One plan, start to finish

A health plan decision is supposed to look like any other six-figure purchase: a market check, a written comparison, a file. Most do not, because the renewal letter arrives as a single number from a single carrier through a single broker, and the only alternative on the table is to ask for a smaller number. This document follows one company through the seven steps of my process, from the public filing that started it to the second-year renewal that needed one page.

Who the company is. A 160-employee Texas mechanical contractor (composite): about 270 people on the medical plan, one carrier since the first filing on record, January 1 renewal. A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. Its filing history follows patterns common in those filings without copying any one of them.

What is modeled and what is not. Every figure about the company is modeled: the filing history (carrier tenure, employees, lives, premium, broker pay, by year), the census, the renewal letter, the claims, the quotes, the decision and the dates. The filing history is built in the shape of a Form 5500 Schedule A, and everything after it is built from that history and from the 52-company peer set, so that a CFO can check the arithmetic. Peer medians are from the 52 filings themselves. Each section says in its first line what it adds. The process, the deliverables, the timing and the fee are exactly as I run them.

The seven steps, one section each

What a CFO should take from it

• In this model, the renewal moved $140,000 the day the group went to market, and nothing about the group had changed.
• The structure that won had a maximum annual cost below the current year, and it had never been offered.
• The broker’s pay was the same number in every quote, disclosed before any quote existed. The incumbent’s would have risen with the renewal.
• The second year needed one page. That is the point: the first year does the heavy lifting, and after that the file does.

Every page ends with the paragraph the CFO used to tell the CEO. If you can retell a decision in five sentences, you understood it; if you cannot, ask more questions before you sign.

Want this run on your plan?

The first step has no fee and asks for nothing: send me the company name and I will read your public filing the way step 0 reads this one. It uses only the public record, so there is nothing to sign. If the numbers say the renewal is worth a look, the next step is the Renewal Snapshot, $0. Nothing paid, and nothing using your documents or claims, starts before the engagement letter and broker-of-record letter are signed.

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. The filing history is modeled in the shape of a Form 5500 Schedule A filed with the U.S. Department of Labor; peer medians are from the 52 filings. No client is identified. Advisory illustration, not legal or tax advice.

Step 0: the public filing, before anyone talked

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. This step shows its filing history in the shape of a Form 5500 Schedule A: employees, covered lives, premium and broker pay, by year, and the date the read was sent.

The company is a Texas mechanical contractor with about 160 employees. It has filed a Form 5500 with a medical Schedule A every year since the 2022 plan year. This is the whole of what the filing shows, and it is what the CFO received by mail on October 20, 2026, 73 days before the January 1 renewal, with no request attached.

Three things a CFO can see without asking anyone:

1. Since 2022 the group has grown about 23% (about 130 to 160 employees) and its covered lives about 17% (about 230 to 270), but the premium grew about 58%. Per covered person the plan went from $4,565 to $6,148, up about 35% in two years, roughly 16% a year. Per employee it now runs about 70% above the peer median (about $10,375 against $6,094). The mix got lighter, from about 1.77 covered lives per employee to about 1.69, so growth explains less than half of the increase. Price explains the rest.
2. One carrier, one broker, every filing. The group is now a different risk than it was at about 130 employees, and it has never been priced by anyone but the incumbent.
3. Broker pay edged down as a share of premium, from 5.8% to 5.4%, but rose about 47% in dollars, from $60,900 to $89,640, because it is a percentage of a premium that grew about 58%. Nobody decided to pay the broker $89,640 a year; it happened.

What the filing cannot show: the renewal letter, the claims, who is on which tier, what employees pay, or whether the broker has ever put the group to market. Every one of those is a document the CFO already has or is entitled to.

What the CFO told the CEO

Our health plan filing is public and I read it. We paid about $1.66 million in 2024 to one carrier we have never priced against anyone else, and the broker’s pay has risen almost half in two years with the premium without anyone deciding that. I want to find out what the renewal looks like against the market before we sign it. It costs nothing to look.

