Hidden Cost of Hype: Revenue Cycle Realities at HFMA Annual Conference

Every year, the HFMA Annual Conference surfaces the same tension in new packaging: healthcare organizations are being asked to achieve more with shrinking margins, thinning teams, and rising complexity. This year, that pressure had a name everyone recognized: "Do more with less".

After several days of sessions, vendor conversations, and one conversation in particular worth highlighting, here's what actually stood out.

Full disclosure: I was at the conference to listen and participate in a video interview.  Macedon has a No Surprises Act Automation software solution. 

The Payer-Provider Problem is a Data Problem

One of the conference's most candid sessions “The Conversation We’re Not Supposed to Have“, featuring David Weathington (Chief Administrative Officer at Blue Agilis) and Todd Tilghman (Director of Revenue Cycle at Williamson) framed the payer-provider relationship with unusual honesty. The mistrust is structural, the friction is real, and neither side is entirely free of blame.

But the most actionable takeaway wasn't about improving relationships in the abstract. It was about information quality. Providers who show up to dispute resolutions and contract conversations with vague complaint lists or unreviewed file dumps get nowhere. Those who come with clean data, honest percentages, and specific problem definitions get results (supposedly) in the form of authorization denials reversed, administrative errors forgiven, payment process changes negotiated.

The implication for RCM leaders is direct: your reporting infrastructure is your negotiating infrastructure. If your data is messy, your leverage is limited. If you can surface the right information with precision, you can move payers in ways that blanket grievances never will.

The AI Cost Risk

The vendor floor this year was, generously, enthusiastic about AI. Less generously, most of it looked the same. When pressed on questions about cost sustainability and what happens when models underperform at scale, most vendors didn't have satisfying answers. The most honest response came from a vendor using machine learning alongside large language models rather than leaning on LLMs alone, which is a more sustainable and interpretable approach that's worth paying attention to.

The reason that question matters more than most vendors wanted to acknowledge: the data volumes in RCM are enormous. A single patient encounter averages 59 clinical documents (for simple cases) and roughly 1000 times as many words, about the same as a novel.  That volume runs through revenue cycle processes repeatedly, across thousands of encounters. Leveraging large language models against that volume, on every case, every time, isn't just technically complex. It's expensive in ways that compound fast and are difficult to forecast.

The sessions that held up best pointed toward a more disciplined approach. “Myths vs. Realities of AI Governance“ led by Dr. Hassan Tetteh and Amol Joshi, whose backgrounds span decades of AI deployment in high-stakes environments, made the case that governance isn't the enemy of innovation, it's what makes innovation survivable. The framing that stuck: governance is brakes, not guardrails. Brakes let you go fast because you can stop when you need to.

That framing applies directly to cost risk. The opportunity isn't deploying AI everywhere. It's deploying the right AI where it actually creates value, with governance strong enough to know the difference. A platform approach makes that possible. Rather than locking into point solutions that apply the same model to every problem regardless of complexity or cost, a well-architected platform lets you route work intelligently and reserve your most capable (and most expensive) AI for the cases that warrant it, and handle simpler volume through lighter, faster, cheaper methods. That's how you get best-in-class results without best-in-class costs eating your margin.

The Conversation that cut Closest to Home

At the conference, I had the opportunity to sit down with Todd Nelson, Director of Partner Relationships and Chief Partnership Executive at HFMA, for a recorded interview on opportunities in the RCM space. The conversation covered tangible changes health systems can make now. Not theoretical transformation, but specific levers with real financial impact.

The thread running through it: too many organizations have distributed mission-critical revenue cycle functions across a sprawling vendor landscape, and they're paying for it in ways that don't always show up on a single line item. Visibility suffers. Cost control becomes reactive. And when something goes wrong, there's no clean place to look. Bringing those core processes in-house, supported by the right technology and disciplined workflows, changes that picture entirely. You get the outcomes, you control the costs, and you have the data visibility to actually manage the operation rather than just respond to it.

That data visibility also feeds directly back to the payer relationship dynamic discussed earlier. Clean, well-documented data is the difference between a productive dispute and a dead end.

One specific area we discussed: the No Surprises Act Independent Dispute Resolution process. The value of IDR as a recovery mechanism is well understood, but the challenge is operational. Most health systems are either outsourcing it at significant expense, not participating at the scale their volume warrants, or running it on strained teams without the process infrastructure to support. Organizations that bring IDR in-house with the right workflows and technology recover more, spend less, and build the operational muscle to scale and reclaim margin leakage.

What "Do more with less" Actually Looks Like

Winning in the current healthcare environment doesn't require a sprawling landscape of disconnected point solutions that silo your data and fragment your visibility. It requires orchestration. In practice, it comes down to three things: 

1. AI deployed selectively, applied where it creates clear value rather than blanketed across every process, which keeps costs manageable and outcomes predictable.

2. A platform that ties your critical workflows and tools together rather than adding another silo to manage. Orchestration is what turns individual capabilities into operational leverage.

3. Almost as a natural result of the first two: real reportability. When your processes run through a common infrastructure, you finally have the visibility to manage your operation, make the case to payers, and find the next opportunity to improve.

The takeaway from HFMA this year was clear: Stop buying tools. Start fixing the architecture. 

Want to continue the discussion? Let's Talk