Fulfillment Operations

Demand forecasting that balances wholesale and subscription inventory Intro Wholesale orders and subscription boxes pull inventory in different ways. One is lumpy and big, the other is steady and recurring. Without a balanced demand forecast, brands overbuy, miss ship dates, or hurt loyal subscribers. This guide shows how to build demand forecasting that balances wholesale and subscription inventory, with clear steps, practical methods, and metrics. It also highlights 2026 updates you should factor into planning. Key takeaways Unify B2B and subscription data to see true demand and risk. Use different forecasting methods for lumpy and recurring items. Set channel service targets, then allocate with fair share rules. Build what-if scenarios before accepting large wholesale POs. Automate with your OMS, WMS, and EDI to act on the plan. Measure bias and fill rate by channel to improve every cycle. Table of contents Why balancing wholesale and subscription inventory matters in 2026 What is demand forecasting for dual-channel inventory The data model: what to integrate and why Methods: choosing forecasts for wholesale and subscription Inventory policies and fair share allocation Implementation plan: steps to a balancing workflow Mini case: a DTC brand balancing subscribers and a big-box PO Latest developments affecting forecasting in 2026 Metrics and governance that keep plans on track How Fulfillment Hub USA enables balanced planning and execution FAQ Conclusion External sources Internal link Why balancing wholesale and subscription inventory matters in 2026 E-commerce demand keeps growing, and wholesale deals are getting larger. The U.S. Census Bureau reported new highs for online sales in the fourth quarter of 2025, released in February 2026. That means more volume to plan, store, and ship. At the same time, subscription revenue remains strong across many categories, from consumables to hobby kits, according to industry research in 2025. Wholesale demand is spiky and negotiated. It comes as big purchase orders with strict EDI compliance and delivery windows. Subscription demand is cohort based. It follows renewal cycles, churn, and seasonal box themes. If you pool both without a plan, you risk shorting subscribers to fill a single big-box order, or missing OTIF on that order to protect recurring boxes. Forecasting must match these patterns, then convert them to inventory targets, vendor buys, and staffed capacity by site. The payoff is higher fill rates, fewer stockouts, and lower working capital. In short: Balancing both channels improves service and cash flow at the same time. What is demand forecasting for dual-channel inventory Demand forecasting for dual-channel inventory predicts both wholesale and subscription needs, then sets inventory and allocation rules that protect each promise date. It merges statistical forecasts, business inputs, and constraints like lead times and MOQs. Definition Demand forecasting that balances wholesale and subscription inventory is a planning approach that projects units by SKU and channel, applies safety stock by service targets, and runs allocation rules so no channel cannibalizes the other. Example: Use a cohort model to project 42,000 units for July subscribers and Croston’s method to estimate two wholesale reorders, then set fair share allocation for in-month ATP. In short: It is one plan that respects two very different demand shapes. The data model: what to integrate and why A reliable plan starts with clean inputs. Bring these into one model: Orders and returns by channel, including subscription cohorts and gift cycles. Wholesale POs, retailer forecasts, and EDI messages like 850, 855, 856, and 810 acknowledgments. Prices, promos, and events that shift lift and timing. Lead times, MOQs, and supplier reliability by lane. Stock by site, lots, expiry, and kitting rules or bill of materials. Capacity limits for receiving, kitting, pick and pack, and carrier cutoffs. Model your subscriptions by cohort and SKU. Track churn, reactivation, upsells, and box swaps. For wholesale, map each banner’s calendar, OTIF targets, and chargeback risk. Use ABC-XYZ segmentation to rank items by value and variability. Map substitutes and components for kits to prevent hidden shortages. FHU tip: Connect your OMS and subscription platform to the WMS for daily nettable ATP. That way the forecast flows into waves and labor plans without manual work. In short: One integrated model reduces surprises and supports faster, safer decisions. Methods: choosing forecasts for wholesale and subscription Different demand shapes call for different math. Choose methods by SKU and channel: Subscriptions and stable DTC items: exponential smoothing or ETS, with seasonality and event effects for promos and holidays. Wholesale reorders and lumpy B2B: Croston or intermittent demand variants, plus scenario inputs around line reviews and resets. New items or kits: simple analogs or Bayesian models that borrow shape from similar SKUs until history builds. Short holiday bursts: weekly models with constrained horizons and blackout logic after season end. Comparison of forecasting methods and best-fit use | Method | Best for | Notes and cautions | |——————–|———————————–|—————————————–| | ETS/Exponential | Subscriptions with seasonality | Capture trend and holidays with events. | | ARIMA/Auto ARIMA | Stable items with rich history | Needs stationarity checks and tuning. | | Croston/SBA | Intermittent wholesale reorders | Forecasts demand size and interval. | | Prophet/TBATS | Multiple seasonal patterns | Useful for weekly and monthly cycles. | | Analogs/Bayesian | New launches and kits | Blend priors until data accrues. | Blend statistical forecasts with business inputs. For subscriptions, build from active subscribers by cohort times expected take rate. For wholesale, include retailer guidance and attach rates for displays, multipacks, and shippers. In short: Match the model to the pattern, then add business context. Inventory policies and fair share allocation Once you have a demand plan, set policies that protect both channels. Start with service targets by channel, like 98 percent for active subscribers and 95 percent for wholesale replenishment. Translate targets to safety stock by SKU and site using forecast error and lead time variability. Use postponement when possible, holding generic components until late-stage kitting. Decide whether to pool inventory or segment it by channel: Pooled inventory increases flexibility and lowers total stock, but needs clear allocation rules. Segmented inventory protects key promises, but