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. The filing history is modeled in the shape of a Form 5500 Schedule A filed with the U.S. Department of Labor; peer medians are from the 52 filings. No client is identified. Advisory illustration, not legal or tax advice.

Steps 1 and 2: the conversation, then the paperwork

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. New in this step: the dates and the conversation.

The CFO called on October 27, a week after the filing review arrived. Forty minutes. No quotes were discussed, because there was nothing to quote from. Two questions came up that the public record had already answered (who is the broker, what are they paid) and one it had not: what the renewal letter said. It said 11.8%.

Three documents were signed on November 3, before any work on the company’s documents began:

1. Engagement letter. Scope, the two deliverables, and the Cost Review fee: $7,500 for the standard scope, $12,500 for the extended scope that prices a self-funded structure. The Renewal Snapshot is $0, names which scope applies, and may end the engagement.
2. Broker-of-record letter. Moves the agent of record on the existing contract with the incumbent carrier. This is what lets me pull the carrier’s experience reports and the renewal workbook. The carrier does not change the commission schedule or the premium until renewal, so for November and December the commission already built into the contract, the incumbent’s 5.4%, was paid to me instead: about $17,600 for the two months, stated in dollars on the disclosure the CFO signed. Nothing about the CFO’s cost changed; the same money went to a different broker. My $35 per employee per month starts with the January 1 renewal.
3. Compensation disclosure. In dollars and per employee, in advance, under the Consolidated Appropriations Act. For this group, from the January 1 renewal: $35 per employee per month on the medical plan, $67,200 a year at 160 employees, the same number in every quote that follows. For comparison, the filing shows the incumbent was paid $89,640 on the 2024 premium, and 5.4% of the renewal as issued would have been about $118,000.

73 days from the first envelope to the new plan year

Dated steps for a January 1 renewal. The Process Standard sets the order; the dates are modeled.

Nothing is priced until November 3, when the paperwork is signed.

Why paperwork comes first: Nothing paid, and nothing using your documents or claims, starts before the engagement letter and broker-of-record letter are signed. Once a broker has quoted, the quote is the pitch and the paperwork is the close. In this order, the CFO knows what I am paid before I know what the market says, and the decision file starts on day one.

What the CFO told the CEO

I signed three things: an engagement letter, a broker-of-record letter that moves our existing policy to a new broker, and a disclosure of his pay in dollars. Nothing about our coverage has changed. The first deliverable is $0 and tells us whether the second one is worth paying for.

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. The filing history is modeled in the shape of a Form 5500 Schedule A filed with the U.S. Department of Labor; peer medians are from the 52 filings. No client is identified. Advisory illustration, not legal or tax advice.

Step 3: the documents

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. New in this step: the census, the renewal letter, the contribution schedule and the claims report.

Everything arrived within a week of the paperwork, because every item already existed. Names came off the census; no Social Security numbers were sent.

26 family contracts carry 30% of the premium

Annual premium by coverage tier at 2026 rates, 160 employees, 270 covered lives. Census and rates modeled.

Total $1.95M a year, $1,018 per employee per month. Employees pay 20% of employee-only and 50% of the dependent increment.

Two things stood out before any analysis. The family tier is 26 contracts, 16% of employees, and 30% of the premium, so the family contribution is where the largest employee decisions live. And the plan had never received a written broker compensation disclosure, which the law has required since the 2022 renewal.

What the CFO told the CEO

Everything they asked for we already had, except a claims report, which the carrier sent once the broker change was in. We had never been given the broker’s compensation in writing, which we were supposed to get every year since 2022. That is now fixed going forward.

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. The filing history is modeled in the shape of a Form 5500 Schedule A filed with the U.S. Department of Labor; peer medians are from the 52 filings. No client is identified. Advisory illustration, not legal or tax advice.

Step 4: the Renewal Snapshot, five business days, $0

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. New in this step: the renewal letter and the claims figures.

The Snapshot answers one question: is this renewal worth a Cost Review, or should the CFO sign it and move on? It was delivered November 17, five business days after the documents were complete.

