Most companies that call us about a failed sales hire want to talk about interviewing. They describe someone who was impressive in the room, who half closed them by the second conversation, and who then produced almost nothing across eight months. What they want from us is a better question set, or some technique that would have exposed the gap earlier.
Interviewing is rarely where the hire was lost. In nearly every case we can trace the failure back to a document written weeks before anyone was interviewed, which is the role definition. The company knew it wanted a salesperson. It had not decided what kind of selling that person would actually do, what a good month looked like in numbers, or which parts of the job were genuinely difficult. A vague brief produces a vague shortlist, and no interview technique recovers from that.
Leadership IQ tracked 5,247 hiring managers across 312 organisations and found that 46 percent of new hires fail within 18 months, with only 19 percent achieving unequivocal success. Of the failures, 89 percent came down to attitude and motivation rather than technical skill. Those are precisely the qualities a job advert never describes and a CV cannot show.
What follows is the full sales hiring process we run, written from the hiring company's side. The short version is that knowing how to hire a salesperson is mostly knowing what to decide before the advert goes out.
SalesFuel's 2026 research put the average cost of a bad B2B sales hire at $177,171. The same research found companies spend close to three months hiring a rep before that person makes a first sales call, then another five and a half months before they reach the productivity of their peers.
Read those two numbers together. The direct cost is the salary and the recruitment spend. The larger cost is the eight and a half months of a territory sitting underworked while you find out whether the hire was right. A process that takes an extra fortnight and raises your hit rate is not slow, it is the cheapest intervention available to you.
.jpg)
Job titles in sales mean almost nothing across companies. An account executive selling a €400 monthly subscription to marketing managers and an account executive selling a €600,000 platform into a procurement committee share a title and essentially no daily work. Hiring one when you needed the other is the most common error we see, and it survives the whole process because the title matched.
Start with the motion instead. Write down who the buyer is, how deals begin, the average deal value, the average cycle length in weeks, how many people are involved in a decision, and what this hire is being brought in to change.
What problemdoes this hire solve that the current team cannot? If the honestanswer is capacity, say so, because that is a different candidatefrom one hired to open a market.
What does monthsix look like, in numbers? Not the annual quota, but the specificpipeline, meetings or closed revenue you would accept by month sixgiven your ramp reality.
Which part ofthis job is hardest? Every sales job has one, whether that is gettinga first meeting in a market that does not know you, holding priceunder pressure, or keeping a nine-month cycle moving. The hardestpart is what you should be selecting for.
What has gonewrong in this seat before? If it has churned, the reason is usuallystructural and will churn the next person too.
"Fiveyears of B2B sales experience" screens for almost nothinguseful. Five years of taking inbound demos at a well-known brandtells you very little about whether someone can open a cold market,while two years of grinding outbound at an unknown company might tellyou a great deal.
Replace theyears with the motion. If the job requires generating pipeline fromnothing, ask for evidence of self-sourced pipeline. If it requiresmulti-stakeholder deals, ask for evidence of running one. Therequirement should describe something the person has done, not aduration they have survived.
The scorecardis a single page listing the outcomes the hire owns, the competenciesrequired to deliver them, and the evidence that would satisfy you oneach. Write it before the advert, because the advert derives from it,and before the interview, because the interview tests it.
Keep it to fiveor six competencies. More than that and every candidate scoresmiddling on everything, which tells you nothing. Write down whatstrong evidence looks like in advance. "Good at prospecting"is a wish. "Has personally generated at least 40 percent oftheir own pipeline for two consecutive years, and can describe thechannel mix and reply rates" is something a candidate either canor cannot demonstrate.
You arecommitting to your standard before you meet anyone charming. Salescandidates are, by selection, persuasive people, and deciding whatgood looks like while you are still calm is the main defence youhave.
Two thingsdetermine whether the right people apply. The first is whether theadvert describes the actual job rather than a generic one. The secondis whether it gives a salary range.
Describe themotion you wrote down in step one. Say what the person will sell, towhom, at what deal size, and against which competitors, then say whatthe first six months involve, including the parts that are hard.Strong salespeople are not deterred by a difficult brief, they aredeterred by a vague one, because vagueness signals that the quotawill be invented later.
On pay, publishthe base and the on-target earnings, and say what proportion isvariable and whether it is capped. Withholding the range filters outexactly the in-demand candidates you are trying to reach, becausethey have other conversations available and will not spend ascreening call finding out whether yours is worth having.
.jpg)
A shortlistassembled from one channel inherits that channel's bias. Inboundapplications alone over-represent people currently looking, which insales correlates with having recently missed target. That is not arule, and plenty of excellent people are between roles for ordinaryreasons, but the skew is real and worth correcting for.
Run threechannels in parallel. Post the advert so active candidates can findyou. Ask your team and your investors for referrals, which remain thehighest-quality source in most searches. Then approach peopledirectly who are not looking, which is where the strongest candidatesusually sit. Most companies that set out to recruit salespeople skipthis third channel, because it is slow and the reply rates areunflattering.
If you are notgoing to do the third one properly, that is the honest point at whichto bring in a specialist. The reason to use a headhunter is notaccess to a database. It is that someone is having continuousconversations with people who are performing well and are not on themarket, and knows what those people currently cost.
A screeningcall in sales has one job, which is to establish whether the numberson the CV are real and whether they were produced in conditionsresembling yours. Twenty minutes is enough if you ask properly.
Ask for thequota in euros, the attainment as a percentage, how many peoplecarried the same quota, and how many of them hit it. A rep at 95percent on a team where two of fourteen hit target is a strongersignal than a rep at 110 percent where everyone cleared it.
