Demand Planning and S&OP: Process, KPIs & Execution (2026)

Demand planning and Sales & Operations Planning (S&OP) solve different parts of the same management problem: how to convert uncertain future demand into one feasible, financially aligned operating plan. A forecast estimates demand. A demand plan combines that forecast with commercial intelligence and agreed assumptions. S&OP then tests whether the business can actually supply that demand, what the financial consequences are, and which trade-offs executives should approve.

The distinction matters because many organizations still treat planning as a forecasting exercise. They spend heavily on forecast accuracy while sales, operations, procurement, inventory and finance continue working from different assumptions. The result is familiar: excess inventory in some areas, shortages in others, last-minute expediting, unstable production schedules, missed service targets, and repeated debates about which number is “right.”

A stronger operating model creates a controlled planning cycle. Demand signals are translated into a baseline forecast, exceptions and commercial assumptions are challenged, the demand plan is agreed, supply and capacity constraints are tested, financial implications are reconciled, and executives approve one plan that downstream teams can execute.

Key Takeaways

Demand Planning, Forecasting and S&OP Are Not the Same Thing

Forecasting is the analytical task of estimating future demand. It may use statistical time-series methods, causal models, machine learning, sales input, market intelligence, or a combination of approaches.

Demand planning turns that estimate into a managed view of expected demand. It adds business context such as promotions, pricing changes, launches, customer pipeline, contracts, lost sales, cannibalization, seasonality and known market events. ASCM’s current demand-planning career guidance describes the agreed demand plan as the number handed to supply so capacity and materials can be planned against it.

S&OP sits above that demand plan. Oracle describes S&OP as an integrated planning process that aligns demand, supply and financial planning so executives can approve a feasible and profitable operating plan. In other words, demand planning asks, “What do we expect customers to want?” S&OP asks, “Given what customers may want, what should the company commit to doing?”

That distinction prevents a common failure: asking demand planners to solve supply constraints by lowering the forecast. If expected market demand is 100,000 units but the business can currently supply only 85,000, the demand plan should not quietly become 85,000. The gap is a management decision. The organization may add capacity, prioritize customers, pre-build inventory, use alternate suppliers, change price, delay promotions, allocate scarce supply or deliberately accept lost sales. S&OP exists to make that trade-off visible.

Start with a Clean Demand Signal Architecture

A planning process is only as good as the demand signals it interprets. Historical shipments are useful, but they can be distorted by stockouts, order batching, promotions, allocations, channel shifts and supply constraints. A product that shipped zero units because it was unavailable did not necessarily have zero demand.

Useful signal categories include:

• Actual customer orders and shipments: the most direct evidence of what was requested and fulfilled.
• Lost-sales or stockout indicators: evidence that shipments understated true demand.
• Commercial pipeline: qualified customer opportunities, tenders, contracts and account commitments.
• Promotion and pricing plans: events that can create temporary uplift, pull-forward or cannibalization.
• Product lifecycle information: launches, substitutions, phase-outs and portfolio changes.
• Market and external drivers: regulation, macroeconomic indicators, weather, commodity exposure, competitor actions or sector-specific leading indicators where genuinely useful.

The planning team should define which signals are trusted, at what aggregation level, who owns them, and how late changes are handled. Without those rules, the forecast becomes a negotiation log rather than an analytical baseline.

Segment Demand Before Choosing Forecasting Methods

Not every SKU, customer or market should be forecast the same way. A stable high-volume item behaves differently from a seasonal fashion line, an intermittent spare part or a new product with no history. Segmenting demand allows the planning effort to match the decision problem.

Segmentation can combine volume, variability, lifecycle stage, margin, strategic importance and lead time. The goal is not mathematical elegance. The goal is to decide where sophisticated forecasting adds value, where simple rules are adequate, and where management attention should focus on exceptions rather than every item.

Build a Baseline Forecast Before Adding Human Overrides

A baseline forecast creates a neutral starting point. Human input can improve it when people possess information the model does not have, such as a confirmed promotion, customer shutdown, launch delay or contract win. Human overrides can also make it worse when they simply express optimism, commercial pressure or last-minute reaction.

A disciplined process therefore separates:

1. Baseline forecast: generated from the approved analytical method.
2. Proposed override: who changed it, by how much and why.
3. Evidence: promotion calendar, customer commitment, market event, price change or other documented driver.
4. Post-period review: whether the override added value once actual demand is known.

This creates learning. If a sales team repeatedly raises forecasts without improving outcomes, the issue becomes visible. If commercial overrides consistently improve launch forecasts in a particular market, that information becomes part of the planning design.

Measure Forecast Error, Bias and Stability Together

Forecast accuracy matters, but accuracy by itself is not a complete planning scorecard. IBF’s Summer 2026 guidance makes the same point: the same MAPE can describe very different demand behavior and very different business consequences. Bias and forecast stability should also be monitored.

