The Retention Business Case
A 1,200-employee distribution company has 260 voluntary leavers a year. Average salary is $58,000 and the benefits load is 22%. Recruitment cost per hire averages $3,800 and onboarding plus initial training averages $2,100 per new hire. Time-to-fill averages 55 days, during which HR estimates 40% of the role's output is lost; a working year is 260 days. HR proposes a $180,000 retention program projected to cut voluntary turnover by 20% relative to today. The CFO wants a business case with the arithmetic shown before the next quarter, and has said plainly that she does not accept 'soft costs' without justification.
1. Calculate the annual cost of voluntary turnover. Show hard and soft components separately and state your assumptions.
Read the worked answer
Hard (cash) components - 260 leavers:
- Recruitment: $3,800 × 260 = $988,000
- Onboarding and initial training: $2,100 × 260 = $546,000
- Hard subtotal: $1,534,000
These are defensible without argument: they appear in budget lines and they recur every time a seat is refilled.
Soft component - vacancy productivity loss:
- Daily salary: $58,000 ÷ 260 = $223.08
- At 40% output loss: $89.23 per day
- Over 55 days: $4,908 per leaver
- × 260 = $1,276,000
Total: approximately $2.81 million, of which $1.53 million is hard.
Assumptions to state explicitly, not bury:
- All 260 leavers are replaced, and every replacement incurs the full recruitment and training cost.
- Vacancy productivity loss is not fully additive - some output is absorbed by overtime, by redistribution to colleagues, or is simply lost revenue that never appears in the accounts. This is the number the CFO will attack, correctly.
- Salary is used unloaded. Loading at 22% would raise the vacancy component to roughly $1.56 million and the total to about $3.09 million - present it as a sensitivity, not as the headline.
- Turnover is treated as homogeneous. It is not: regretted and non-regretted leavers have very different costs, and the retention program will only prevent regretted ones.
The professional discipline: lead with $1.53 million. It is unarguable, and the case still works.
2. Compute the payback period and ROI on the $180,000 program under a hard-cost-only view and a total-cost view. What do you recommend?
Read the worked answer
Avoided leavers: 20% of 260 = 52.
Hard-cost-only view:
- Avoided cost: $1,534,000 × 0.20 = $306,800
- Program cost: $180,000
- Net benefit: $126,800
- ROI: 126,800 ÷ 180,000 = 70%
- Payback: 180,000 ÷ 306,800 × 12 = about 7 months
Total-cost view:
- Avoided cost: $2,810,000 × 0.20 = $562,000
- Net benefit: $382,000
- ROI: 212%
- Payback: about 3.8 months
Recommendation: present the hard-cost case as the decision basis and the total-cost case as upside. A 70% first-year return with a seven-month payback is a good investment on numbers the CFO cannot dispute. Leading with 212% invites her to dismantle the vacancy assumption and, having won that argument, to discount the whole case.
The bigger risk in the proposal is not the cost side - it is the 20%. That projection is the load-bearing assumption. Ask what it is based on. A 20% relative reduction means turnover falls from 21.7% to 17.3% - plausible for a well-targeted program, implausible for a generic engagement initiative.
3. The CFO pushes back: 'Half of these people would have left anyway, and I don't believe the vacancy cost.' How do you respond?
Read the worked answer
Concede the vacancy point immediately - she is substantially right. Vacancy loss is partly absorbed by overtime, by redistribution, and by work simply not being done. That is precisely why the hard-cost view carries the recommendation. Conceding the weakest assumption early buys credibility for the ones you will defend.
Then answer the 'they'd have left anyway' challenge, because it is the more important one. It is true that some turnover is unavoidable and even healthy. The response is not to argue - it is to segment. Split the 260 into regretted and non-regretted leavers, by performance rating and by whether the role was refilled. If, say, only 120 are regretted, the addressable cost is roughly half, and the program's target should be expressed against that base, not against all 260. That is a smaller prize, honestly stated, and far harder for the CFO to dismiss than an inflated one.
Then ask what would change her mind. In practice: a costed pilot on two depots with a control group, measured on 12-month regretted attrition, with a decision gate at six months and a defined exit. That converts an argument about assumptions into an experiment with a price.
Also offer the counterfactual she should care about. The $180,000 is discretionary; the $1.53 million is not - it is already being spent. The real question is not whether to spend on retention but whether to keep buying replacements at the current rate.
Document the assumptions in the business case itself, with the sensitivity range, so the number is auditable rather than asserted.