Thinking back on those who had provided insights and inspiration during 2025, we caught up again with Bil Westerfield, President & Chief Data Scientist at Magpie Health Analytics, to discuss how mounting operational pressures, rapid policy change, and increasing scrutiny are reshaping how agencies and industry partners think about analytics, efficiency, and trust.
You can check out Bil’s other recent interviews here: Magpie Health Analytics: Telling Stories, Connecting Data, & Building Trust to Transform Healthcare and Podcast Flashback: CMS Secrets to Success with Raghu Akkapeddi, Bil Westerfield, and Robert Hicks.
Rethinking the Way Forward: The External Reality
Across government, 2025 has marked a real inflection point—not because the challenges are new, but because long-standing ideas about efficiency, modernization, and operational change are finally moving from discussion to decision.
“Many of the policy ideas and operational changes we are seeing have been percolating for a decade or more, but we’re starting to see decision-makers in agencies move on these ideas and really think through the downstream impacts, not only within government, but across contractors and into the industries and beneficiaries they serve.”
Agencies across the board are being pushed to rethink how they do business and how they operate more efficiently, and that pressure is showing up in fundamental ways. “They’re trying to make policy decisions. They’re trying to reimagine policies. They’re trying to reimagine whole payment systems and entire contracting ecosystems that they’ve been operating under for 20-plus years.”
While the higher-level decisions are being made by agency leadership, much of the real burden shifts to group and division leaders who are responsible for translating policy into action. These leaders are increasingly looking for operational analytics that help them understand cost savings, tradeoffs, and the potential unintended consequences of different paths forward.
“As we’re handing information to our business leaders, who then must have discussions with their leadership and peers who may or may not agree on the direction forward, they need to be able to present the whole picture, speak to the ‘what ifs,’ and address the ‘may happen’ scenarios.”
The need to understand how decisions will play out operationally has become one of the defining challenges agencies and their industry partners now face.
Speed vs. Risk: Trust Under Scrutiny
As agencies are pushed to move faster, the tension between speed and risk has become unavoidable. New tools and technologies, particularly AI, promise to accelerate analysis and decision-making, but they also raise a more fundamental question: can leaders trust the answers well enough to act on them?
“Leadership has to be able to confirm the information is accurate, but also from a business perspective confirm that the savings or efficiencies are real and be able to defend that the decision was the right one.”
That expectation changes the role of analytics. Model outputs need to be produced fast and results must hold up under scrutiny whether that scrutiny comes from auditors, oversight bodies, or internal stakeholders who may disagree.
For industry partners, this raises the bar. “It means being able to clearly articulate data sources, methods behind the models, and the messaging leaders can use to interpret and defend the outputs.” In practice, building that confidence often requires slowing down just enough to ensure understanding before speed is applied.
“Sometimes that means sitting with your business sponsor to align on what the data shows and how it was produced, ensuring the findings and conclusions can stand up across subsequent discussions.”
Without that shared understanding, speed becomes a liability. With it, analytics can help leaders move forward with confidence rather than hesitation.
Operational Analytics: Anticipating Outcomes Before Acting
As agencies confront the need to move faster while remaining accountable, operational analytics has taken on a more forward-looking role. “Rather than simply measuring performance after decisions are made, leaders are increasingly asking a different question: Will this change work the way we think it will?”
Operational analytics starts with measuring real work in real systems processes, procedures, and day-to-day activities. How many claims are processed? How many beneficiaries are enrolled in the program? What costs are incurred, and how do workloads shift as changes are introduced?
From there, the focus turns to scenario evaluation. Leaders want to understand whether proposed changes are likely to produce the intended outcomes before they commit to them. Will a policy change improve throughput? Will an operational adjustment reduce costs without creating new bottlenecks? Will efficiencies realized in one area simply shift burden elsewhere?
This kind of analysis helps decision-makers evaluate tradeoffs in advance rather than discovering them after implementation.
Only after expected outcomes are examined does the lens widen to unintended consequences. Changes that look sound on paper can behave very differently once they encounter contracts, systems, and workflows that have evolved over decades.
“What we’re increasingly asked to help agencies do is evaluate how changes intended to improve efficiency or oversight play out operationally and that’s where analytics make a real difference.”
When operational analytics are done well, it gives leaders the evidence they need to move forward with confidence. It allows them to explain not just what decision they made, but why they believed it would work, what outcomes they expected, and how they considered the tradeoffs involved.
Without that grounding, even well-intended changes risk creating friction, confusion, or exposure once they hit execution.
Direction Forward: AI, Operational Analytics, and Defensibility Under Scrutiny
Looking ahead, Bil Westerfield predicts artificial intelligence will move beyond language—beyond writing and coding—and increasingly shape how data analysis is performed and how decisions are made across government. Decision cycles will continue to compress, but expectations around oversight, auditability, and accountability will not.
In this environment, the value of analytics will not come from speed alone. As AI becomes more embedded in decision-making, leaders will need confidence not just in the answers produced, but in their ability to explain, defend, and stand behind those answers when questioned.
“Speed only helps if you can stand behind the decision that follows. If you can’t explain or defend the answer, faster just means risk shows up sooner.”
As AI moves beyond conversational use cases and into decision support, the challenge shifts from generating responses to encoding real domain understanding. At this stage of AI, subject-matter expertise in data, analytics, and policy isn’t displaced it is essential.
“You need experience that understands how the data works, how it flows from its source, the transformations that apply to it, and the policies that govern it. That’s how you teach AI and language models a finite domain, so the answers the model producesare reliable and defensible.”
Going forward, organizations will be judged not only on how quickly they can act, but on how well they can justify their decisions. The future belongs to analytics that help leaders anticipate outcomes, understand trade-offs, and make decisions that hold up under scrutiny.
