A startup recruiting story from 8 to 19 people with no recruiter

This startup recruiting case study follows a seed-stage New York SaaS company that hired 11 people across engineering, sales and support without a recruiter. It sourced and interviewed in The Cognitive. Customer-reported: $720 per hire against agency quotes of $15,000+, and founder time per hire down from 22 hours to 4.

What changed for the New York seed startup?

The first 4 rows are the customer's own figures. The last 2 rows are our arithmetic on them, and the working is further down the page.

Customer figures from our case study record dated August 2026. The before column for cost is the agency quote the founders received, not money spent.

Where did the startup start, and why did founder-led hiring break?

The customer is a seed-stage SaaS company in New York with 8 people and open roles in engineering, sales and support. It had no recruiter, no spare time and no employer brand that would make strong candidates come looking. The founders did the hiring themselves, and they had a product to build at the same time.

In practice that meant scanning applicants at midnight. Screens got squeezed between product work, replies went out days late, and the best candidates were often already deep in another company's process by the time the founders called. The record puts it plainly: they kept losing people to companies that simply responded faster.

By the founders' count, each hire took 22 hours of founder time, taken straight from an 8-person roadmap. The market made it harder: the BLS put the information industry's job openings rate at 4.3% for August 2026 (JOLTS, read 5 October 2026), and Gem's 2026 benchmarks found recruiters carrying 13.4 open roles each (Gem, read 5 October 2026).

It's the recruiting team we couldn't afford, for less than our coffee budget. Co-founder and CEO, Seed-stage SaaS Company, New York

Why didn't a recruiting agency work for a seed startup?

They looked at recruiting agencies first. The quotes came back at $15,000+ per placement, with fees of 20 to 25% of salary. Across 11 hires that would have eaten most of the seed round's hiring budget, and an agency would still have needed founder time for briefs and final rounds.

So the founders absorbed recruiting themselves: posting roles, scanning inbound and fitting screens around everything else. Roles kept moving, but slowly. The constraint was never a shortage of people to contact. It was the hours two founders could give to reading, screening and replying.

The agency quote also sits far above wider benchmarks. SHRM's benchmarking put the average cost per hire at nearly $4,700 (SHRM, read 5 October 2026), an average that includes companies with in-house recruiters. The founders' $720 per hire came in at about 15% of that average, on our arithmetic.

How did they find engineers and a founding AE without a recruiter?

A plain-English brief is read into editable filters, then every person is judged against its must-haves.

For the outbound roles, the founders wrote each brief in plain English and searched ~900M public profiles. The brief becomes filters they could edit. People who miss a must-have drop below people who meet them, and nobody is removed. Verified emails and phone numbers can be revealed for the people worth contacting.

How did the startup run recruiting on one platform?

The record describes the rollout. Each step below is written in today's product terms, because the setup dates from mid 2026 and the product has changed since.

1. Load the open roles One role each for engineering, the founding AE and support. Each role got its own weighted rubric, so a sales interview and an engineering interview were scored on different things. Example AE rubric, ours: discovery 30, objection handling 30, explaining the product 20, working without a playbook 20. Weights must total 100.
2. Write a sourcing brief per outbound role Plain English, with the must-haves spelled out. The founding AE brief started too broad, and the founders tightened it before the results felt usable. Example brief: "Has sold B2B software as one of the first sellers at a startup, has run full cycles alone, New York or remote US."
3. Reveal contacts and start sequences Reveal verified emails and phones only for people worth contacting, then send email sequences from your own mailbox. Replies or opt outs stop the sequence. Credits: 1 per person a search shows, 4 per email revealed, 12 per phone, charged only when a value comes back.
4. Interview sourced and inbound candidates the same way Every candidate, whether found by search or applying on their own, was invited to a live, two-way AI video interview. Today a role runs at 10 or 20 minutes. A setup to copy: 20 minutes for engineering and the founding AE, where follow-ups need room, and 10 minutes for a support first round.
5. Let candidates book their own slot Candidates pick a time without creating an account and get email reminders. No founder calendar is involved in the first round. Open slots outside office hours inside the role's window, so employed candidates can book around their current job.
6. Read reports, not every profile The founders' main adjustment was trusting the shortlist instead of reading every profile. They reviewed reports over coffee and opened recordings when a score looked borderline. A review rule to copy: a 4 or 5 on the top weighted criterion gets a full read; any 3 gets the recording; a 1 or 2 gets a skim.
7. Run only the final conversations A founder decided who moved on and ran the last round with each finalist. Offers and closing stayed with the founders. The 4 founder hours per hire cover this step plus reading reports: the last conversation, the decision and the offer.

What should the AI interview probe for each startup role?

Example probes, written by us to show the competencies, not the customer's questions. More in our account executive, software engineer and customer support banks.

A buyer agreed to a call but has not said why. You have 15 minutes. How do you run it?