The renewal asked for 11.8%. Nothing underneath it moved that fast.

Year-over-year increase. Renewal letter and claims are modeled; peer median is from the 52 filings.

Claims trend: 2026 paid claims per employee per month against 2025, net of claims above the pooling point.

What it found.

• The renewal asked for 11.8%. Carrier trend quoted for 2027 ran about 8.5%. The peer median increase across the 52 filings has been 5.3% a year. The group’s own claims, net of two large claimants, rose 3.1%.
• The 24-month loss ratio was 79% in 2025 and 67% in 2026. Two claimants, $312,000 and $168,000, account for the difference between this group and an ordinary one. Both are above the $125,000 pooling point, which means the carrier’s own underwriting formula is supposed to spread them, not charge them back in full.
• The group has about 270 lives. That is large enough for a level-funded contract from three carriers and for self-funding with stop-loss. It has never been offered either.

The verdict. Worth a Cost Review, extended scope, $12,500, because at about 270 lives the self-funded structure belongs in the comparison. Not because the renewal is outrageous (it is within what carriers ask of a group with two large claims) but because the group has never been priced by anyone else and has structures available to it that return the claims data permanently.

Sometimes the Snapshot says the opposite: the renewal is at or under trend, the claims support it, and the right call is to sign. That verdict is also $0, and the CFO walks away with a one-page record that the plan was checked.

What the CFO told the CEO

The renewal is 11.8%. Market trend is around 8.5% and our own claims, apart from two large ones the carrier is supposed to pool, grew about 3%. We have never been quoted by another carrier or shown a level-funded option, and at our size we qualify for both. I am authorizing the Cost Review at the extended scope, $12,500 flat, two to three weeks. If it comes back and the renewal is fair, we sign it knowing that.

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. The filing history is modeled in the shape of a Form 5500 Schedule A filed with the U.S. Department of Labor; peer medians are from the 52 filings. No client is identified. Advisory illustration, not legal or tax advice.

Step 5: the Health Plan Cost Review, three weeks, $12,500

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. New in this step: every quote, the claims and the carrier responses.

Delivered December 8. Three structures priced side by side, each with what the broker is paid written next to it, plus the claims analysis and the renewal rebuilt from the carrier’s own workbook.

Who was asked, and who answered

Two carriers write most of the 52 peer groups, and both were in this comparison. The point of asking a third carrier and the stop-loss market is that it keeps the first two honest.

How the quotes landed

How the 2027 quotes landed: renewal as issued, then the market

Annual medical cost for 160 employees, with broker pay in each option. Modeled from the group’s 2026 premium and 2026 market relativities for Texas groups of this size. Broker pay: incumbent at 5.4%, the rate on its last filing; mine at the disclosed $35 per employee per month.

Level-funded options shown at maximum annual cost.

Three things to read off that chart:

1. The renewal moved $140,000 the day the group went to market. Nothing about the group changed between November 3 and December 1. What changed is that the incumbent knew it was being compared, so the first number was not a market price.
2. The level-funded maximum came in under the current year. Carrier B’s maximum annual cost, the most the group can pay if claims run badly, was $1.93 million against $1.95 million current and $2.18 million renewal as issued. If claims run as they did in 2026 net of the two large claimants, the group gets a share of the surplus back and, for the first time, keeps its own claims data.
3. Broker pay is the same number in every one of my quotes. $67,200. The incumbent’s pay on the renewal as issued would have been about $118,000, and a percentage rises with every renewal. A per-employee commission is not automatically lower than a percentage; here it is, and either way the CFO can see it.

What the claims said

Two claimants, not the whole group, drove the 2025 loss ratio

Paid medical claims by month against monthly premium. Carrier experience report, 24 months, modeled. Pooling point $125,000.

Net of the two claimants above the $125K pooling point, claims ran 66% of premium over 24 months.

Net of the two claimants above the pooling point, the group ran a 66% loss ratio over 24 months. A group running that loss ratio is a candidate to keep its own good years rather than leave them with a fully insured carrier. It is also the single fact that the fully insured renewal letter did not mention.

