ICA Revenue Calculator

How the per-episode impression figure is derived:

• Primary source — Nielsen multiplatform ratings: When trades report L+7 or L+35 figures (live + DVR + streaming, captured 7 or 35 days post-airing), we use that as the anchor. Example: a P+ show reported at "4M L+7 multiplatform avg/ep" → 4,000,000 impressions/ep.
• Platform-direct disclosures: Paramount/CBS press releases (e.g. "biggest P+ drama debut ever," "X million viewers in Y days") provide the network's preferred number. We back-calculate per-episode from total/season figures where needed.
• Cross-reference: When 2–3 sources disagree (Nielsen vs. trade vs. platform), we pick the median and document range in our internal notes.
• Confidence: ±20%. These are the highest-quality numbers in the model.

Why this tier is most accurate: Networks and Nielsen actively publish first-run numbers (it's marketing fuel for them). The data is fresh and contested across multiple sources, so triangulation is possible.

How the per-episode impression figure is derived:

• Primary source — Nielsen Top 10 streaming charts: Weekly "minutes viewed" figures for top library titles → divided by average episode runtime → divided by episode count = rough per-episode reach. Example: a show with 800M weekly minutes / 60-min eps / 30 eps available → ~444K viewers/ep/week.
• Halo adjustment: Live catalog gets a measurable bump (typically 1.5×–2.5×) in the 4–8 weeks leading into a new-season premiere as viewers binge to catch up. We model the steady-state, not the halo peak.
• Rule of thumb: When direct chart data is unavailable, live catalog per-ep ≈ 30–50% of first-run per-ep for the same show. We apply 40% as a default mid-point.
• Platform anniversary releases: "X billion minutes streamed of [show] this year" provides annualized totals we deconstruct to per-episode.
• Confidence: ±35%. Less direct measurement; dependent on what gets charted.

What's harder: Most live-catalog viewing happens off the Nielsen Top 10, so a long tail of mid-tier libraries are extrapolated rather than measured directly.

How the per-episode impression figure is derived:

• Primary source — platform retrospectives: Anniversary or milestone announcements ("10 years on Paramount+ — Y billion minutes streamed of [show]") give us total catalog-lifetime figures. Total minutes ÷ total runtime ÷ episode count = average viewers per episode per cycle.
• Category benchmarks: Back-catalog per-ep typically runs 10–25% of original first-run reach in steady state. We apply tier-specific defaults (prestige drama: 20%; broadcast procedural: 15%; sitcom: 25% — sitcoms hold up best in catalog).
• Cultural rediscovery overrides: Some shows defy benchmarks when they hit a wave (e.g. Suits on Netflix in 2023, Yellowstone post-finale). When a back-cat show is in a rediscovery moment, we boost the estimate and flag it as "rediscovery" in notes.
• Episode-count compounding: Back-cat looks low per-episode but a 7-season show has 70+ episodes accumulating revenue year-round, which is why total catalog earnings can still be material.
• Confidence: ±50%. Most uncertain tier. Numbers should be treated as directional planning estimates, not measurement-grade.

How a real deal tightens this: If Paramount shares first-party platform data (their own consumption telemetry), our back-cat estimates can be replaced with measurement-grade numbers per episode.

Audience watches the show on-demand through a streaming subscription service. This is where most premium first-run and library content lives today. Ads appear in the ad-supported tiers; ICA placements live inside the content itself and play across all tiers.

• How it's delivered: Streaming app on connected TV, mobile, web
• Audience profile: Subscriber base; on-demand and binge behavior
• Ad model: Subscription + ad-tier streaming CPMs (typically $25–$32)
• Examples: Paramount+ first-run (Landman, 1923, Tulsa King)
• Measurement: Nielsen streaming ratings, VideoAmp, platform first-party telemetry

In the data: streamingImps field. The "Streaming CPM" input in Step 3 prices this window.

Audience watches the show live (or via DVR) on a traditional channel at a scheduled time. This is "old-school" TV — antenna broadcast (CBS) or cable bundle (Paramount Network). Linear has the largest single-airing reach for hits but is shrinking year over year.

• How it's delivered: Free over-the-air antenna (CBS), cable/satellite/vMVPD subscription (Paramount Network)
• Audience profile: Live + DVR viewers; older skew on average
• Ad model: Traditional 30-second linear spots; Nielsen-rated CPMs ($22–$28 typical)
• Examples: Yellowstone on Paramount Network, Marshals on CBS
• Measurement: Nielsen live + DVR ratings, C3/C7 commercial ratings

In the data: linearImps field. The "Linear CPM" input in Step 3 prices this window. P+ exclusives have linear = 0.

Long-tail on-demand consumption of older or concluded content sitting in a platform's library. Distinct from "first-run streaming" because the audience is steady-state rather than concentrated in a premiere window. Per-episode reach is lower, but a deep catalog accumulates real volume year-round.

• How it's delivered: Same apps as streaming, but for library content (often older seasons or concluded series)
• Audience profile: Discovery-driven, binge-watchers, nostalgia/comfort viewers
• Ad model: Library / back-catalog CPMs (typically lower, $20–$24 range)
• Examples: Yellowstone back-catalog, 1883 library on P+
• Measurement: Nielsen library streaming, platform anniversary disclosures

In the data: vodImps field. The "VOD CPM" input in Step 3 prices this window. Most LIVE first-run shows have vod = 0; BACK-CAT shows lean on this column.