Opens with questions about how the buyer works today, listens for a cost or pain they name in their own words, and agrees a dated next step instead of pitching features.

The buyer says they are happy with their current tool. What do you ask next?

Why would a buyer take a risk on a company of under 20 people?

Turns the size into a reason, such as founder access and fast fixes, and says plainly what a small vendor cannot offer yet.

Their security team asks for a certification you do not have. What now?

Tell me about something you built end to end with no spec. What did you decide on your own?

Names the trade-offs they chose without being told, what shipped first and what they would cut again. Early engineers need that judgment more than a long tool list.

What broke after launch, and how did you find out?

A customer says an export is missing rows. Walk me through your first 10 minutes.

Tries to reproduce it before replying, asks which filters were set, and tells the customer what they are checking and when they will hear back.

Engineering says it works as designed. What do you tell the customer?

What changed, number by number?

Team size went from 8 to 19, which is 11 hires across engineering, sales and support. Recruiter headcount stayed at 0 the whole way, and the record says no recruiter was needed before Series A.

Cost per hire averaged $720 against agency quotes of $15,000+. That is at least $14,280 less per hire, or about 95% lower. Over 11 hires, the gap is about $157,000 that stayed in the bank. Against SHRM's average of nearly $4,700, $720 is about 85% lower.

Founder time per hire fell from 22 hours to 4, which is 18 hours saved per hire and about 82% less. Across 11 hires that is 198 hours, roughly 200 founder hours returned to the product. At 40 hours a week, that is about 5 working weeks of one founder. For a company this size, the record argues that hours mattered as much as cash.

How was the startup hiring math worked out?

Inputs are customer-reported. Everything after an equals sign is our arithmetic.

HIRES
19 people minus 8 people = 11 hires.

COST PER HIRE
Agency quote $15,000+. Actual $720.
$15,000 minus $720 = $14,280 less per hire, at minimum.
$14,280 / $15,000 = 95.2% lower.

SPEND FOR 11 HIRES
Agency quote: 11 x $15,000 = $165,000+
Actual: 11 x $720 = $7,920
$165,000 minus $7,920 = $157,080, the record's roughly $157,000.

FOUNDER HOURS
Before: 11 x 22 = 242 hours
After: 11 x 4 = 44 hours
242 minus 44 = 198 hours, the record's roughly 200.
22 minus 4 = 18 hours saved per hire. 18 / 22 = 81.8%.
198 hours / 40 hours a week = about 5 founder weeks.

AGAINST THE SHRM AVERAGE
$720 / $4,700 = 15.3%, so about 85% lower.

WHAT THE AGENCY QUOTE IMPLIES
$15,000 at 25% of salary = a $60,000 salary.
$15,000 at 20% of salary = a $75,000 salary.

What did a founder's hiring week look like before and after?

Before follows the record's account of founder-led hiring. After follows the rollout it describes, at the reported 4 founder hours per hire.

Hours per hire are customer-reported. The 10 or 20 minute length describes today's product.

What went wrong during the startup's rollout?

Calibration. The first brief for the founding AE was too broad, so early results were a mix of people who matched the title and people who matched the job. The founders tightened the brief, naming the must-haves for an early sales hire, before the recommendations felt usable.

The second adjustment was behavioural. Founders who are used to reading every profile have to learn to read a report instead. The record names this as the main change for them: trusting a shortlist built against written criteria, and spending their own time on the final conversation.

What did the founders read after each AI interview?

An example interview report: a 1 to 5 score per criterion, a weighted score out of 100 and resume claim checks.

Each criterion gets a 1 to 5 score, with overall written feedback on strengths, areas to improve and the reasoning behind the recommendation. The report suggests a verdict and nothing is rejected automatically. A founder reads it, opens the recording if needed, and decides.

What does this startup case study not show?

• One company and 11 hires across 3 functions: a planning data point for seed teams.
• The $15,000+ is an agency quote, so the savings compare against money never spent.
• An agency fee also buys someone to sell the role and close; here the founders did that.
• The setup dates from August 2026. The product has changed since, and this page describes today's workflow.

What still needed a founder in this startup recruiting setup?

The final conversation, the decision and the close. An agency fee also pays for someone to sell the role and negotiate the offer, and the record is honest that the comparison is not exact. Here the founders did all of that themselves, which is part of why 4 hours per hire remained.

The Cognitive is not an ATS either. It has no job posting, careers page, offers or human interview scheduling. Inbound candidates still came from wherever the roles were posted, and the founders kept their own record of offers. People sourced inside The Cognitive are not written into an ATS automatically.

Which startups does this approach fit, and which does it not?

Do

• Seed teams hiring a handful of roles across functions with no recruiter.
• Founders who can write what a strong hire looks like before searching.
• Roles where a spoken conversation shows the skill, such as sales or support.
• Teams that will spend their saved hours on final rounds and closing.