Ask where thepipeline came from, whether self-sourced, marketing-generated, orinherited from a predecessor's account list. This single questionseparates candidates faster than anything else, because a rep who hasonly ever worked handed-over pipeline has never done the job you maybe hiring for.
Ask about thedeal they lost that they still think about. The answer revealswhether they analyse their losses or narrate them. Reps who say"procurement went with the cheaper option" and stop thereare usually the same reps who will not know why your deals arestalling either.
Structure isthe part that makes your decisions predictive. Schmidt and Hunter'smeta-analytic review of roughly a century of selection researchplaced structured interviews at around 0.51 for predicting jobperformance against roughly 0.38 for unstructured conversations, withwork sample tests among the strongest predictors available. The samequestions asked of every candidate, scored against your scorecard,will keep beating a free-flowing chat even when the chat feels moreinsightful at the time.
Scoreindependently before the panel confers. That part matters more thanit sounds, because once one interviewer says "I loved her",the others adjust, and you have converted three opinions into one.
For anycustomer-facing sales role, ask the candidate to sell. Give them yourproduct, a defined prospect, and two or three days to prepare, thenrun a thirty-minute meeting where somebody plays the buyer with arealistic objection.
What you arewatching for is not polish. It is whether they researched before thecall, asked questions before presenting, connected your product to abusiness problem rather than listing features, handled the objectionwithout either collapsing or steamrolling, and asked for a next step.Those five behaviours, observed once, tell you more than fourconversations about their philosophy of selling.
Most referencecalls are a formality where a former manager confirms dates and sayssomething pleasant. Ask for something specific instead. Ask what thequota was and what the candidate attained, then compare it with whatthey told you. Ask where they ranked, and ask what the manager had tocoach them on, which gets real answers because every manager coacheseveryone on something and the shape of the answer is informative.
Then askwhether they would hire the person again into the same role, andlisten to the pause as much as the answer. Two references from directmanagers beat five from colleagues and customers. If a candidatecannot produce a single manager from their last two roles, that isitself data.
Offer stage iswhere organised processes lose candidates to disorganised ones thatmoved faster. Ashby's analysis of 230,000 applications reaching offerstage between January 2021 and March 2024 found an average acceptancerate of 78 percent, rising to 81 percent in 2023, with business rolesaveraging 84 percent against 73 percent for technical ones.Candidates who eventually declined spent around six days in the offerstage, roughly twice as long as those who accepted.
That lastfigure is the operationally useful one. Time in offer stage is not aneutral administrative delay, it correlates with the outcome, becausea candidate sitting with your offer for a week is usually a candidaterunning another process. Make the verbal offer by phone the same dayyou decide, have the base, commission structure, quota and rampallowance already agreed internally, and send paper withintwenty-four hours of the verbal yes.
The hiringprocess ends when the person is productive, not when they sign, andthe gap between offer acceptance and start date is wherecounter-offers land and second thoughts form. Use it. Send theproduct materials and the territory data, introduce them to the teamover a call, and have day one actually ready, with the laptopconfigured, CRM access working, and a named person responsible forthe first week.
Five to sevenweeks from brief to accepted offer is realistic for an individualcontributor role. Roughly one week to write the brief and scorecardproperly, two to three weeks of sourcing and screening in parallel,one to two weeks of interviews and the work sample, and a few days toclose. Sales leadership searches run eight to twelve weeks, becausethe pool is smaller and the people worth hiring are almost alwaysemployed.
If your lastthree hires took wildly different amounts of time, the variation isitself the finding. It usually means the process restarts fromscratch each time, and the cost of that shows up as inconsistenthires.
Should Ihire for industry experience or sales ability?
Sales ability,in most cases. Industry knowledge can be taught in a quarter, whilethe person who understands your market but cannot run a deal will notbecome a closer through exposure. The exception is genuinelytechnical sales where the buyer will not engage with someone whocannot hold a specialist conversation. Even then, a rep who has soldcomplex multi-stakeholder deals in an adjacent sector adapts fasterthan one who knows your industry from a simpler sales model.
How manyinterview rounds should there be?
Three, plus thework sample. A screening call, a hiring manager interview, thesell-back exercise, and a final conversation with whoever else theperson will work with closely. Beyond four touchpoints you startlosing candidates without improving the decision.
Is it worthusing a recruitment agency for one sales hire?
It depends onwhether you can reach passive candidates yourself. With a strongnetwork, a recognised employer brand and someone who has time fordirect outreach, you may not need help. If you are entering a newmarket, hiring a profile you have not hired before, or you havealready run one failed search, the arithmetic changes quickly givenwhat a mis-hire costs. It is worth testing the market before youcommit to running the whole thing alone.
What do I doif the best candidate wants more than the range?
Decide beforeyou meet them what the ceiling is and why. If you break the band forone person you will do it again, and you will have a compensationstructure your existing team eventually discovers. If the marketkeeps telling you the range is wrong, that is real information, butchange the band deliberately rather than in a negotiation.
Every stageabove is useful, but they are not equally weighted. If you run amediocre interview against a precise, honest, specific roledefinition, you will usually still hire someone who works out. If yourun a brilliant interview against a vague one, you are choosingbetween people you cannot properly compare, and you will pick whoeverperformed best in the room. In sales, that is the one selectioncriterion guaranteed to be gamed.
Recent Blogs
Suggesting that the blog offers guidance for every stage of the job search.
.png)
.png)