> ## Content Index > Fetch the complete content index at: https://www.thefringetheory.com/llms.txt > Use this file to discover other available public pages before exploring further. # Mending the trust gap - URL: https://www.thefringetheory.com/benefits-broker-rfp-what-to-do-first/ - Published: 2026-09-11T11:24:43.000Z - Updated: 2026-09-11T11:24:43.000Z - Description: The advisor brings the ideas, the employer decides the fit. But who owns the outcome? - Author: Derek Winn - Tags: Theory Most plan sponsors want the same thing, which is better benefits for less money. The market has no shortage of answers to that, and few of them have ever been accountable to a number. A good broker or consultant can put a limitless array of solutions in front of you, and most of them have worked somewhere for somebody. The health plan is frequently the largest line item in the business that does not get run with an owner and a number. The challenge is two fold. Every one of those solutions needs a willing sponsor, and nearly every one of them also needs willing participants. The second half is where it generally stops. Changing your broker or consultant is a change to one party. Asking every employee and every family member on the plan to change how they buy healthcare is a change to countless parties. Putting a plan in action because it made sense at the time, approving it by committee, and never going back to remeasure it is an accountability gap. I liken this conversation to when I was a high school teacher. Typically I would have class for 45 to 90 minutes. In that time, I worked hard every day to educate students, and some would still fail exams. What separated the students who passed from those who failed was the work they did when class was not in session. The exam was never a surprise. They knew the standard going in, and so did I. Real change sounds like work, not only for plan sponsors but for employees, and that generally means more work for administrators too. It also carries a fear of trouble with attraction and retention, HR turning into a complaint department, and more. That fear is not irrational, which is exactly why it wins so often. What gets left out of that calculation is whether the plan was keeping score the whole time. Neither change, the broker swap or asking members to buy differently, generally had a goal written down, and neither had a result measured against one. That leaves political capital as the only quantity in the room anyone can measure. In a survey of roughly 1,900 employers published by the brokerage industry's trade association, the share saying they did not plan to switch brokers for the foreseeable future fell from 31% in 2024 to 18% in 2025 \[1\]. Among the top reasons employers gave for switching, at 27%, was lack of proactive support with managing health and benefit program costs. Most commentary reads that as a trust problem. I think trust is the symptom. The thing underneath it is that very little in this market has been accountable to a number, and that includes the part employers own. When I started Fringe Theory the goal was to write the version of this I would want to read if I were the plan sponsor, along with [the tools to do something with it](https://www.thefringetheory.com/bam/). Acknowledging a trust gap belongs on everyone's list. Measuring your own internal accountability gap is the harder half, and there is no product that does it for you. Three questions worth holding as you read. The piece is about the last one. - When something you bought did not work, how did you find out, and how much time did it consume? - How many advisor or vendor relationships have you started or ended in the last five years? - Did anything measurably improve along the way, and did you change your own approach? ## Twenty years, and what it bought In 2006 the average annual family premium for employer coverage was $11,480\. In 2025 it was $26,993, an increase of 135% \[2\]. Across that span employers bought most of what the market recommended, and the two biggest sellers were high deductible plans and wellness programs. The evidence on both is now in. High deductible enrollment with a savings option ran about 4% of covered workers in 2006 and 33% in 2025 \[2\]. I have written about [that design before](https://www.thefringetheory.com/what-a-deductible-increase-saves/), so the short version. The cleanest natural experiment moved a whole workforce onto a high deductible plan in one year \[3\]. Preventive services fell 7.5% in the first year and another 5.2% in the second, though preventive care stayed free and sat outside the deductible. Wellness has been through two large randomized trials, and both came back null, one of them checked again at three years \[4\]\[5\]. Set side by side, the two biggest sellers of the last twenty years look like this. | | High deductible plans | Wellness programs | | --------------------------------------------- | ---------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------- | | Sold as | Members with skin in the game shopping on price | Healthier members, and lower spend to follow | | What the randomized evidence found | Spending fell 11.8% to 13.8%, all of it from less care. No price shopping in either year \[3\] | Null on spending, behavior and health at 4,834 employees \[4\]. Null on all ten clinical markers at 32,974 \[5\] | | What the typical buyer could check it against | Nothing written down | Nothing written down | Sources: Brot-Goldberg and others, Quarterly Journal of Economics, 2017, for the high deductible findings. Jones, Molitor and Reif, Quarterly Journal of Economics, 2019, and Song and Baicker, JAMA, 2019, for wellness. Full citations at the foot of the piece. RAND is not a trial but it points the same way. Looking at ten years of one large employer's data, it