A drug can have strong clinical data, get favorable payer coverage, and still miss forecast by a wide margin. The reason is almost never the molecule. It is the patient who never reached the point where the drug could have worked.
This is the layer of the forecast that most commercial teams do not see until after launch.
The Three Forecast Stages Most Companies Run
Most pharma forecasts are built in three stages, by three different teams, on three different timelines.
Stage one is HCP intent. Modeled by the brand strategy team 24 to 36 months before launch. Asks prescribers what they would do based on the clinical profile.
Stage two is market access. Modeled by a different team 6 to 12 months before launch. Layers in prior auth requirements, formulary restrictions, step edits, and patient cost sharing.
Stage three is patient flow. Often left to a vendor and reviewed quarterly. Tracks the actual journey from pre diagnosis to treatment initiation.
By the time the third number arrives, the first two are locked. The manufacturing plan is committed. The field force is being hired. Then launch happens, and the math turns out to be wrong by an order of magnitude.
The Diagnostic Bottleneck Almost Nobody Models
The hidden bottleneck is often diagnostic and referral leakage. Consider an illustrative NSCLC scenario where a precision therapy requires a positive EGFR mutation test before it can be prescribed.
Journey analytics maps what actually happens. Only 44 percent of patients receive an EGFR test. Only 31 percent have an EGFR positive result returned before first line treatment. 29 percent of those tested start chemotherapy before results come back. Once on chemo, patients rarely switch mid cycle.
By the time the targeted therapy arrives in the addressable patient pool, the system has already halved the eligible population twice. The drug never had a chance, and the commercial team never modeled the chance.
Eligible vs Accessible Patients
This is the difference between eligible and accessible patients.
Eligible is a clinical number, defined by trial inclusion criteria and FDA label. Accessible is a system number, defined by referral patterns, diagnostic infrastructure, prescriber familiarity, and payer architecture.
They are rarely the same, and the gap is where most NPV evaporates.
How the Companies That Get This Right Operate
The launches that capture eligible populations treat the diagnostic and referral system as a commercial activity, not a medical affairs side project.
Bayer and Loxo Oncology invested in pre approval diagnostic education for VITRAKVI rather than waiting until launch to find the bottleneck. Field resources shifted from standard prescriber targeting to the testing and referral system that determines whether eligible patients ever reach the drug.
The companies that wait inherit a leakage map they cannot fix.
Why Single Source Data Cannot Solve This
No single dataset sees the full pathway. Claims data misses lab results. Lab systems sit outside claims. Progression and line of therapy changes are often inferred, not directly observed.
The commercial value is in the linkage, not in any one dataset. That linkage is also where most pharma analytics functions hit a wall, because integrating claims, EHR, lab, specialty pharmacy, and genomics requires capabilities most brand teams do not own.
The Implication for Forecasting
If your forecast does not include a patient flow stress test alongside HCP intent and payer access, you are presenting a forecast that is structurally incomplete. The number you are walking the board through is half the math.
Patient journey analytics is not a marketing dashboard. It is a leakage map. The commercial story is rarely why adoption missed forecast. It is how many patients never made it to the point where the drug could have worked.
The companies that build that map two years before launch are the ones whose forecasts hold. Everyone else is preparing a board meeting they could have avoided.