Do not

• Companies that want someone else to sell and close every candidate.
• Teams that need job posting, offers or a careers page in one tool.
• Founders unwilling to act on a report without rereading every profile.

Try it on one role

Write one brief, interview this week's candidates and read the reports before deciding anything else. Start free at https://app.thecognitive.io/signup.

Where to go next

Recruiting software for startups · How to find candidates · How AI sourcing works · Hire an account executive · Pricing · All case studies

Sources and notes

Figures reported by the customer; setup dated August 2026. Source: The Cognitive case study record for Seed-stage SaaS Company, New York, created 4 August 2026 and read 4 October 2026. The record does not name the company, so neither does this page. Derived figures are our arithmetic and are labelled where they appear. Example rubric, brief and probes are ours.

External context, all read 5 October 2026: SHRM, The Real Costs of Recruitment (11 April 2022, average cost per hire); SHRM recruiting benchmarking (2026, median time to fill); BLS JOLTS Table 1 (August 2026, released 29 September 2026); Gem 2026 Recruiting Benchmarks key takeaways (1 December 2025).

Frequently asked questions

Can a startup hire without a recruiter?

Yes. This seed-stage New York SaaS company went from 8 to 19 people with no recruiter. It sourced outbound candidates and ran first-round AI video interviews in The Cognitive, and the founders spent their time on final conversations. It reported 4 founder hours per hire, down from 22.

What is the best recruiting software for startups with no recruiter?

Look for one place that finds people, contacts them and runs a first-round interview, so founders only see finalists. That is the setup in this case study. Check that the tool shows its reasoning, keeps a recording and leaves the decision to you. Our roundup of recruiting software for startups compares the options.

How does $720 per hire compare to a recruiting agency?

The agency quotes this company received were $15,000+ per placement, or 20 to 25% of salary. On that quote, $720 is about 95% lower. The comparison is not exact, because an agency also sells the role and helps close. Here the founders did that part themselves.

How many founder hours does a hire take?

This team reported 22 founder hours per hire before the change and 4 after. Across 11 hires that is 198 hours back. Your number depends on how much of screening you hand off. Track it for one hire before you change anything.

What is the cheapest hiring software for a small business?

On The Cognitive, AI Sourcing and AI Interview each start at $99/month, on monthly plans you can cancel anytime. Each AI Interview plan includes a monthly interview allowance, and past it you buy an interview top-up. Separately, the 7-day free trial includes 2 AI interviews plus 100 sourcing credits.

Can AI interview candidates for a founding sales role?

Yes. A live, two-way AI video interview can probe discovery, objection handling and how a candidate explains a product, with follow-ups when an answer is thin. The founders still ran the final conversation and made the call on their founding AE.

Does the AI reject startup candidates automatically?

No. Each report suggests a verdict from the weighted score, and a person accepts or overrules it. Nothing is rejected automatically. Integrity flags such as tab switches are logged for review and do not change the score.

How do sourcing credits work for a small team?

A search costs 1 credit per candidate it returns. Revealing an email costs 4 credits and a phone number 12, and a reveal is charged only when a value comes back. Small teams usually reveal contacts only for the people they plan to message.

Do candidates know the interview is run by AI?

Yes. The invitation tells candidates the interview is run by AI. They book their own slot without creating an account and get email reminders before it starts.

Does The Cognitive work with a startup's ATS?

It connects to 60+ ATS tools, including Ashby, Greenhouse, Lever and Workable. Candidates imported from the ATS get a note with the score, summary and report link. Many seed teams have no ATS yet, and a CSV upload plus a read-only report link works too.

What is the average cost per hire for a startup?

SHRM's benchmarking put the average cost per hire at nearly $4,700 across employers (SHRM, read 5 October 2026). Agencies sit far above that: this startup was quoted $15,000+ per placement, or 20 to 25% of salary. It reported $720 per hire across 11 hires by sourcing and running first-round interviews in The Cognitive.

How long does it take to fill a role?

SHRM's 2026 benchmarking puts the median time to fill a nonexecutive role at 39 calendar days (SHRM, read 5 October 2026). For a founder, the cost that matters is hours inside those days. This team reported 22 founder hours per hire before and 4 after, because the first round ran on the candidate's schedule.

How do you hire a founding account executive?

Write a brief that names what an early seller must have done, such as selling without a known brand behind them. Source against it, interview every candidate on discovery and objection handling, and run the final round yourself. This startup's first AE brief was too broad, and tightening it is what made the results usable.

Should a startup use a recruiting agency or recruiting software?

An agency sells the role, screens and helps close, for a fee this team was quoted at 20 to 25% of salary. Software hands you the sourcing and first round and leaves the selling and closing to you. This team chose software and spent the 4 remaining founder hours per hire on the final conversation and the close.

Explore: Hiring software for small business · Recruitment agency fees · What is cost per hire · AI recruiting for tech hiring · Seed startup engineering case study · Recruiting outreach software