How long does each piece of content earn its keep?
№ 083 · Content cost amortization curve
Definition
MetricCumulative share of content cost amortized since first availability
Unit% of capitalized content cost expensed by year N
Single line showing content cost recovery per month for a content release. Content that does not get viewed in the first 90 days probably never pays back. Build by allocating content cost to view-driven retained revenue per period. Example: $8M show, $1.2M month-1 recovery, plateaus at $6.4M by M9 (80% recovery).
Benchmarks
| Bottom 30% | Median | Top 30% |
|---|---|---|
| Straight-line or 8-10 year schedules on front-loaded viewing, deferring cost past consumption | Over 90% expensed within 4 years of first availability, ~45% annual declining-balance equivalent (large-streamer policy reference) | Curve tracking viewing: ~40-50% expensed in year 1, declining-balance thereafter |
Subscription-monetized original and licensed video content, US GAAP film-cost accounting, policies disclosed by public streamers 2018-2025. This is a policy shape reference, not a performance percentile distribution. · No cross-operator distribution exists; the reference points are audited disclosures. Netflix: over 90% within 4 years, film faster than series, 10-year cap. Discovery historically expensed 40-50% of program cost in year 1 over at most 4 years for flagship networks. Content with genuine long-tail replay (kids, evergreen catalog) legitimately supports a flatter curve; the failure mode is a flat curve on front-loaded viewing.
Category split omitted: Amortization schedules are disclosed per company under GAAP; no two independent Tier 1-2 sources publish a multi-operator distribution by content category.
When it looks bad
The amortization curve sitting well below the cumulative viewing curve for the same titles, so cost is recognized after the audience has already left the content.
A slate delivering 80% of lifetime views in months 1-6 while 45% of its cost is still unamortized after year 2.
What to do about it
- Rebuild curves per content class from observed viewing decay: license deals and topical titles on steep 1-2 year curves, evergreen kids and catalog titles on flatter curves justified by measured replay, mirroring the film-faster-than-series logic in Netflix's policy
- Run a quarterly impairment screen on titles with viewing below plan; WBD booked $326M of content impairments in 2023 when the slate was realigned, and small operators need the same discipline at their scale
- Window or license out under-viewed originals to third parties to recover residual value before the asset fully amortizes with no audience
- Tie greenlight ROI models to the amortization curve, not cash cost, so a title's payback is judged against the years its cost actually hits the P&L
Sources
- Netflix / SEC Amortization accelerated over shorter of availability window, estimated use, or 10 years; on average over 90% amortized within 4 years of first availability; film faster than TV sec.gov ↗
- Warner Bros. Discovery / SEC $14.6B production costs grouped by monetization strategy (individually vs as a group); acquired libraries amortized straight-line or accelerated sec.gov ↗
- Warner Bros. Discovery / SEC Released group-monetized content expected amortization front-loaded: $2,312M year 1, $1,284M year 2, $804M year 3 sec.gov ↗
- Discovery Communications / SEC Accelerated pattern expensing approximately 40-50% of program cost in year 1, total life no more than four years for flagship networks sec.gov ↗
- Accounting Perspectives (Wiley) Netflix's 90%-in-4-years disclosure equates to a declining-balance amortization rate of approximately 45% annually onlinelibrary.wiley.com ↗
- Behind the Balance Sheet Independent scrutiny of whether the 4-year accelerated curve matches an ageing content asset base; useful counterweight when setting your own curve behindthebalancesheet.substack.com ↗