- **Hire #3, second engineer.** One engineer is a bus-factor problem and a review problem (someone has to check the AI-assisted output, including the founding engineer's). A generalist second engineer roughly doubles product surface.
- **Hire #4, product/design generalist or first customer-success person.** Pick based on your failure mode. If customers churn because the product confuses them, you need someone who can [own design and product quality](/blog/how-to-hire-product-designer). If customers love the product but onboarding eats founder hours, you need customer success.
- **Hire #5, your bottleneck.** Track your own calendar for two weeks. Whatever category of work consumes the most founder hours that someone else could do at 80% quality is hire #5. For some founders that's ops and finance; for others it's more engineering; for a few it's recruiting itself.

Specialists come after product-market fit, when the role to specialize in is finally legible. Hiring a growth-marketing specialist before you know your channel, or a DevOps specialist before you have scale, is one of the classic [startup hiring mistakes](/blog/startup-hiring-mistakes) that burns both money and a precious team slot.

## Equity for your first five

Equity drops fast with each hire, so sequence affects compensation more than most founders realize. Carta's H2 2025 compensation data shows the median equity ladder for early engineering hires:

| Hire | Median equity grant |
|---|---|
| Engineer #1 | ~1.5% |
| Engineer #2 | ~0.85% |
| Engineer #3 | ~0.50% |
| Engineer #4 | ~0.44% |
| Engineer #5 | ~0.33% |

Treat these as midpoints, not rules. The spread for hire #1 is wide: the 25th to 75th percentile runs from roughly 0.6% to over 4%, depending on seniority, salary trade-off, and how early they really are. A founding engineer joining a two-person company pre-product should land at the high end; a fifth engineer joining a post-revenue team should not.

The takeaway for sequencing: your earliest hires are the expensive ones in equity terms, which is exactly why they should be the people whose judgment compounds, generalists who grow with the company, rather than specialists you'll outgrow.

## When (and whether) to get recruiting help

Most seed founders should run their first hires themselves, and the honest trigger for outside help is time, not anxiety. Recruiting platform Dover's rule of thumb: when you're spending **15+ hours per week** sourcing, screening, and scheduling, it's time to get help. According to Dover's own cost framing, fractional recruiters run roughly $2,000 to $7,000 per hire versus $30,000+ per hire for a traditional contingency agency, and a free ATS avoids $3,000 to $10,000 per year in software spend. Those are a vendor's marketing numbers, so weigh them accordingly, but the direction is right: agency economics rarely make sense for a five-person company.

There's also a real argument for *not* delegating early hiring even when you can afford to. Your first five hires set the bar for the next fifty. Interviewing them yourself is how you learn what your bar actually is. Tooling that removes the administrative drag, without removing you from the decisions, is the sweet spot at this stage.

## Running the sequence with Kit (free)

Knowing the sequence is half the job; running five searches with no recruiter, no employer brand, and no inbound is the other half. This is the exact situation Kit was built for, and the free tier covers all of it, which removes the software-cost objection entirely.

Here's how the sequence maps to the product:

- **One pipeline per hire, from a template.** Kit ships [pre-configured process templates](/templates), so you're not designing an interview funnel from scratch for each archetype: founding engineer, founding AE, first marketer.
- **Score for judgment, not speed.** For the founding engineer, combine Kit's code assignment stage with team review. Use an assignment that's deliberately AI-assistable, then have reviewers score what the candidate chose to build and how they reasoned. Kit's assignment payouts (on by default) compensate candidates for their time, which matters when you're a no-name company asking for hours of work.
- **A careers page in minutes.** Kit's career portal gives a four-person company with zero employer brand a credible, branded careers page, so your founding AE candidates don't land on a Notion doc.
- **Outbound, because you have no inbound.** At seed stage, the candidates you want are not applying; you go get them. Kit's outreach campaigns handle the sourcing sequences for the founding-engineer and founding-AE searches.
- **A talent pool for hires #4 and #5.** The great candidate you meet during hire #2 but can't afford until hire #5 shouldn't vanish into an inbox. Kit's talent pool keeps them warm so your future searches start half-done.

<div class="blog-inline-cta">
  <p><strong>Making your first hire this quarter?</strong> Kit gives you the pipeline templates, code assignments, careers page, and outreach to run the whole sequence yourself, free.</p>
  <p><a href="/users/sign_up">Start with your first role</a></p>
</div>

The sequence in one line: founding engineer, then sales-or-marketing by deal size, then a second engineer, then product or customer success, then your bottleneck. Hires one and two are well-evidenced defaults; three through five are yours to adapt. At a median team size of four, each of these decisions is a fifth of your company. Make them in the right order, screen the founding engineer for judgment over speed, and keep yourself in the interviews even when the admin gets heavy. The tooling can be free; the judgment has to be yours.