Most of the people I talk to land in one of two buckets, and they both think they're describing the same problem: "hiring is broken."
Bucket one has never hired a salesperson before. No recruiter, no playbook, no idea what "good" looks like beyond a gut feeling. They're doing product, support, and sales themselves, and now they're supposed to also interview, coach, and manage someone into a role that doesn't have a defined process yet. They don't actually know if they need a closer or someone to just qualify leads and get demos on the calendar. Nobody's told them the honest answer: hiring an AE before you have a repeatable motion just means two people flying blind instead of one.
Bucket two has a recruiting team. HR sources, screens, and forwards a shortlist. On paper, this should be the easier path. In practice, the sales leader keeps getting profiles that technically match the JD and still feel wrong. HR is screening through a different lens than the one the sales leader actually needs. One TA lead I know is the only recruiter across 40-50 open roles company-wide and gets through maybe a fifth of the applications that come in. Volume isn't the problem there. Calibration is.
Different starting points. Same ending: a role stays open for months, and nobody can point to the exact moment it went wrong, because it didn't go wrong once. It went wrong six times, in six different places in the funnel. Here's where.
It starts with the JD, and the JD is usually wrong
Most sales JDs describe activity, not outcomes. "2-5 years of B2B sales experience, strong communication skills" tells you almost nothing about what the person will actually be doing on day 30. I've seen a JD written for what was, on paper, a mid-level AE. The actual job for the first ninety days was qualifying inbound leads and getting meetings booked. Not the same role. Not the same person. The company hired for the title on the org chart instead of the work that needed to happen first.
This isn't a wording problem, it's a thinking problem. Before a JD gets written, someone needs to answer questions most companies skip:
What exact part of the funnel will this person own in the first 90 days?
Who's the buyer, and what industry, geography, or deal size do they need real experience with?
What does success look like in hard numbers: meetings booked, pipeline generated, deals closed?
Is there already a working sales motion, or is this person supposed to build one from nothing?
What's non-negotiable, and what's trainable?
Skip this, and every downstream step, sourcing, screening, interviewing, inherits the confusion.
Sourcing looks productive. Mostly it's noise reduction, not quality control
Naukri, LinkedIn, referrals, inbound. Every channel promises signal and mostly delivers volume. I've watched a search start at 260 applicants and get filtered down to 46 through paid search tools, and every one of those 46 still needed a human to check tenure, actual company relevance, and whether "5 years in sales" meant five years of quota-carrying work or five years of adjacent activity. LinkedIn's version of filtering, asking applicants to write a note to the hiring manager, knocks out 80-90% of people on effort alone, which is useful, except it also means you're now reading AI-generated notes from the remaining 10-20%, and you have to interrogate those too.
CVs, generally, are not reliable evidence anymore. In one process I watched closely, the majority of candidates who looked strong on paper couldn't clearly explain their own numbers when asked: what their quota was, how pipeline moved, what "achieved 150% of target" actually meant in rupees and deals. The CV said one thing. The person, under two follow-up questions, said something else. That gap is the entire job of screening, and most sourcing channels aren't built to catch it. They're built to collect applications, not validate claims.
There's a second layer here that a CV never surfaces at all: sales motion match. A CV tells you where someone worked and what number they hit. It doesn't tell you whether the sales cycle at that company looks anything like yours. Someone who crushed quota closing a self-serve product in two-week cycles hasn't necessarily done anything close to a six-month enterprise deal with procurement and legal in the room. Someone who's only ever sold to a single decision-maker hasn't had to navigate a buying committee. I've seen candidates with genuinely strong logos on their CV fail completely in a new role, not because they can't sell, but because they'd never sold this particular way before. The CV says "5 years of B2B sales." It doesn't say which kind.
Screening and scheduling: this is where good candidates quietly disappear
Assume you've sourced well. Now the candidate has to actually get through the process, and this is where I've seen the most silent attrition. In one funnel I tracked closely, well under half of the candidates who were sent an assessment ever completed it, and a chunk of those weren't lost to disinterest. They were lost because nobody was tracking who'd been sent something and never followed up.