found the 87% of participation going to lifestyle management produced 13% of the savings \[6\]. The sharpest finding there is a selection one. In the 13 months before the Illinois program started, the employees who would go on to participate were already spending $115.30 a month less than the ones who would not \[4\]. A program can look like it saves money purely by attracting people who were already cheaper. What the two have in common is a missing measure. Call it a BLACK BOX. Not a solution that conceals its results, a solution whose results the buyer was rarely in a position to check. Both were sold with a return figure, and in the HDHP's case that figure generally showed up in the rate before anyone bought it. Neither came with a measure the buyer could hold it to. The employer who bought a wellness program in 2014 was sold something the market believed in at the time. They were handed a number, or an ROI, and no way to tell later whether it had been measured. ## Two places accountability went missing Accountability went missing in two places, and only one of them gets talked about. **The vendor was rarely held to a measure.** Not because anyone refused, but because asking at purchase was never anyone's job. Ask most vendors today what they moved and a good number will tell you they were never given a baseline to move it against. **And the employer did not measure either. This is the part that gets talked about least and the part you can actually do something about. The peer reviewed survey work on plan administrators finds that plan administrators measure, on average, very little of what they buy \[7\]. A baseline only exists if someone wrote it down.** That is not for lack of interest. Plenty of organizations do parts of this well, and there is usually one person behind that who decided it mattered. Even there, the measure tends to get assigned after the purchase rather than before it. Benefits is one of a long list of things on an HR director's desk, and most of the claims data sits with the carrier rather than the plan. The renewal calendar also rewards getting to a decision rather than going back to check the last one. The measure was rarely assigned to anyone, and a measure with no owner does not get produced. Overall, benefits is one of the few line items where the work and the understanding both get outsourced. Nobody decided it should be that way, it is just how the market grew up. There is one number almost everybody does check, which is trend against the benchmark. Did we come in under the average or over it? That is a real measure and it is worth having, but it describes the market rather than the plan, and it cannot tell you which of the things you bought moved anything. An employer can beat trend in a year none of its program spend moved anything, and miss it in a year the programs worked. **What makes that second gap odd is that these same companies already know how to do this everywhere else.** Most businesses of any size run on some methodology. EOS out of Traction, Scaling Up, OKRs, whichever one leadership landed on. They all do roughly the same four things: name the priority, put one person on it, attach a number, and check it on a cadence. Very seldom does any of that reach the benefits committee, which generally meets a few times a year, approves a renewal, and carries no scorecard, no owner per line, and no starting number. The plan is often the largest line item in the business that does not get run that way. Overall, we all want trusting and accountable partners. That is a reasonable thing to want and most people in this business are trying to deliver it. The division of labor here is not complicated. The advisor brings the ideas and the evidence behind them, the employer decides which ones fit the plan and says so plainly, and both sides own the measure. Where it generally breaks down is that last part, because the two sides rarely agreed at the start what would be measured or who would produce it. ## The reflex, and the introspection it skips Often the answer to a plan that is not performing starts with an RFP. That is the long standing move and generally the first thing anyone suggests. Let's break this down. On the merits an RFP is frequently the biggest time sink in a renewal cycle, and in a great many markets it is low impact besides. If two hospital systems own your market and both sit in every network you would seriously consider, [a discount comparison has very little to find](https://www.thefringetheory.com/what-a-network-discount-buys/). More administrative drag, roughly the same result. An RFP changes the vendor. It rarely changes the process around it, seldom produces a baseline, and does not answer what the last three programs actually moved. An employer who has been through several relationships in the past several years is looking at a pattern. At some point it is worth asking what all three of them were missing, and whether it was something only you were in a position to hand them. That is not a comfortable sentence to write as a benefits advisor and I mean it plainly rather than as a criticism. A good advisor hired into the same gap generally lands in the same place. So the move I would make before the market check is introspection. What did we ask for, what did we write down, and would we have known if it worked. Then go to your advisor with questions rather than relying solely on their advice. In my experience most advisors would welcome that conversation. ## The arithmetic One count, runnable this quarter, and every input is already yours. Take the last five years. List every program, vendor, point solution and advisor relationship you started or ended, with the annual cost beside each. For a mid-sized self funded employer that list generally runs ten to fifteen lines, and