By Greg Sanderson, VP of Analytics
Over the last several years, pharmacy analytics has become increasingly visual.
Dashboards, charts, and side-by-side comparisons now play a central role in how clients review PBM proposals and explain options to clients. And to be clear, that shift isn’t a bad thing. Clear visuals can make complex information easier to digest and help guide conversations.
But as the industry leans more heavily on front-end views, it’s worth asking a more fundamental question: What determines whether the numbers behind those visuals are actually right?
Because in pharmacy, accuracy doesn’t come from the dashboard. It comes from the modeling behind it.
PBM modeling is not a surface-level exercise
Evaluating PBM proposals is fundamentally different from comparing many other benefit components. Pharmacy pricing is layered, conditional, and constantly changing. Small differences in definitions, exclusions, specialty classifications, or rebate treatment can materially change outcomes.
Accurate PBM modeling requires depth — both in data and in logic.
At a minimum, it means working at the claim level, not just with summary assumptions. It means understanding how drugs are classified, which claims are eligible for rebates, how exclusions apply, and how program participation realistically impacts cost. And it means updating that logic continuously as PBM terms evolve.
When modeling doesn’t account for these details, results may look clean on paper but fail to reflect how a contract will perform in practice.
True PBM modeling pulls from a wide range of inputs
Claims data is only the starting point. Accurate analysis also incorporates drug attributes, specialty and limited distribution lists, contract language, pricing terms, rebate structures, program rules, and evolving market conditions.
But data breadth alone isn’t enough. Each assumption applied to that data needs to be validated.
- Which definitions are being used?
- Which exclusions apply?
- How are rebates calculated — and under what conditions do they change?
- Which savings programs are members actually eligible for, and what happens when participation varies?
These questions don’t always have simple answers, and they rarely remain static year over year. That’s why PBM modeling isn’t a “set it and forget it” exercise. It’s an ongoing process of refinement.
Why assumptions matter more than most people realize
In pharmacy, assumptions drive outcomes.
Savings projections often depend on expectations around program enrollment, rebate eligibility, or utilization patterns. If those assumptions are overly optimistic, or simply outdated, the model may tell a compelling story that doesn’t materialize at renewal.
That’s why assumption validation is one of the most critical (and most overlooked) parts of PBM modeling. It requires reviewing contract language closely, understanding how PBMs operationalize pricing, and applying realistic expectations based on experience — not just theoretical maximums.
This is also where human expertise matters most. Even the most sophisticated analytics framework needs informed oversight to interpret nuance, identify inconsistencies, and explain tradeoffs.
Dashboards have a role — but they aren’t the whole story
Dashboards are valuable. They help visualize outcomes, compare scenarios, and communicate results more clearly to employers.
But dashboards are the end of the process — not the process itself.
They don’t show how assumptions were built, how definitions were applied, or how changes in the market were incorporated. They don’t explain why two scenarios differ, or what risks may exist beneath the surface.
For clients advising on PBM decisions, that context is essential. Clients don’t just need to see the numbers, they need to trust them.
What clients should expect from PBM modeling today
As pharmacy pricing grows more complex and market conditions continue to shift, clients should expect more from the analytics supporting PBM decisions.
At a minimum, that includes:
- Claim-level modeling grounded in real data
- Continuously updated logic as pricing terms and definitions evolve
- Clear validation of assumptions that drive projections
- Human expertise to interpret results and explain implications
- Transparency into what’s driving cost — not just what the total looks like
PBM decisions carry long-term financial and operational consequences. They deserve an analytics process that reflects that complexity.
Dashboards help tell the story.
But accuracy — the kind that stands up at renewal and under scrutiny, is built behind the scenes.
And that’s where the real work of PBM modeling happens.
