Reference
Is our benefits program actually competitive?
The short answer
Benchmark on three axes rather than one: what you spend against comparable employers, what employees actually receive in plan value, and how the program performs on the dimensions your workforce cares about. Cost per employee alone is misleading, because two employers spending identically can deliver very different value depending on plan design, network fit and contribution strategy. The comparison set matters more than the data source — industry, geography, size band and workforce demographics all move the benchmark.
Why cost per employee is the wrong single number
It conflates three things: how rich the plan is, how much of it the employer funds, and how well the employer bought. An employer spending less may be buying better, funding less, or offering less — and those call for entirely different responses.
Separate them. Total plan value, employer contribution share, and rate competitiveness for the plan you have are three questions with three different remedies.
Choosing the comparison set
A national average across all employers is nearly useless. What matters is industry, region, size band and workforce demographics, because all four move both cost and expectation. A 400-person healthcare employer in South Florida and a 400-person software firm in the Triangle are not the same benchmark, whatever the headcount suggests.
Where a genuinely comparable set is unavailable, several directional sources exist and are worth reading as direction rather than as a target.
The axis most employers skip
What employees actually value. A plan that benchmarks well on cost and badly on the specific hospitals and specialists your workforce uses will read as a cut regardless of what it cost you. Network adequacy against where your people actually live is a solvable analysis, and it is rarely run before selection.
Comprehension is the other half. Employees make a consequential decision in twenty minutes, default to last year’s election, and discover the deductible in March. An employer can pay for the richer option and receive none of the credit.
When to run it
Roughly two quarters before renewal, so the findings can still change something. A benchmark produced alongside a renewal quote is a description rather than a lever.
Every two or three years is usually sufficient unless something has changed materially — a large hire wave, a new location, a demographic shift, or a second consecutive difficult renewal.
| Finding | What it means | Remedy |
|---|---|---|
| Spend high, plan value high | Buying a rich plan | Contribution or design, if cost is the concern |
| Spend high, plan value average | Buying badly | Market the program; check funding |
| Spend low, plan value low | Under-invested | A retention question, not a cost one |
| Spend average, poor network fit | Wrong plan for this workforce | Network analysis before renewal |
| Good plan, low perceived value | A communication failure | Year-round education, not one email |
Common questions
Where does benchmarking data come from?
How often should we benchmark?
Does a PEO benchmark for us?
Where this sits
This page supports Benefits Administration — the practice that does this work.