it usually totals more than anyone in the room expects, because it has rarely been assembled on one page. Now put three columns against it. What was this supposed to move. Where did that measure sit the day we bought it. Who has produced a figure since, and where the program was sold on savings, was that figure against paid claims or billed charges. One thing worth saying plainly before you try it. **This is committee work rather than analyst work**, and it wants HR and finance in the same room. HR knows what was bought and why. Finance knows what it cost and what that money was up against. Neither half produces the list alone, and if you do not have a benefits committee at all, that is a common situation and probably a different piece. The number that matters is how many lines you can complete. A baseline will not prove a program caused anything and I would not claim otherwise. What it does is make someone accountable for it, which is the difference between a program you can retire on evidence and one that renews forever because nobody can say either way. ## What to do Two lists, because the stop side is what frees the time for the start side. **Stop doing** - Running an RFP as the first response to poor performance. Run it when you have a reason that survives being written down. - Buying anything whose result you cannot rate. If nobody can say what measure it should move and where that measure sits today, you are buying a black box. - Accepting a savings figure without asking which quantity it is. Against billed charges and against paid claims are not the same number, and the difference is frequently the entire savings figure. **Start doing** - Start keeping score before you buy. Write down what the program is supposed to move and where that measure sits today. It costs one line and it is the only thing that makes the next conversation possible. - Ask every current vendor what they moved and how they measured it. Most will want to answer, and the ones who cannot were generally not given a baseline to answer against. - Put one name and one date against every program you keep. A measure with no owner and no review date is a measure that rarely gets produced. None of that requires a consultant in the room, and it works perfectly well with one. The difference is between taking advice and being able to talk about it on even terms. ## The Tail There is a version of this where the measurement arrives from somewhere else. A vendor decides that proving it is worth more than claiming it, or a carrier hands the plan a clean read on what each program actually moved. Some of that is starting to happen, and it is worth encouraging. It still will not produce the list. The list runs across vendors, across years, and across two departments, and no one party in the arrangement can see all of it. The only desk with a view of the whole thing is the plan sponsor's, which is generally also the desk with the least time. That is the part I do not have a good answer for. The worksheets behind all of this are on the site. Free, no gate, no email, and each one carries an AI prompt you can run your own claims file through. [The Benefits Accountability Matrix, the four boxes and the grid](https://www.thefringetheory.com/benefits-accountability-matrix/) [BAM: The Iceberg Review, where the money is actually concentrated](https://www.thefringetheory.com/iceberg-review/) [BAM: Where the leverage is, twelve cost levers ranked](https://www.thefringetheory.com/where-the-leverage-is/) See also [From snout to tail](https://www.thefringetheory.com/what-drives-employer-health-plan-cost/), which is where this publication started and which argues that cost shares decide who finances the spend rather than its size. ## Sources 1. Leader's Edge (The Council of Insurance Agents & Brokers), "The State of Benefits Satisfaction," August 29, 2025, reporting the McKinsey Employer Health Benefits Survey 2025 of approximately 1,900 employers. 2. Kaiser Family Foundation / HRET, *Employer Health Benefits Annual Survey*, 2006 and 2025 editions. The 135% premium increase and the enrollment comparison are my arithmetic from the two survey years. 3. Brot-Goldberg ZC, Chandra A, Handel BR, Kolstad JT, "What Does a Deductible Do? The Impact of Cost-Sharing on Health Care Prices, Quantities, and Spending Dynamics," *Quarterly Journal of Economics*, 2017;132(3):1261-1318. 4. Jones D, Molitor D, Reif J, "What Do Workplace Wellness Programs Do? Evidence from the Illinois Workplace Wellness Study," *Quarterly Journal of Economics*, 2019;134(4):1747-1791. 5. Song Z, Baicker K, "Effect of a Workplace Wellness Program on Employee Health and Economic Outcomes: A Randomized Clinical Trial," *JAMA*, 2019;321(15):1491-1501, reporting the eighteen-month results; and "Health And Economic Outcomes Up To Three Years After A Workplace Wellness Program," *Health Affairs*, 2021;40(6):951-960. 6. RAND Corporation, *Do Workplace Wellness Programs Save Employers Money?*, RB-9744, 2014, reporting RAND's analysis of ten years of one Fortune 100 employer's data. The underlying peer reviewed paper covers seven of those years: Caloyeras JP, Liu H, Exum E, Broderick M, Mattke S, "Managing Manifest Diseases, But Not Health Risks, Saved PepsiCo Money Over Seven Years," *Health Affairs*, 2014;33(1):124-131. 7. Richman BD, Monahan AB, Pfeffer J, Singer S, "ERISA and the Failure of Employers to Perform Their Fiduciary Duties: Evidence from a Survey of Health Plan Administrators," *Journal of Law, Medicine & Ethics*, 2026;54(1):14-20. Fringe Theory is independent and unaffiliated. Views expressed are my own and do not represent those of my employer. Nothing here is legal, tax, medical, or investment advice. Figures described as modeled or derived are my own arithmetic from the public sources listed, not disclosed data.