The fix
Why this is happening — and what to do about it.
Operators tell us "we cannot find call center agents" roughly once a week. We have learned to ask one follow-up question before saying anything else: "How many people apply, and how many pass the phone screen?"
The answer almost always reveals the actual problem. Applications are usually fine — sometimes higher than they were two years ago. The breakdown is downstream: most candidates do not survive the phone screen, the few who do drop off in offer stage, and the cohort that finally makes it to nesting attrits at 30 to 40 percent before they ever touch a live queue. That is not a sourcing problem. That is a screen problem dressed up as a sourcing problem.
The "sourcing pool" myth
There is a story you will hear in almost every operations meeting right now: the labour market has changed, the agents are not out there, the gig economy is eating the pool, the new generation will not work phones. Some of that is true at the margins. None of it is the reason your funnel is broken.
We staff 14 countries and we run live engagements every week. The pool is not the problem. The pool is bigger than it was five years ago in most of our regions. What has changed is that the people in the pool are harder to evaluate with a 1990s phone screen — and most call center recruiting funnels are still using a 1990s phone screen.
When operators say "the pool has changed," what they almost always mean is "our funnel stopped working." The fix is not to mourn the pool. The fix is to rebuild the screen.
Why generalist firms can't fix this
A general staffing firm will respond to the "we cannot find agents" complaint by widening the funnel — more job boards, broader sourcing parameters, lower keyword bar on resume parsing. Volume goes up at the top, conversion stays flat, and three months later the operator is running another search.
Specialist call-center recruiters do the opposite. We narrow the funnel by sharpening the screen, because we know the QA scorecard, we know the call types, and we know what tenure looks like in the role. The result is a smaller shortlist, but with higher conversion at every downstream stage. Operators routinely tell us they receive fewer candidates from us and place more of them — because the candidates are right.
If your current staffing partner cannot tell you what your specific QA scorecard rewards and how their screen reflects it, they are running a generalist screen against a specialist role. That is the structural problem.
What recruiting against a QA scorecard looks like
The single highest-leverage move in fixing a broken screen is to align the screen with the QA scorecard your floor actually uses. Most recruiters interview for resume keywords, availability, and a fuzzy notion of "customer-service mindset." QA teams score against empathy, problem identification, ownership, voice and tone, and call control. Those are very different rubrics.
When the recruiting screen and the QA scorecard agree on what "good" looks like, the predictive validity of the screen jumps dramatically — agents who pass the screen pass QA, agents who pass QA stay on the floor, and the funnel starts working again.
- Pull the actual QA scorecard your floor uses. Read it with your recruiting team, line by line.
- Identify the three to five behaviours QA scores most heavily. These become the screening criteria.
- Rewrite the phone-screen guide so every question maps to one of those behaviours, not to a resume bullet.
- Score every candidate on the same rubric, with the same weighting QA uses on a live call.
- If your QA team and your recruiting team have never met, they should. They are screening for the same job — they should be using the same definition.
Voice samples and scenario tests
Resumes do not predict who can hold a courteous, clear conversation under mild pressure. Voice samples and scenario tests do — and they are the cheapest screening tool you have.
A two-minute recorded voice sample, captured during the screen, eliminates more bad hires than any other single intervention. It tells you tone, clarity, energy, and basic call-handling instinct. It costs nothing. And almost no in-house funnel uses one.
A short scenario test — "a customer calls in upset that their order has not arrived, walk me through the first thirty seconds" — tells you problem identification, ownership, and empathy in 60 seconds. Score it against the QA rubric. The candidates who score above 4 out of 5 typically pass nesting at 80+ percent. The ones below 3 usually wash out before day 30.
Fix the screen, not the volume
The temptation when you cannot find call center agents is to spend more on top-of-funnel ads. In our experience this is almost always the wrong move. More volume against a broken screen produces more bad hires faster — which raises attrition, which empties the floor again, which raises the pressure to lower the bar further.
The right move is the opposite. Tighten the screen. Add the voice sample. Add the scenario. Score against QA. Watch your conversion rates per stage, not your top-of-funnel volume. Operators who make this shift typically see phone-screen pass rate climb from 18 percent to 35 percent inside two cohorts — because the screen is now selecting for behaviours that survive the floor instead of resume keywords that do not.
The way we do this every day is documented in /services/call-center-recruitment, and the underlying philosophy is in /how-we-work.
Why retainer firms drop the bar in week 4
One pattern worth naming: retainer-based generalist recruiting firms have a structural incentive to lower the bar around week 4 of an engagement. The fee has been paid. The hours are sunk. The path of least resistance is to push more candidates through and call the engagement complete.
We have seen this pattern enough that we built our model around it. We charge per agent, not per retainer, so our incentive aligns with yours: the agent has to make it onto the floor and through guarantee for us to be paid in full. That is the model in /why-us if you want the longer read.
