Speeding Up a Hiring Process That Is Already Leaking
A regional credit union with 45 branches hires about 210 member-service representatives a year against total headcount of 1,050. Offer acceptance has fallen from 88% to 71% in 18 months; 90-day attrition in the role is 34%; time-to-fill averages 41 days; cost per hire is $4,300. Exit data show most early leavers cite unpredictable scheduling and "the job wasn't what was described." Branch managers, under staffing pressure, want to remove the post-secondary requirement and replace the structured behavioural interview with a 20-minute informal conversation, expecting offers to move faster. A candidate survey (n=180) shows the starting rate sits 12% below the local market midpoint and that candidates wait an average of nine days between final interview and offer.
1. Separate the two proposals from the branch managers and assess each on its merits. Which of the presented problems does each actually solve?
Read the worked answer
Treat them as two unrelated changes that have been bundled into one 'speed' story.
Proposal 1: replace the structured behavioural interview with a 20-minute chat.
- It does not address the speed problem. The delay candidates report is the nine-day gap between final interview and offer - an approval and scheduling problem, not an interview-length problem. Compressing the interview shrinks a component that is not the constraint.
- It does address something else, badly. Unstructured interviews are among the weakest predictors of job performance and increase susceptibility to similarity and affinity bias. The problem the organization most needs to fix is the 34% 90-day attrition - which is a selection validity and realistic-job-preview problem. This proposal makes it worse.
- Better route to the same goal: keep structure, compress the loop into one day, pre-delegate offer approval within a salary band, and commit to a 48-hour offer.
Proposal 2: remove the post-secondary requirement.
- Defensible if the requirement is not job-related and does not predict performance for a member-service role. It is low cost, widens the pool, and can reduce adverse impact.
- Defensible only after checking: is the requirement in the competency profile, has it been used as a proxy for something else, and does incumbent performance differ by credential? If the answer is no to all three, remove it.
- Do not remove it by decree on a busy Tuesday. Test it in a pilot and measure what happens.
2. Design the pilot. What will you change, how will you measure it, and what thresholds would make you stop?
Read the worked answer
Scope. A 90-day pilot in a matched pair of branch clusters (similar volume, similar vacancy rates), with the remainder as comparison. Change two things only: (a) the education requirement is removed for the pilot roles; (b) the interview loop is compressed to a single day with pre-approved offer bands. Do not also de-structure the interview - one variable at a time.
Metrics. Applicant volume and quality mix; days from application to offer; offer acceptance; 90-day attrition; six-month performance rating; manager satisfaction; cost per hire.
The honest measurement problem. Ninety days and two clusters will not produce a statistically meaningful read on 90-day attrition - the base rate is 34% and the cell sizes will be tiny. Say so up front and rely on leading indicators: acceptance rate, offer-to-start time, and structured reference/onboarding quality checks. Pre-register the claim you are making so the pilot is not retro-fitted to whatever happens.
Stop/go thresholds, set in advance. Proceed to full rollout if offer acceptance improves by at least 8 points and time-to-offer falls below five days with no deterioration in manager-rated new-hire readiness at 60 days. Stop and redesign if acceptance is flat while screening volume rises without conversion, or if early attrition indicators (no-shows, first-30-day absences) worsen.
Governance. Give branch managers a written commitment that the pilot answers their speed problem, and hold them to the measurement in exchange.
3. The CFO says a general market adjustment to close the 12% pay gap is unaffordable. What do you propose instead?
Read the worked answer
First put the current cost on the table, because the CFO is comparing an unknown cost against zero.
- 34% of 210 annual hires is roughly 71 early leavers. At $4,300 per hire, replacing them costs about $305,000 a year in recruiting alone, before onboarding, supervisor time and member-service disruption. That figure is a replacement cost the organization already pays, annually, invisibly.
Then separate the two failures - they need different fixes.
- Offer acceptance (71%) is driven by pay and by nine days of silence. Speed is nearly free. Fix it first.
- 90-day attrition (34%) is driven by schedule predictability and a broken realistic job preview - both cited in exit data. These are also largely free: publish schedules two weeks ahead, show real shifts and weekend rotation in the interview, let candidates shadow a rep for two hours.
Then target the money rather than generalizing it. Options in ascending cost:
- Raise the entry rate only, leaving the scale alone - cheapest way to fix acceptance.
- A shift/schedule premium for hard-to-fill branches, which buys the thing candidates actually name.
- A capped incumbent top-up for reps below a defined compa-ratio threshold, funded by carving 0.5 points out of the merit budget.
The argument to make: the CFO is being asked to fund a 12% lift across the population to fix a problem concentrated at the entry point and in scheduling. Cost the targeted package; it will be a fraction, and it addresses the actual causes.