What Is CPFR?

You're forecasting to one number. Your supplier is planning production to a different one. Neither of you knows the other's number exists, and the gap between them shows up months later as either a stockout or a warehouse of dead stock. CPFR exists to close exactly that gap.

CPFR (Collaborative Planning, Forecasting, and Replenishment) is a model where a brand and its trading partners share one agreed demand forecast and jointly plan replenishment. Instead of each side guessing, they align on a single plan. CPFR reduces the demand distortion that drives stockouts and overstock up the supply chain.

Key takeaways

• Two private forecasts are the real enemy: you plan to one number, your supplier produces to another, and the gap surfaces months later as a stockout or dead stock.
• The model is a four-step loop: joint business plan, shared forecast, joint replenishment planning, exception review, repeated every cycle.
• CPFR is partnership; VMI is delegation: both sides keep decision rights in CPFR, while VMI hands the reorder decision to the supplier.
• The lightweight version captures most of the value: one shared forecast, a visible promotions calendar, and a standing monthly exception call.

What is CPFR?

CPFR is a formal way of planning with your trading partners instead of at them. You and a supplier (or a retail partner) agree on one demand forecast, plan replenishment against it together, and meet on a rhythm to handle the exceptions. The output is a single number both sides commit to, rather than two private forecasts that meet for the first time as a late order.

Where the model came from

CPFR was formalized in the late 1990s retail world, born out of pilots between large retailers and their suppliers who kept planning against each other's guesses. The acronym and framework were standardized by the industry body VICS (now part of GS1 US). The context matters only for this: it was built for exactly the situation where two companies' plans depend on each other, which is now everyday reality for any brand whose supplier capacity constrains its growth.

How does the CPFR model work?

The model runs as a four-step loop: agree on a joint business plan, build a shared forecast, plan replenishment together, and review exceptions. Most of the work happens in the last step, because the whole point is that the two sides handle surprises together rather than discovering them separately.

• Joint business plan: agree what you're both trying to do this season (the promotions planned, the launches coming, the service levels expected) so the forecast has shared context, not just shared math.
• Shared forecast: build one demand forecast both sides can see and challenge. It doesn't matter whose system it starts in; it matters that there's one number.
• Joint replenishment planning: turn the forecast into order and production plans together, so the supplier reserves capacity for what you actually expect to need.
• Exception review: meet on a set cadence to work only the gaps, where actual demand is diverging from the shared forecast, and adjust both sides' plans while adjustment is still cheap.

The collaboration loop and shared data

The loop lives or dies on what's shared. A working CPFR relationship refreshes four things on the agreed rhythm:

• The demand forecast itself: the one number both sides plan against.
• Current inventory positions: yours and theirs, so neither side plans blind to stock already in the chain.
• The promotion and launch calendar: sharing sell-through but hiding the promotion that will double it defeats the purpose.
• Known supply constraints: capacity limits, material lead times, and blackout windows, surfaced before they become missed orders.

Most brands run this pragmatically: a shared sheet or portal plus a standing monthly call covers the mechanics for a growing-brand relationship. The formality of the original framework matters far less than the discipline of one forecast and a fixed exception rhythm.

What problem does CPFR solve?

Demand distortion. When each layer of a supply chain forecasts privately and pads its orders for safety, small swings in real demand amplify into large swings upstream: the bullwhip effect. The supplier sees your padded orders, not your real demand, and pads their own production plans on top. CPFR dampens the distortion at its source by letting both sides plan against the same real number, so the padding (and the overstock and shortages it creates) never enters the chain.

One shared number is also an internal discipline before it's a partner discipline. Conative AI gives your own team that single source: demand forecast per SKU across every channel (DTC, wholesale, marketplaces) in one place, updated as sales move. AI-powered demand forecasting keeps the number current; you decide how much of it to put in front of a partner. When the forecast you share is one your own buying already runs on, partner planning meetings get short. See it on your own catalog: see a demo.

CPFR vs VMI: what's the difference?

CPFR is a partnership; VMI is a delegation. Under CPFR, both sides keep their decision rights and align them around one shared forecast. Under vendor managed inventory, you hand the replenishment decision itself to the supplier, who runs it against your data inside agreed limits. CPFR costs more effort and returns more alignment; VMI costs less effort and returns less control.

• Who forecasts. cpfr: Both sides, one shared number; vmi: Supplier, from your data feed
• Who decides replenishment. cpfr: Jointly planned; vmi: The supplier, inside agreed bands
• Data shared. cpfr: Forecast, inventory, promotions, constraints; vmi: Stock levels and sales velocity
• Relationship type. cpfr: Strategic partnership; vmi: Transactional delegation
• Best fit. cpfr: Partners whose capacity constrains your growth; vmi: Steady SKUs you want off your desk

Frequently asked questions

What are the four steps of CPFR?

Joint business planning (agree the season's goals and events), shared demand forecasting (one number both sides see and challenge), joint replenishment planning (turn the forecast into order and capacity plans together), and exception review (meet on a set cadence to work only the gaps between forecast and actuals). The loop then repeats each planning cycle.

Who uses CPFR?

It started with large retailers and their suppliers, and the formal framework still fits that world best. In practice, any brand whose supplier capacity limits its growth runs some version of it: sharing a forecast with a key manufacturer, agreeing capacity ahead of peak, and reviewing gaps monthly. The principles scale down even where the ceremony doesn't.

What's the difference between CPFR and S&OP?

Scope. S&OP (sales and operations planning) is an internal process that aligns your own sales, marketing, finance, and operations around one plan. CPFR does the same job across company lines, aligning you and a trading partner around one forecast. They complement each other: a brand with working S&OP has a much better number to bring into CPFR.

Is CPFR only for big retailers?

No, but the full formal framework mostly is. A growing brand gets most of the value from a lightweight version: one shared forecast with a key supplier, a promotions calendar they can actually see, and a standing monthly exception call. The test isn't ceremony; it's whether both sides plan from the same number instead of guessing at each other.

What data is shared in CPFR?

The core set: the demand forecast itself, current inventory positions on both sides, the promotion and launch calendar, and known supply constraints like capacity limits or material lead times. The rule of thumb is to share whatever would change the other side's plan if they knew it, on a rhythm both sides commit to.

How does CPFR reduce the bullwhip effect?

The bullwhip effect grows because each layer forecasts privately and pads its orders, so distortion compounds upstream. CPFR removes the main input to that padding: the supplier plans against your shared real forecast instead of reverse-engineering demand from your order sizes. Less guessing means less padding, which means smaller swings amplifying up the chain.