Here's the trap built into every sales hire: you're evaluating a professional persuader using a persuasion exercise.
Salespeople are trained (literally, professionally trained) at the exact skills that win interviews.
Rapport, confident answers, objection handling, closing for next steps.
The candidate who dazzles you in the interview has proven exactly one thing: they can sell themselves, one time, with weeks to prepare.
Whether they can sell your product every day for a year is a separate question, and your gut can't answer it.
That's why sales hires fail more often than any other role, and why learning how to hire a salesperson is mostly about learning to stop trusting the interview.
I've hired sales teams across 20+ years at The Motley Fool, The Hustle, HubSpot, and Hampton. And here's the system.
The short version: hiring a salesperson means never trusting just the interview. Decide whether you need a setter or a closer, source from adjacent products and direct outreach instead of job boards, require a prepared mock pitch, score every stage identically, and verify their numbers with former managers before you sign anything.
When to hire a salesperson (and when you're not ready)
The uncomfortable rule: don't hire a salesperson until you've sold the product yourself, repeatedly.
If you can't write down your sales motion (where leads come from, what the pitch is, the common objections, how long deals take, why you lose), then you don't have a sales process for a rep to run.
You have a hope that someone else is going to figure it out.
Founders who hire a rep to "figure out sales" are paying a salary to outsource discovery work only the founder can do.
The rep fails through no fault of their own.
You're ready when you've personally closed enough deals to see the pattern, you have more qualified conversations available than time to run them, and you can hand over a written playbook. Even a rough one.
Not ready yet? Keep selling, and hire elsewhere first. Sequence matters.
Why sales hires fail more than any other role
Three compounding reasons:
The evaluation is rigged. You're using an interview to assess someone whose profession is winning interviews.
The lag is long. A bad ops hire can surface in a few weeks. A bad sales hire hides behind "ramp time" for one, two, three quarters or can blame things on a “bad macro environment”. All that while they’re burning your warmest leads.
The cost is multiplied. Salary, plus the pipeline they mishandled, plus the market time you lost. DePaul's sales turnover research put the average cost of replacing a single rep near $98,000, and that's before you count the deals that never happened.
The fix isn't better instincts. It's changing what you evaluate. Less telling, more selling.
Decide what you're actually hiring
"Salesperson" is not a role. Before anything else, figure out which job this is:
A closer. Runs demos and pitches, gets contracts signed. Needs your pipeline to already produce conversations.
A setter (SDR). Prospects and books meetings that you or a closer runs. The right first hire when your time is the bottleneck. Also much easier to evaluate, because the activity is visible daily.
An account manager. Grows and keeps existing customers. A different animal entirely. Don't expect hunting.
Most founders should split the role before hiring for it.
If you're great at closing but starved for at-bats, hire a setter, not a "VP of Sales." Seniority inflation kills more first sales hires than any other scoping mistake.
How to hire a salesperson only on commission (the honest version)
Short version: you probably can't, and shouldn't.
Every experienced rep reads commission-only as "the founder wants sales risk for free." The candidates who accept it are, with rare exceptions, the ones who can't get a base anywhere else. If your budget can't support a base salary, you may not be ready for a sales hire. Do the founder-selling work until it can.
What works instead: a modest base plus meaningful commission, structured so on-target earnings land around a 50/50 to 60/40 base-variable split for closers.
Commission-only has a narrow legitimate lane: established rep networks in industries where it's the norm, like some insurance and manufacturing rep models. If you're not in one of those, don't fight the market.
Where do you find salespeople worth interviewing?
Job boards underperform for sales roles. The best reps are employed and hitting quota, not browsing listings.
Better sources: reps selling adjacent products to your same buyer (they bring the relationships), direct LinkedIn outreach written like a sales pitch (you're selling the job, and badly written outreach filters you out), and your network's answers to "who's the best rep you've ever worked with?"
One more filter that costs nothing: ask every applicant for a 3-minute video on why this role. For a sales candidate, that video is a work sample. Energy, clarity, persuasion, on camera, exactly like their future demos.
The evaluation system: make them sell, don't let them tell
This is the heart of it. Structure every stage so the candidate demonstrates selling instead of describing it.
1. Evidence-based interview. Skip "what's your superpower." Ask for deal autopsies: "Walk me through the last deal you lost. What happened?" "Take me through your current pipeline math, activity to meetings to closes." Real sellers talk in specifics and numbers. Performers talk in philosophy.
