Replace the top-line curve with a subscriber flow
A simple forecast takes last year's subscribers and applies a growth rate. That approach is weakest when penetration is high, competition is intense, and households rotate among services. The paper describes precisely that environment. Forecasting should instead begin with a flow: opening subscribers plus gross additions and reactivations, minus voluntary and involuntary churn, equals the closing base.
Build the flow monthly and by meaningful segment. Separate direct, app-store, platform, telco, and retail bundles; premium and ad-supported tiers; new and returning users. The closing base is then the consequence of acquisition and retention assumptions, not an independent target. This makes the plan explainable when actual performance changes.
Forecast acquisition from drivers
Paid additions should be linked to budget, channel-level CAC, conversion, and available audience rather than inserted as a management goal. Organic additions can be tied to brand demand, product launches, content events, seasonality, and historical baselines. Bundle additions require partner-specific launch timing, eligibility, activation, and revenue-share assumptions.
Capacity constraints belong in the model. A campaign may encounter rising marginal CAC as high-intent audiences are exhausted. A bundle partner may report sign-ups that do not become active viewers. A major content release may pull demand forward rather than create entirely incremental subscribers. Scenario ranges should make those uncertainties visible.
Model retention by tenure and event
New subscribers often churn differently from established viewers, so use tenure-based survival curves. Include cancellation, failed payment, pauses, and reactivations separately. Then layer known events such as price changes, the end of a promotion, the conclusion of a sports season, a content release, or a shift in ad load.
The paper emphasizes bundles, engagement, content utility, and predictive churn intervention. Each should become a testable retention assumption. For example, a bundle may reduce cancellation but lower direct economics; an ad tier may preserve a price-sensitive household that would otherwise leave. Forecast the saved contribution, not just the saved subscriber.
Connect every subscriber to monetization
Closing subscribers should feed a revenue and contribution model by tier and channel. Subscription users need realized price after discounts, taxes, platform share, and payment failure. Ad-supported users need active viewing, ad opportunities, fill, effective CPM, and supply costs. Bundled users need partner revenue, engagement, upsell, and data-access assumptions.
This step prevents an attractive subscriber forecast from masking weak economics. Two plans can reach the same closing base while producing very different revenue, CAC payback, retention risk, and first-party data. Management should approve the forecast that produces durable contribution under realistic downside conditions, not the one with the largest undifferentiated audience number.
- Opening active subscribers
- Gross additions by channel and tier
- Qualified bundle activations
- Voluntary and payment-related churn
- Reactivations and tier migrations
- Closing base, revenue, and contribution
Run scenarios and reconcile monthly
At minimum, build base, conservative, and upside cases. Vary the assumptions that management can act on: CAC, organic demand, activation, churn, bundle share, price, ad-tier mix, and advertising yield. Do not vary every input at once. A scenario should tell a coherent story about the market and the organization's response.
Each month, replace forecasts with actuals and preserve the original plan for comparison. Explain variance through acquisition, activation, churn, reactivation, mix, or monetization. Update forward assumptions only when evidence supports the change. This rolling discipline turns subscriber forecasting from a target-setting exercise into a shared learning process for growth, product, content, advertising, and finance.
Decision implication
US streaming subscriber forecasting is most credible when every change in the base has a driver and every subscriber has an economic profile. A cohort-based flow shows where growth comes from, how long it lasts, and whether it creates contribution after acquisition, platform, content, and service costs.
Build a cohort-based US streaming subscriber forecast with PyxiVisio.
Use a PyxiVisio decision-intelligence model to connect assumptions, delivery, economics, risk and approval conditions.