“This is not real-time data.”
Jennifer Sanchez says it almost apologetically, and then immediately explains why that lag is exactly what makes MGMA’s Provider Compensation benchmarks trustworthy. In an era of dashboards that promise instant answers, the most widely used dataset in medical practice management is intentionally a year behind. That tension — between speed and rigor — runs through nearly every misunderstanding MGMA sees around provider compensation data.
Sanchez and Liz Gurley, data strategists at MGMA, joined Daniel Williams, senior editor and host, on the MGMA Insights Podcast Network to open a limited series on how MGMA data is collected, validated, and applied. Their first stop was the MGMA dataset practices rely on most — and often misinterpret most.
The data most practices use, and why it feels deceptively simple
MGMA DataDive begins with data collection and ends as the framework leaders use to evaluate pay, performance, and practice economics. “DataDive in general is benchmarking data,” Sanchez said. “We collect nationwide data, from small private practices all the way up to very large organizations.”That's because provider compensation is rarely meaningful in isolation.
The dataset includes annual compensation, work RVUs, collections, encounters, and call pay for physicians and advanced practice providers. “It really helps provide compensation, data, productivity, financials,” Sanchez said. “And it allows organizations to ask those questions: 'Are we paying our providers appropriately? Are we paying them at a competitive market rate? What is their productivity against their peers?'”
Gurley emphasized that the label “provider compensation” often obscures what leaders actually use. “We also have productivity: work RVUs, collections, revenues, even total encounters,” she said. “If you’re just trying to determine how many patients your provider should be seeing or where your provider falls within the market, that’s reported annually.”
The most common mistake happens here: treating a single benchmark as an answer rather than a prompt. Sanchez put it bluntly later in the conversation: “Not just focusing on one benchmark — you have to use them together in combination.”
Why the data is always a year behind and why it's non-negotiable
The annual lag is the cost of accuracy. “When we say 2026 reporting, it’s based on 2025 data. We are benchmarking last year because we collect data over a full 12-month period," explained Gurley. Partial-year data would distort productivity, compensation, and seasonality.
Gurley detailed what happens after submissions close. “We have a data solutions team made up of analysts who are verifying, validating, and aggregating all of that data,” she said. “We’re analyzing trends, looking for outliers, and making sure that if there are outliers, they’re explained.”
That validation window is why MGMA can say, with confidence, that published benchmarks are stable enough to support contracts, recruitment, and governance decisions. “When we do start to publish, we’re very confident in what we’re putting out to the market,” Gurley said.
Percentiles: What the 50th actually means and what it doesn’t
Few concepts generate more anxiety than percentiles. Sanchez sees it constantly. “When we’re looking at the 50th percentile, that is the median of what our participants are giving us,” she said. “Half are under, and half are over. There is no set guideline saying this is where you have to be.”
Percentiles are meant to show you where you stand, not tell you what to do. “If we’re under, it brings the question why? 'If we’re over, why are we over? What are we doing well?'” Sanchez said. The benchmark is just the beginning of inquiry.
Gurley reinforced that there is no “correct” percentile. “We’ve had groups say we’re going to measure at the 40th percentile because that works best for our physicians,” she said. “Others might target the 65th.” Context is important, especially when ranking providers. “It wouldn’t make sense for a new provider to be at the 80th percentile compared to someone in practice for 25 years,” she added.
Best practice, Gurley said, is pragmatic: “Start at the 50th and work your way out.”
Peer-to-peer benchmarking and why filters matter more than averages
MGMA’s dataset is only useful if comparisons are truly comparable. That’s where filters — and thresholds — come in.
Sanchez explained that MGMA enforces minimums before publishing segmented data. “We need a minimum of three groups and at least 10 individual physicians or APPs to publish,” she said. “We never want members benchmarking against two or three providers.”
When filters narrow the field too far, asterisks appear — a signal that statistical reliability is breaking down. Gurley emphasized that this is where human support matters. “We have SMEs who will help you understand why those asterisks are there and do some back digging,” she said.
Geography, ownership model, and practice type all materially affect compensation. “You can compare Eastern, Midwest, Southern regions side by side,” Sanchez said. “Or limit it to just the Western region if that’s where you operate.” Without those controls, “market” becomes an illusion.
From benchmarks to conversations: Using data without weaponizing it
Provider-level benchmarking is where data either builds trust or destroys it.
Sanchez described the common pattern: “You might see a provider at the 40th percentile for work RVUs and the 60th for compensation,” she said. “That’s where education comes in — especially during those MOR conversations.”
Gurley acknowledged this discomfort. “It can be a difficult conversation to lay out a report that says your goal is the 50th percentile and you’re producing at the 30th,” she said. “There’s a lot more to that story — access, scheduling, coding, support.”
MGMA’s tools are designed to support those conversations, not shortcut them. The My Provider Report Card translates internal numbers into percentiles automatically. “It tells you exactly where that provider is next to MGMA’s 50th percentile,” Sanchez said. “You don’t have to figure it out manually.”
The newer pay-to-production plotter visualizes compensation against work RVUs. “Physicians and leaders love to see it,” Sanchez said. “It puts it on a scatter plot so patterns are obvious.”
Total compensation and keeping comparisons truly apples to apples
“Total compensation” means exactly what shows up on a W‑2, and that precision is crucial.
“It includes bonuses and incentives,” Sanchez said. MGMA also separates base compensation, medical directorships, and call pay. “On-call data is for both physicians and APPs, because they both take call,” she noted.
Ownership structure and academic status also change the math. “We do have academic compensation as a separate dataset,” Sanchez said, distinguishing university-based practices from non-academic organizations.
The discipline is in resisting shortcuts. “Too much, too fast is the biggest mistake,” Sanchez warned earlier. “Being under the 50th percentile doesn’t mean leadership is failing, and being over doesn’t mean you take your foot off the gas.”
MGMA data: A tool in the toolbelt
For Gurley, the role of MGMA data is clear. “We want to be a tool in your tool belt,” she said. “We’re not writing the prescription — but we do provide the foundation.”
Sanchez echoed that forward-looking stance. “Never feel lost,” she said. “There is somebody here who will help you get what you need and make it make sense.”
This MGMA data conversation is just the beginning. Tune into the MGMA Podcast Network mid-June for management and staff benchmarks and mid-July for financials and operations.
Resources
Email us at dwilliams@mgma.com if you would like to appear on an episode. If you have a question about your practice that you would like us to answer, send an email to advisor@mgma.com. Don’t forget to subscribe to our network wherever you get your podcasts.
