When evaluating a healthcare market, organizations are usually trying to answer two fundamental questions: How big is the opportunity, and where should we focus our efforts?
Answering these questions requires looking at healthcare activity through two distinct data lenses: observed claims and estimated volumes. Observed claims capture documented clinical activity and provider relationships, while estimated volumes apply statistical modeling to measure the likely size, direction, or potential of a market.
Neither approach is inherently better. The path forward depends on matching the data to the decision. As a guiding principle, we recommend using estimates when deciding how large the opportunity is and turning to observed claims when deciding whom to act on.
Why the distinction matters
Claims data provides a critical window into real-world healthcare utilization in the U.S., but no claims dataset captures all healthcare activity.
Observed claims provide a specific, traceable view of documented clinical activity. However, factors such as payor mix, geography, care setting, and data source coverage can limit how fully observed claims represent total market volume. This happens for a variety of reasons including that not all clearinghouses and payors vend their data, cash pay care doesn’t generate a claim for submission to a payor, and some IDNs own their own payors. Modeled estimates can help address these gaps by providing a more complete view of market activity, but they rely on statistical assumptions that introduce a degree of uncertainty.
Neither approach is inherently superior. They each offer different lenses on healthcare activity, with distinct strengths and trade-offs. The key is understanding what each data type can—and cannot—tell you and matching it to the specific decision at hand.
Observed claims vs. estimated volumes
| Dimension | Observed claims | Estimated volumes |
| Primary value | Specificity, traceability, and verifiable real-world behavior | Broader market coverage, baseline comparability, and scale |
| Trade-off | Inherently partial view; can understate total market activity | Introduces statistical assumptions and modeled inference |
| Core function | Measure what is directly observed in the data | Estimate what has likely occurred beyond what was directly observed |
When to use observed claims
Observed claims are most appropriate when you need to identify or understand documented entities, events, or relationships. Consider the following use cases:
- Provider and account targeting. Prioritize providers and organizations with documented patient or procedure activity.
- Clinical trial site selection. Identify sites treating real patients who may meet relevant eligibility criteria.
- Referral and influence mapping. Measure actual patient movement and documented provider-to-provider relationships.
- Patient journey analysis. Reconstruct real sequences of diagnoses, treatments, procedures, and care settings.
- Provider and organizational profiling. Understand documented clinical volume, specialty focus, and treatment behavior.
For these use cases, the value of claims data lies in its precision. Observed claims provide concrete evidence of where care was delivered, which providers and organizations were involved, what services were performed, and how patients moved through the healthcare system.
When to use estimated volumes
Estimates are most appropriate when the objective is to understand market magnitude, opportunity, or direction. Common use cases include:
- Market sizing and TAM. Quantify the likely total market rather than only the portion visible within a particular claims dataset.
- Territory design and resource allocation. Compare geographic opportunities without allowing differences in claims capture to distort the results.
- Market trending. Distinguish changes in underlying market activity from changes in data contribution or claims capture.
- Market baselining. Establish a stronger view of current market activity to support downstream forecasting and planning.
- Cross-market or cross-payor comparisons. Normalize systematic differences in visibility across populations and segments.
Estimates attempt to correct for under-capture of observed claims, making them better suited to questions about overall scale, relative opportunity, and market direction.
The hybrid approach: Estimation informed prioritization based on observed claims first
In many workflows, the strongest approach is not choosing between observed claims and estimates but learning how to use both in tandem, matched to the right decision. A typical sequence looks like this:
Consider a medical device company assessing a new market. Early in the process, the strategy team needs to understand the overall market opportunity, including:
- How many procedures are performed nationally?
- Is demand growing or declining?
- Which regions represent the greatest potential?
For these questions, estimated volumes provide a more complete view of the market. Later, the commercial team needs to determine where to focus field resources. For example:
- Which health systems are performing the procedure?
- Which surgeons are actively treating patients?
- Which accounts should sales teams prioritize?
For these decisions, observed claims provide the documented activity needed to act.
Together, observed claims and estimated volumes provide both the context to plan and the evidence to act. In practice, this might look like: “We observe 12,000 procedures within this integrated delivery network (IDN) and estimate total volume at roughly 18,000.” This approach gives the account team a documented figure to stand behind and a modeled figure to plan against.
Common misapplications
Aligning the wrong dataset to a decision can risk building an inaccurate view of the market or acting on insights that are not well aligned with reality, which can misdirect time and resources.
Common misapplications include:
- Sizing a market from observed claims. This can understate the opportunity because observed data represents only captured activity.
- Targeting from estimates. This can direct resources toward providers or accounts whose modeled volume is not supported by sufficient observed evidence.
- Trending raw observed volume. A change in observed claims may reflect a change in data capture rather than a true change in the market.
- Treating capture rate as universal. Capture can vary by payor, geography, procedure or therapy, care setting, provider type, and time period. A single enterprise-wide capture-rate assumption can create false precision.
Match the data to the decision
Observed claims ground decisions in documented activity. Estimates provide a broader view of market scale and opportunity. The goal is not to choose one universally. It’s to match the data to the decision in front of you. When used together, observed claims and modeled estimates complement one another, giving your organization a more credible and useful view than either approach could provide on its own.
Does your organization have a complete view of the market, or are gaps in visibility shaping your decisions? Request a demo of Definitive Healthcare to see how using observed claims with modeled estimates narrows your focus to the opportunities that matter most.
