White Paper · Retention Economics

Executive summary

Turnover is one of the largest costs most employers never put on a budget line. It does not arrive as an invoice. It arrives as a position posted, a manager pulled off the floor to interview, a new hire learning the job at half speed, and a schedule run short in the meantime. The cost is real, it recurs, and in the industries that run on hourly and variable-hour staff it recurs constantly.

It can be managed, and the lever is more affordable than the cost it offsets. Workers consistently rank health coverage as the benefit that most influences where they work and whether they stay. The employer who extends coverage the workforce values reduces the rate at which people leave, and every departure prevented is a replacement cost avoided.¹

This paper sets out that math in four parts: what a single departure actually costs, where the cost concentrates in federal quits data, how an accessible benefit reaches the variable-hour workforce conventional coverage screens out, and the return calculation that weighs benefit cost against avoided replacement cost. The figures that size the cost are third-party estimates, and the paper says so plainly; the figures that size the churn are federal data.²

The direction the math points is not in dispute, and the threshold the benefit has to clear to earn its cost turns out, on the employer's own arithmetic, to be low. That is the whole argument: a modest, accessible benefit applied where churn is worst is, more often than not, the cheaper side of the ledger.

Section 1 · The cost

What a departure actually costs

The cost of losing an employee is larger than the cost of hiring one, and the gap is where most of the money hides. The visible cost is recruiting: the posting, the screening, the interviews, the background check, and any agency fee. A widely cited SHRM figure puts the average direct cost per hire near 4,700 dollars, a third-party estimate, and for a finance leader it is the least of it, because it captures only the part of the loss that generates a receipt.³

None of these four lands on an invoice, and that is exactly why turnover is underestimated. A recruiting fee lands in a specific account a finance leader can see; the vacancy, ramp, team-load, and knowledge costs diffuse into lower revenue, overtime, and reduced output, showing up as a general softness rather than a line anyone can point to. A cost that is never named is a cost that is never managed.⁴

Section 2 · The churn

Where the math hits hardest

Turnover is not distributed evenly. It pools in the industries built on the hourly and variable-hour workforce, and federal data measures the pattern directly. Through its Job Openings and Labor Turnover Survey, the Bureau of Labor Statistics publishes the quits rate, the share of employment that voluntarily leaves each month, by industry. Across the whole economy in early 2026 it sat near 1.9 percent; in accommodation and food services it ran more than double that.⁵

The two facts multiply. Replacement cost is a per-departure number; the quits rate is how often departures happen. An employer in a low-churn industry pays the per-departure cost rarely; an employer in accommodation and food services pays it again and again across the year, on a workforce already large relative to revenue.⁶ Both terms are high at once, which is why a retention lever that would barely register in a low-churn business can change the economics of a high-churn one.⁷

When departures happen matters as much as how often. In these industries a large share of separations occur in the first weeks or months, before the worker has returned the cost of hiring and training them, and an early departure is the most expensive kind. A benefit that takes effect immediately, with no eligibility waiting period, aims directly at that early-departure window, so the employer recovers a hiring investment it would otherwise write off.

Section 2 · The churn, by vertical

Nine industries, one shape

The pattern holds across the specific industries this firm serves, though the data that documents it differs from one to the next. Where the federal quits series measures the vertical directly, it is shown as a monthly rate; where it does not, the documentation is the sector's own annual research, identified as such. The two measures are never equated.

Across all nine the shape is the same even where the numbers differ. The industries that run on hourly, contingent, and variable-hour staff carry both the highest voluntary churn and the thinnest benefit offers, and the departures are overwhelmingly worker-initiated, which is the turnover a benefit can actually move.

Section 3 · The lever

How benefits change the retention equation

A retention lever only works if it reaches the people doing the leaving and moves their decision to stay. Health coverage does both. A Randstad US study found 66 percent of workers regard a strong benefits package as the largest single factor when they weigh a job offer, and 61 percent would accept a lower salary for better benefits; asked which benefit mattered most, 75 percent named health insurance ahead of every other category.¹²

The difficulty is that conventional coverage does not reach this workforce. Major-medical eligibility is keyed to full-time hours, which screens out the part-time and variable-hour staff who make up much of a high-churn payroll. An employer can offer a rich major-medical plan and still leave most of its hourly workforce uncovered, so the retention lever never touches the people most likely to leave.¹³

• Optiv AccessTier 1 · funded layer
  For the contingent and very-low-hour staff who cannot make a meaningful pre-tax election, an employer-funded layer delivers day-one virtual care and prescription access on eligibility terms the employer sets. No Section 125 election, no payroll-tax recovery, no take-home effect for the worker.
• Optiv AdvantageTier 2 · §125 supplemental
  For the employees who can elect it, a supplemental benefit funded through a Section 125 cafeteria plan that sets its own eligibility class. Guaranteed issue, no waiting period, pairing around-the-clock virtual care and a 400-plus-medication prescription benefit with a fixed indemnity benefit on a substantiated medical event.

Together, the two layers put a usable benefit in front of essentially the whole workforce: Access for the remainder, Advantage for the employees who can elect it. The retention math depends on that completeness, because a benefit that reaches only part of the workforce reduces turnover only in that part.¹⁴ And the effect reaches beyond the individual worker: in high-churn industries information about employers travels fast, so an employer known for real coverage shortens its time to fill and raises its applicant quality, lowering recruiting cost on top of the retention effect.

Section 4 · The distinction

Which turnover a benefit can actually reduce

An honest return calculation has to be precise about which departures a benefit can prevent, because not all turnover is the same and not all of it is addressable. An employer that credits a benefit with reducing every separation will overstate the return and lose the argument the first time a skeptical controller examines it.

The Bureau separates total separations into three categories; the quits rate isolates the departures a worker chooses to make and a better offer might change.¹⁵

Within voluntary quits, a further distinction sharpens the picture. Practitioners separate regrettable turnover, the loss of a worker the employer would have kept, from non-regrettable turnover. A benefit's value concentrates in preventing regrettable departures, and a calculation that counts only those as addressable is the one a finance leader should trust, because it does not claim the benefit can do something it cannot.

Section 5 · The return

Calculating the return against replacement cost

The retention case becomes a financial case when the cost of the benefit is set against the replacement cost it avoids. The cost side has two components, and an honest calculation keeps them separate.

The Section 125 supplemental layer generates a payroll-tax recovery, because the amount a worker elects pre-tax is excluded from the Social Security and Medicare wage bases, which for 2026 carry a combined employer rate of 7.65 percent up to the Social Security wage base. That recovery offsets part or all of the employer's cost of the supplemental layer. The funded layer, by contrast, is a flat employer cost with no recovery, so it enters the math as a cost the employer chooses to absorb to cover the workers the supplemental layer cannot reach.¹⁶

Section 5 · The break-even

How few avoided departures it takes

Because a single avoided departure saves a substantial fraction of a worker's annual pay, a benefit only has to change a modest number of stay-or-go decisions to cover its own cost. The illustration below shows the shape of the calculation. It is not a projection.

The durability is the point. Even a deliberately conservative version, turnover of 30 percent rather than 50, replacement cost held at the low-end 50 percent of pay, and benefit cost doubled to $1,200 per employee, pays for itself by preventing roughly seven of 30 departures, a reduction of less than a quarter of the voluntary turnover. The case rests on the structural gap between a cost measured per departure and a cost measured per employee, not on an optimistic multiplier.