2. The numbers audit. Ask for specific quota attainment, deal sizes, and cycle lengths. Write them down. You'll verify with references later, and inflated claims decay fast under specific questioning.
3. The mock pitch. Non-negotiable. Give the candidate your actual product, a few days to prepare, and 30 minutes to sell it to you while you play your real buyer, objections included.
You're not grading polish on day zero. You're grading preparation, discovery questions, objection handling, and whether they ask for the close. Anyone offended by the exercise has told you everything.
4. Scorecard the whole thing. Same stages, same scores, every candidate. Chemistry is how sales candidates beat you. The scorecard is how you take that weapon away. (Scorecards come from the non-negotiables you set before anyone applied. That’s Step 2 of Start the Hire, free, and it’s what keeps a great pitch from rewriting your standards mid-process.)
Red flags that should end the process
A few patterns experienced sales hirers treat as disqualifying, whatever the charm level:
Vague numbers. "I crushed quota most years" instead of "108% in 2024, 92% in 2023, and here's why." Real producers know their stats cold.
All philosophy, no autopsy. Ten minutes on "building relationships," but they can't walk you through one specific lost deal.
They blame every miss. Bad territory, bad leads, bad product, bad manager, four jobs in a row. The common denominator is sitting in your interview.
They won't do the mock pitch. "My track record speaks for itself" means "I don't want you to see me sell."
They negotiate like you're the enemy. How they sell you on their comp is a live demo of how they'll treat your prospects.
None of these are quirks to coach out later. In a role where the interview is the work sample, disqualifiers are data.
Reference checks that actually derisk the hire
Sales references are uniquely verifiable, because there are numbers. Call former managers and check the claims: "They told me 115% attainment in 2024. Does that match?"
Ask where they ranked on the team, whether the manager would hire them again for this kind of motion (SMB vs enterprise, hunting vs farming), and how they behaved in a losing quarter.
Mismatched numbers or a hesitant former manager are disqualifying, not discountable.
What it costs (quick reference)
Rough US bands to sanity-check your budget. Verify current numbers for your market before setting comp:
Role | Base | OTE | Time to meaningful production |
SDR / setter | $45-65K | $65-90K | 1-3 months |
SMB-market closer | $60-90K | $120-180K | 3-6 months |
Enterprise rep | Well beyond | Well beyond | 6+ months |
These skew B2B SaaS, where most published sales comp data comes from. Adjust down outside tech.
Under-market comp doesn't save money. It selects for candidates nobody else wanted, which is the most expensive thing you can buy.
Budget reality: base, plus employer costs, plus tools, plus the leads they'll consume while ramping. Expect 1-3 months before meaningful production for a setter, 3-6 for a closer with a real sales cycle. The Bridge Group's 2025 data has SDR ramp at a 15-year low of 3.0 months while closer ramp hit an all-time high of 6.2. The gap is widening, which is another argument for starting with a setter.
Set your standards before they charm you: Start the Hire (free)
A salesperson will out-talk your instincts. The fix is writing down what you require before anyone applies: the role, the non-negotiables, the JD. Plus the 10 Red Flags checklist to run every finalist against. Three steps, one sitting, free.
FAQ: hiring a salesperson
What is the 70/30 rule in sales (and in sales interviews)? In selling: the prospect talks 70%, the rep 30%. Discovery beats pitching. Use it in the interview too. Candidates who monologue at you will monologue at your prospects. And watch the mock pitch for it: the good ones ask questions first.
Should a startup's first sales hire be a leader or a rep? A rep, or a setter. Sales leaders manage systems and people. With no system and no people, a "VP of Sales" either does rep work at triple the price or builds process for a motion you haven't proven. Hire leadership after 2 or 3 reps are producing.
How long should ramp take before I worry? Set the bar in writing before day one: activity metrics from week one (calls, meetings booked), pipeline creation by day 30-60, closed revenue on a timeline matched to your sales cycle. "Still ramping" without visible activity is not ramp. It's the lag phase of a mis-hire.
Jordan DiPietro has spent 20+ years hiring for The Motley Fool, The Hustle, HubSpot, and Hampton, and is the creator of Hiring OS, the opinionated hiring system for growth-stage founders.
Related reading:
Looking For More?
—> Here are 3 ways I can help you:
