See your savings

Free benchmark ยท All 50 states plus DC ยท Plan year 2026

Where does your plan cost sit against the market?

Not whether you're overpaying. Where you land in the distribution of what comparable groups actually filed.

How to use this

Four inputs, one percentile

  1. Pick your state. Small group rates are filed by state, so this sets the distribution you're measured against.
  2. Enter total monthly plan cost. Employer share plus employee share, medical and Rx only.
  3. Enter enrolled employees. The count on the bill, not headcount.
  4. Enter covered members. Employees plus every dependent on the plan.

The bar shows where you sit against filed 2026 small group premiums, in your state and nationally.

A high percentile isn't proof you're overpaying. An older census, heavy dependent enrollment or an active high-cost claimant can all price correctly and still land high. It tells you whether to ask, not what the answer is. Nothing you enter is transmitted.

Total plan cost means employer share plus employee share, medical and Rx only. Leave out dental, vision, life and EAP. If you're level funded, use the maximum funding rate.

The 30-second version

What am I actually looking at?

Six things worth knowing before you quote your percentile to anyone.

01

A position, not a verdict

It tells you whether the question is worth asking. It doesn't tell you what the answer is.

02  ·  BELOW THE 40TH

Competitive for your state

Look at other levers before you touch funding structure. The savings here are elsewhere.

03  ·  40TH TO 70TH

Normal range

Worth a conversation at renewal. It isn't an emergency, and it isn't a reason to shop the same plan harder.

04  ·  ABOVE THE 70TH

Worth understanding why

Find out what's driving it before you accept the next renewal, not after.

05

High can still be correct

An older census, heavy dependent enrollment or an active high-cost claimant all price up legitimately.

06  ·  THE COMMON TRAP

PMPM and PEPM aren't the same

A group with heavy family enrollment looks cheap per member and expensive per employee. Know which one you're quoting.

Built from 2026 small group rate filings submitted to CMS. All 50 states plus DC. Nothing you enter is transmitted.

READING THE RESULT

Where did you land, and what does it mean?

Most rate tools hand you a verdict. This one hands you a position, and that's deliberate.

BELOW THE 25TH

Cheap, and worth checking why

A young census, a lean plan design, or a carrier buying market share. A low rate on a plan nobody can afford to use isn't a win.

25TH TO 75TH

Ordinary for your market

Your rate reflects your census and your state. Savings here come from structure, not from shopping the same plan harder.

75TH TO 90TH

Something specific is driving it

Usually an older census, high dependent enrollment, or a rich plan design. Worth asking which one, and whether it's all three.

ABOVE THE 90TH

Where structure moves the most

Usually a claims event, a small pool, or a renewal that compounded for several years running.

The mistake that breaks every comparison

  • Per member and per employee are not the same number. People mix them constantly.
  • A group with heavy family enrollment looks cheap per member and expensive per employee.
  • The same group reads either way depending on which one you quote.
  • This tool benchmarks on cost per member per month, because that's the basis filed rates use.
  • Your per-employee figure sits alongside it, because that's your budget line.

One benchmark, two displays. Never two benchmarks.

Questions

What else should you know before you quote this?

What the data is, what counts as plan cost, and why a high percentile isn't a verdict.

A percentile tells you whether to ask. These tell you what to ask.

What would the individual market cost my people?

Lowest-cost Silver by county and age, plus household totals for spouse and children. All 3,137 counties.

Look up my county

What actually moves the number?

Six funding structures by group size and industry, and the cost levers each one makes possible.

Compare structures

Would a fixed budget beat a renewal?

CHOICE Arrangement, formerly ICHRA. How it works, when it wins, when it loses, and what can be bundled on top.

How it works

You have the position. Now the levers.

If you could change one thing about your health plan without changing your budget, what would you change?

A percentile tells you whether to ask the question. What moves the number is structure, and most employers have never been shown what theirs makes possible.

See what your structure unlocks

Methodology and important disclosures

What this tool is. A comparison of your reported plan cost against published market data. It is not a quote, an offer of coverage, a rate review, tax or legal advice, or a finding that any rate is incorrect. It does not evaluate your census, claims history, plan design or carrier. No carrier or plan is named, priced or recommended anywhere on this page.

Benchmark source. CMS Unified Rate Review Public Use File, plan year 2026, posted 2026-03-27. Small group market filings across all 50 states plus DC. The metric is projected total premium per member per month as filed. Distributions are weighted by projected member months. Percentiles are interpolated between the 10th, 25th, 50th, 75th and 90th points of each state's distribution.

Limits. Filed premiums describe the fully insured small group market and are not adjusted for plan design, actuarial value, census age, dependent enrollment or claims experience. States with fewer filings produce thinner distributions, which is why the national percentile is shown alongside the state percentile. A percentile is a position in a distribution, not an assessment of whether a rate is appropriate for a specific group.

Your inputs. Figures you enter are processed in your browser. Nothing is transmitted, stored or retained by this page.

Insurance placement. Any insured products referenced elsewhere on this site are marketed and offered through licensed agents in states where licensed. Save with Benefits provides education and program design. Last reviewed September 2026.