What the CFO told the CEO

Four carriers were asked, three quoted. Our current carrier dropped its renewal from 11.8% to 4.6% once it knew we were shopping. Another carrier offered a level-funded plan where the most we can pay next year is less than we pay now, we get money back if claims are good, and we own our claims data. Every quote shows the broker’s pay next to it, and it is the same $67,200 on each. I have a recommendation and the file behind it.

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. The filing history is modeled in the shape of a Form 5500 Schedule A filed with the U.S. Department of Labor; peer medians are from the 52 filings. No client is identified. Advisory illustration, not legal or tax advice.

Step 6: the decision, and the file behind it

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. New in this step: the decision and the figures in it.

Signed December 11 by the CFO as plan administrator. The file is six pages and this is its first.

Decision

Move the medical plan to Carrier B on a level-funded contract effective January 1, 2027. Keep the plan design (deductible, out-of-pocket, network tier) within 5% of current so employees see a carrier change, not a benefit cut. Keep dental and vision with the incumbent for one more year; they renewed at 4% and flat and were not the problem.

Why this and not the others

What it cost employees

Twenty-three employees had a provider outside the new network; three had ongoing treatment and were moved under transition-of-care. Contribution percentages were unchanged; because the composite rate went down, the family contribution fell about $9 a month.

What is in the file

The Snapshot, the Cost Review, all five quotes as received, the carrier’s revised renewal letter, the compensation disclosure, the census summary without names, the claims report, and this page. The next person to ask “how was this decided” gets the file, not a recollection.

The modeled maximum cost for 2027 is $254,000 below the renewal as issued. That figure is in the file because it is what the decision was measured against, not because it was promised in advance. Nobody quoted a savings number to this CFO before the paperwork was signed.

What the CFO told the CEO

We are moving medical to a new carrier on a level-funded plan January 1. Worst case we pay $1.93 million, which is less than this year and $254,000 less than the renewal we were handed. If claims run as they have, we get a share back. Twenty-three people had to check their doctors and three had care transferred. Dental and vision stay put. Everything that went into this is in one file with my signature on it, and the broker’s fee is on page two.

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. The filing history is modeled in the shape of a Form 5500 Schedule A filed with the U.S. Department of Labor; peer medians are from the 52 filings. No client is identified. Advisory illustration, not legal or tax advice.

Step 7: 120 days before the next renewal

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. New in this step: the 2027 claims and the 2028 renewal.

September 3, 2027, 120 days before January 1, 2028. The process starts again at step 3, and this time the documents include something the group has never had: twelve months of its own claims, on a contract it controls.

What the year showed

• Paid claims ran 68% of the maximum funding level. One claimant exceeded the specific stop-loss; the carrier absorbed it, which is what the stop-loss is for.
• Surplus at year end: $84,000, of which the contract returns 50%. $42,000 came back to the plan in March 2028.
• Carrier B’s 2028 renewal on the level-funded contract: +3.9%, against a quoted trend of about 7.5%.

The Renewal Snapshot, second year

Delivered September 15, $0. Verdict: no Cost Review needed. The renewal is under trend, the claims support it, the broker fee is unchanged, and the group now has the data to go to market any year it wants to. One page went in the decision file: renewal reviewed, market checked by reference to last year’s quotes, signed.

That is the ordinary outcome, and it should be. The point of year one was not the $254,000. It was that every year after it starts from a file instead of a letter.

The standing calendar

What the CFO told the CEO

Year one of the level-funded plan: claims ran 68% of the maximum, we got $42,000 back, and the renewal is 3.9%. The snapshot says sign it, so we are signing it. Same broker, same fee, one page added to the file. We will go to market again when the numbers say to, and we will know when that is because we have the numbers.

A composite built from the 52-company peer set. No real company’s filings or claims are shown; every figure is modeled so the arithmetic holds together. The filing history is modeled in the shape of a Form 5500 Schedule A filed with the U.S. Department of Labor; peer medians are from the 52 filings. No client is identified. Advisory illustration, not legal or tax advice.