Most employers do not choose a medical plan. They inherit one, then renew it eleven more times. The decisions that actually move cost and satisfaction — how the plan is funded, whose network it uses, where the deductible sits, and what an employee pays for family coverage — usually get made once, under time pressure, and then quietly repeat.
If the plan types are new to you, start with what an HMO is and what EE, ES and EC coverage tiers mean.
This guide walks those decisions in the order they matter, rather than the order a renewal packet presents them. It is written for Michigan employers of roughly ten to five hundred employees, where every option below is genuinely open.
Start With the Funding Model
Funding comes first because it determines what every later number means. There are three realistic structures.
Fully insured. You pay a fixed premium and the carrier keeps both the risk and the claims data. Simple, predictable, opaque. In Michigan’s small group market the rate is community-rated, which means your own claims history does not set your price — protection if your group runs hot, a penalty if it runs cold.
Level-funded. You pay a fixed monthly amount that splits into claims funding, administration and stop-loss premium. If claims land under projection, some of the surplus can come back. You see claims data. Underwriting is real, so a group with known large claims may not be offered it at a rate that makes sense.
Self-funded. You pay claims as they arrive, buy stop-loss to cap the downside, and hire an administrator. Cash flow varies month to month. In exchange you stop paying the carrier’s risk margin and premium tax, and the plan becomes yours to design.
A rough rule: under about fifty employees the real question is fully insured versus level-funded. Over about a hundred, self-funding deserves a serious look. In between, the answer depends more on your cash position and risk tolerance than on your headcount.
The Network Decides More Than the Deductible
Employees experience the network, not the funding model. A plan that saves eight percent and drops the health system where half your staff already receives care will be remembered as a pay cut.
Before comparing plan designs, map where people actually live and who they actually see. In Michigan that question has a different answer in Grand Rapids than in Macomb County, and a different answer again for an employer with a plant downstate and an office up north. Ask each carrier for a disruption analysis against your own enrollment file. A carrier that will not run one is telling you something.
Narrow and tiered networks are real tools, not traps — but they only work when the narrowing misses your employees. That is an empirical question, and it is answerable before you sign.
Plan Design: The Four Numbers That Matter
Strip a plan summary down and four numbers carry most of the employee experience: the deductible, the coinsurance split, the out-of-pocket maximum, and the prescription tiers. Everything else is detail.
For 2026, a non-grandfathered plan cannot set an out-of-pocket maximum above $10,600 for self-only coverage or $21,200 for anything other than self-only. If you want the plan to be HSA-qualified, it has to sit inside a tighter box: a minimum deductible of $1,700 self-only or $3,400 family, and an in-network out-of-pocket maximum no higher than $8,500 or $17,000.
Those HSA limits are worth knowing before you fall in love with a design, because a plan that misses them by fifty dollars is not a qualified plan and no one can contribute to an HSA against it. The 2026 contribution limits are $4,400 for self-only and $8,750 for family, plus a $1,000 catch-up at age fifty-five.
The prescription tiers deserve more attention than they usually get. Specialty drugs are where a mid-size employer’s trend now comes from, and two plans with identical medical designs can differ sharply once you read how each one handles specialty fills.
Contribution Strategy and the Affordability Test
What you charge employees is a separate decision from what you buy, and for employers with fifty or more full-time equivalents it is also a compliance test.
For plan years beginning in 2026, coverage is affordable if the employee’s cost for the lowest-priced self-only option that meets minimum value does not exceed 9.96% of income. For 2027 that figure rises to 10.22%. Because household income is unknowable to an employer, the test is run against one of three safe harbors: W-2 Box 1, rate of pay, or the federal poverty line.
Two things employers regularly get wrong here. The test looks only at self-only cost, so a generous single contribution and a punishing family contribution still passes. And it looks at the lowest-priced minimum-value option, so adding a cheaper plan to the menu can fix an affordability problem without changing anything else.
Passing the test is not the same as designing well. A family contribution that quietly pushes people onto a spouse’s plan is a retention problem the compliance test will never flag.
What to Have in Hand Before You Go to Market
Most of the delay in a benefits marketing exercise is not the carriers. It is waiting on the employer’s own documents. Four things do almost all the work:
- The renewal letter, in full, including the rate pages
- The current plan summary or schedule of benefits
- The contribution split by tier, as it actually runs in payroll
- Enrolled counts by tier
If the group is fifty or larger, add large-claim information and, where the funding allows, claims experience. Nothing else is needed to get a real quote, and every additional document a broker asks for is a reason the process takes another week.
Running the Renewal on a Real Timeline
A renewal is not an event that arrives in the mail. By the time the letter reaches you, the claims that set it have already been paid, and the questions worth asking were worth asking months earlier.
Working backward from an effective date, a sane schedule looks like this: data assembled and a decision made about whether to test the market by 120 days out; the market tested and proposals in comparable form by 90 days; a decision made by 60 days; and enrollment communication running by 30 days. Compressing that is possible, and it costs leverage.
Going to market every single year is not automatically the right answer either. A group that shops annually teaches carriers that its business is available cheaply and teaches employees that the network is unstable. The judgment about whether to go out is part of the advice, not a formality that precedes it.
Questions We Get
Is a lower premium always the better deal?
No. Premium is one of four costs in a medical plan; the others are the deductible and out-of-pocket exposure employees carry, the disruption cost when a network changes, and the administrative cost of moving carriers. A plan that saves six percent on premium and moves two hundred employees off their current physicians has not saved anything a CFO would recognise a year later.
How much can we actually influence a renewal?
More than most employers assume, and less than a marketing exercise implies. On a fully insured small group in Michigan, the rate is community-rated and largely set before you see it, so the leverage sits in plan design and contribution strategy. On a claims-rated or level-funded group, the leverage is real and it comes from understanding your own claims before the carrier explains them to you.
What is the difference between level-funded and self-funded?
Both make you responsible for claims rather than buying a fixed premium. Level-funded packages that responsibility into one steady monthly payment with stop-loss built in and a possible surplus return; self-funded means paying claims as they come, buying stop-loss separately, and managing the cash flow yourself. Level-funded is the on-ramp; self-funded is the destination.
Do we have to offer coverage at all?
The employer mandate applies at fifty full-time equivalents. Below that there is no requirement to offer a plan. But several other obligations start well before fifty and are often missed: COBRA at twenty employees, the Form 5500 at a hundred participants, and a Summary Plan Description from the very first covered employee.
What makes a plan HSA-qualified in 2026?
A deductible of at least $1,700 for self-only coverage or $3,400 for family, and an in-network out-of-pocket maximum no higher than $8,500 or $17,000. Missing either figure disqualifies the plan entirely, and no one enrolled in it can contribute to an HSA.
How is affordability measured?
For plan years beginning in 2026, the employee cost for the lowest-priced self-only option that meets minimum value must not exceed 9.96% of income, measured through a W-2, rate-of-pay or federal-poverty-line safe harbor. It rises to 10.22% for 2027. The test ignores family cost entirely.
When should we start the renewal?
About 120 days before the effective date, which is earlier than most employers start. That is the point at which you still have time to assemble data, decide whether to test the market, and get proposals into comparable form without making the decision under deadline.
Should we go to market every year?
No. Shopping every year signals to carriers that your business moves on price alone, and it puts employees through repeated network changes. Test the market when something has changed — the renewal, the workforce, the funding options available to you — and be able to say why.
General information for Michigan employers, not legal or tax advice. Plan-specific questions belong with your counsel or accountant, and we will bring them in.
Related Reading
Unhappy with your current broker? Switch to CFH. Your employees won’t notice. You will.
