Script-data drug sales are gross of rebates and can be 45% above company-reported

Claim

Two published numbers for the same drug, same year, both correct, differing by ~45%:

Source	Wegovy FY2025	Basis
Market/script data (IQVIA-type)	$18.6B	Gross — list price × units
Novo's reported Obesity Care segment	DKK 82.3B ≈ $12.7B	Net — after rebates, discounts, chargebacks

The gap is the gross-to-net spread, which ran at roughly 70% of gross US sales for Novo's GLP-1 franchise in 2025. In US pharma this is not a rounding difference; it is frequently the largest single line between the headline and the P&L.

The trap is that both numbers circulate freely in press coverage without their basis labelled, and both are attributed to the same product and year. A comparison that pulls one competitor's figure from script data and the other's from a 10-K/20-F is corrupted by the entire rebate spread — and the spread differs by drug, by payer mix, and by year.

Guard

1. Label the basis before using any drug revenue figure. Company-reported = net. IQVIA/Symphony/script trackers = gross. Analyst "peak sales" estimates are usually net, but check.
2. Never mix bases in a head-to-head. Both sides from the filings, or both sides from script data. Never one of each.
3. Growth rates are safer than levels — but only if the gross-to-net ratio is stable, which it is not when a franchise is repricing. In 2025–26, GLP-1 net price fell far faster than gross, so gross-basis growth systematically overstates the revenue the company actually books. This is precisely the period where the error is largest.
4. Watch for the same split in "market share." Volume share, value share on gross, and value share on net give three different answers. NVO's 2026 figures ranged from "~55% of US injectable GLP-1 volume" to "39.4% of the US GLP-1 market" — both true, different denominators.

Why it matters beyond arithmetic

The gross-to-net spread is itself a moat metric. A 70% rebate is not a discount, it is a measure of how little pricing power the manufacturer holds against the PBMs. Tracking the spread over time reads directly on competitive position — a widening spread on flat volume means the franchise is being competed away even while gross sales look fine. On NVO the spread widening showed up in reported gross margin (84.7% → 81.0% in one year) before it showed up anywhere in the narrative.

What would falsify this

Nothing about the accounting. The specific ~70% figure is GLP-1-and-2025-specific and will change; the structural gap between gross script data and net reported revenue will not.

Related

[[principle-primary-source-beats-vendor]] — the filing is the net number, and the net number is the one that becomes cash. [[pitfall-vendor-revenue-is-net-construct-for-banks]] — the mirror-image failure, where the vendor reports a net construct and the reader assumes gross. [[theme-glp1-obesity-pharma]].

History

• 2026-07-31 — found during the NVO analysis while reconciling Wegovy's $18.6B press figure against a DKK 82.3B reported obesity segment. Caught only because the two were ~45% apart on the same product and year.