
Four people work from every employer health plan. What each seat should see, what each one owns, and why review belongs at the seams between them.
Four people work on every employer health plan, and each of them needs something different from it. The advisor needs breadth and room to apply judgment. The service team needs the history behind every open item. The employer needs decisions laid out clearly enough to make them. The employee needs a plain answer about their own coverage. When all four draw from the same plan, and each seat is designed around what that person is actually responsible for, the plan gets easier to run for everybody in it.
A few weeks ago I wrote about the living record of the plan, the idea that one plan should have one continuously maintained record underneath it. This piece is about the next question. Once that record exists, what should each person see?
I'd ask the same three questions of any seat, in any tool, ours included. What does this person see? What are they responsible for? And where does a person review what reaches them?
The answers are different for each role. That's the point. A single screen for everyone either buries the employer in detail or starves the advisor of it. Four disconnected tools give you four stories about one plan. The design has to sit in between.
The advisor should see the widest view of anyone. Across the whole book of business, and deep into any one plan when it matters. Signals, plan terms, contracts, vendor performance, prior recommendations, and what has already been tried.
The advisor is responsible for interpretation. That means deciding what a signal means for this particular client, given everything they know about the client's priorities and leadership, and deciding what's worth raising at all.
Review sits with the advisor first. Nothing becomes a recommendation to an employer until the advisor has looked at it and added what they know. The design principle I'd hold here is simple. Give the advisor full context, and make it fast to get from a signal to the plan detail behind it. The work you do as an advisor is harder than most of your clients see, and the tool should make it lighter. What an advisor shouldn't have to do from this seat is assemble. If the first hour of every client review goes to pulling files together, the seat was built for storage instead of judgment.
This is the seat that usually gets designed last, and it's the one that keeps the plan running week to week.
The service team should see the open work. Deadlines, owners, what was promised to the client, what has been done, and why each item exists in the first place.
They're responsible for follow-through. An approved decision only becomes a result when someone carries it the rest of the way, and that someone is usually on the service team.
The principle here is that every task should link back to the decision it serves. When a service professional can see the reason behind the work, they can tell a routine item from one that points to something bigger, and flag it to the advisor before it turns into a problem. That escalation is their review point, and it's one of the most valuable signals a plan has.
The employer should see reviewed priorities, why each one matters, a summary of the evidence, the decision that's needed, the person who owns it, and the outcome once the work is done.
The employer is responsible for the decisions themselves, and for owning them as the plan sponsor. That's a serious responsibility, and the seat should respect it by being built around the decisions the employer actually has to make.
Review here works in two directions. Everything that reaches the employer has already been reviewed by the advisor. And the employer can always ask for the detail underneath any priority, because a sponsor should never have to take a summary on faith. They should be able to see the reasoning whenever they want it. What the employer shouldn't have to do is translate. If a priority arrives in language only an analyst can read, the review step didn't finish its job.
The employee should see their own coverage, their own costs, their options, and a clear next step, in plain language.
They should never see internal analysis, a recommendation that's still under review, or anything about anyone else. The employee is responsible for their own choices, and a well-designed seat respects that the decision is theirs. It explains. It doesn't steer.
Review for this seat runs the other way. Nothing about an individual employee flows upward. What the plan learns from employee questions reaches the advisor and employer only as an aggregate pattern, and only after a person has reviewed it. That boundary protects the employee, and it protects the employer too. No benefits leader should ever be in a position to see one person's private question, and a well-built seat makes that impossible by design instead of by policy.
If you step back, review doesn't live inside any one seat. It lives at the seams between them. A signal moving to the advisor. An advisor's recommendation moving to the employer. A pattern in employee questions moving to anyone at all.
The rule I'd hold every platform to is that nothing crosses from one seat to another without a person deciding it should. Technology can do a great deal of first-pass work inside a seat, and do it quickly. Crossing into someone else's seat is a human decision. That's the difference between a governed experience and a shared login.
For employers, it means expecting to see decisions you can act on, and asking where the review happened before something reached you.
For advisors, it means a tool that makes your judgment more visible to the client, not less.
For service teams, it means the reason behind the work finally travels with the work.
For employees, it means an answer about their plan, without any of the machinery behind it.
None of this is exotic. It's how the best advisory relationships already work, on their best days. The design question is how to make that the normal day. If you sit in one of these four seats and see it differently, I'd value hearing where.