The other failure mode is worse, because it's entirely internal: candidates get shortlisted and then just sit. I've seen four strong AE candidates wait for over a week because the stakeholder who needed to review them was busy with something else. Twenty-plus shortlisted profiles sitting unseen in a dashboard nobody was checking. Five candidates waiting on scheduling with no one clearly owning the next step. None of this is a sourcing problem or a candidate-quality problem. It's an internal ownership problem, and it kills exactly the candidates you worked hardest to find, because good people don't wait around wondering if anyone's going to get back to them.
The shortlist mismatch: recruiter and hiring manager, looking at the same resume, seeing different things
Even when a shortlist gets delivered on time, it can still be wrong. This is the recurring pattern: a recruiter (internal or agency) screens for what looks defensible on paper, years of experience, company pedigree, a clean CV, and the hiring manager is screening for something much harder to write down: can this person actually sell, in this specific market, to this specific buyer. Those aren't the same filter. I've had hiring managers tell me, almost word for word, "these all technically match what I asked for, and none of them are who I need." That's not a communication failure between two people who don't understand each other. It's two genuinely different definitions of "qualified," and nobody catches the mismatch until three or four rounds of it have already happened.
Then come the interview rounds, and the amnesia between them
Once a candidate clears the shortlist, the next failure mode is internal: too many rounds, and no shared memory between them. Each interviewer asks their own version of "tell me about your sales experience" without knowing what the last three people already covered. The candidate repeats themselves. The interviewers don't calibrate against each other's notes because there usually aren't shared notes, just a vague verbal "yeah, I liked them" passed along in a hallway or a Slack message.
This isn't just inefficient, it's expensive. Candidates who are actually good have other conversations happening in parallel, and a process that takes two to four weeks to reach the actual decision-maker will lose people to companies that decided faster, not because the other offer was better, but because it existed first. I've seen strong candidates accept somewhere else purely because nobody on the hiring side moved for three weeks. The fix isn't more rounds, it's fewer rounds with a real division of labor: each interviewer assigned a distinct thing to probe, notes that actually get shared, and a scorecard instead of a gut-feel debrief.
And then, after all of that: the offer, and the ghost
You'd think reaching the offer stage means the hard part is over. It isn't. I've watched a single sales role stay open for six months despite three or four offers going out, every one of them declined or gone unanswered. Sometimes it's a live competing offer the candidate didn't disclose until the last minute. Sometimes it's a compensation gap that was never actually negotiated, a candidate asking for two lakhs more than the offer, and the conversation just stalling instead of closing. Sometimes it's nothing to do with the offer at all: a candidate falls sick the day before joining, or another accepts and then withdraws two minutes before a scheduled call, no warning.
The pattern underneath all of it: most hiring processes have no explicit ownership of the gap between "offer sent" and "candidate joins." Everyone treats the offer as the finish line, when it's actually just the start of the part of the process with the least structure and the most silent risk: competing offers, notice periods, relocation, second thoughts. A process that spent weeks carefully screening and interviewing usually has no equivalent plan for the two weeks after the offer goes out. That's exactly backwards.
The honest version of the funnel
Strip away the job titles and the tools, and this is what actually happens to a sales hiring process, in order: the JD is vague or wrong before anyone applies. Sourcing channels filter for effort and pedigree, not selling ability or sales motion fit. Good candidates quietly drop out of screening because no one's tracking who's stuck. Shortlists get built against the wrong definition of "qualified." Interview rounds repeat themselves because nothing's shared between interviewers. Momentum dies in two-to-four-week gaps, and candidates leave for whoever decided first. And the offer, treated as the end, is where some of the most avoidable losses happen, because nobody owns what happens after it's sent.
None of these are candidate problems. They're process problems, wearing a candidate-shaped disguise. Everyone I've described above, in either bucket, thinks they have one hiring problem. What they actually have is six unrelated problems stacked on top of each other, and no visibility into which one is actually costing them the hire.
Where in this does your process actually break?