Forecast error

Error measures the distance between forecast and actual demand. Common measures include MAPE and WAPE. MAPE is intuitive at item level but becomes problematic when actual demand is zero or very small, because percentage error can become undefined or disproportionately large. WAPE aggregates absolute error relative to total actual demand and can be more stable for portfolio reporting, but it can hide poor performance on low-volume or strategically important items.

Forecast bias

Bias asks whether forecasts are consistently high or consistently low. This matters operationally because a planner can achieve a reasonable average error while still carrying a persistent directional problem. Repeated over-forecasting can create excess inventory and markdown pressure; repeated under-forecasting can suppress service and generate expedites.

Forecast stability

Stability measures how much the forecast changes between planning cycles for the same future period. Some change is healthy because new information arrives. Constant large revisions can be destructive because supply, procurement and capacity decisions may have already been made against the previous signal.

Business outcomes

Demand planning should also be connected to outcomes it influences without pretending the planner controls them alone. Useful downstream measures include fill rate, OTIF, inventory turns, excess and obsolete inventory, markdown exposure, premium freight and capacity utilization.

For the inventory consequences of planning decisions, see our guide to Retail Inventory Management. For the order-to-delivery consequences, see How to Improve Logistics Performance.

Run a Demand Review That Produces a Consensus Demand Plan

The demand review should not be a slide presentation where every department announces its own number. Its purpose is to resolve exceptions and agree one unconstrained demand view.

A practical demand-review agenda includes:

• actual demand versus prior plan and explanation of major error;
• forecast bias and significant forecast changes;
• new products, phase-outs and cannibalization;
• promotions, pricing and customer events;
• market intelligence and known demand risks;
• commercial overrides with documented evidence;
• assumptions requiring executive or financial challenge;
• the final consensus demand plan by the aggregation level used for S&OP.

The important point is that the demand review should remain unconstrained. Supply limitations belong in the supply review. Otherwise the organization loses visibility into unmet market demand and quietly turns capacity constraints into lower forecasts.

Supply Review: Test the Demand Plan Against Reality

The supply review asks whether the consensus demand plan can be fulfilled with available and planned capacity, materials, labor, supplier capability, inventory and logistics resources.

Typical questions include:

• Which product families or sites are capacity-constrained?
• Are critical materials or suppliers unable to support the plan?
• What inventory must be built ahead of peak demand?
• Which lead times make late demand changes impossible to absorb?
• Where would overtime, subcontracting, alternate sourcing or premium freight be required?
• Which constraints can be removed economically, and which require allocation or demand shaping?

This is also where planning connects to Supply Chain Resilience. A feasible plan built around one vulnerable supplier, one route or one plant may still be too fragile. The supply review should distinguish nominal capacity from resilient capacity.

Use Scenario Planning for Structural Uncertainty, Not Routine Forecast Noise

Forecasting and scenario planning should complement each other. Forecasts are useful for near-term operating decisions when relationships remain reasonably stable. Scenarios are useful when the structure itself may change.

Examples include:

• a tariff or regulatory change that could alter sourcing economics;
• a new technology that may change demand or capacity requirements;
• a major customer deciding between two incompatible platforms;
• an uncertain market launch where adoption could follow several plausible paths;
• geopolitical disruption that may alter lead times, supplier access or transport routes.

In those situations, the S&OP team can maintain a base operating plan while using scenarios to test trigger points and contingent actions. Our Scenario Planning for Managers guide explains how to construct and wind-tunnel those alternative futures without pretending one is a guaranteed forecast.

Financial Reconciliation: Translate Units into Economics

An S&OP process is incomplete if it balances units but ignores money. Oracle’s current S&OP framework explicitly includes financial review before executive approval. The operating plan should therefore be reconciled with revenue, margin, inventory, working capital and cost implications.

Finance should challenge questions such as:

• Does the consensus demand plan support or conflict with the financial forecast?
• What is the margin mix, not just the volume?
• How much working capital is required to support the inventory plan?
• What overtime, subcontracting, premium freight or alternate-sourcing cost is embedded?
• What would demand shaping, allocation or price changes do to revenue and margin?
• Are planned investments justified by the duration and value of the constraint?

This is where S&OP becomes a business-management process rather than a supply-chain meeting.

The Monthly S&OP Cycle

Oracle’s current planning-cycle model separates product, demand, supply, financial and executive review stages. Organizations can tailor the exact sequence, but the logic is useful because it prevents critical decisions from being mixed together prematurely.

1. Product / portfolio review

Review launches, phase-outs, substitutions, lifecycle assumptions and major portfolio decisions that will affect demand and supply.

2. Demand review

Approve the unconstrained consensus demand plan, major assumptions, exceptions and forecast-risk ranges.

3. Supply review

Test material, capacity, inventory, supplier and logistics feasibility. Create response options for gaps.

4. Financial reconciliation

Translate demand and supply options into revenue, margin, inventory, cost and cash implications. Highlight conflicts with the financial plan.

5. Executive review

Approve the operating plan, settle unresolved trade-offs, assign actions and make explicit decisions about demand shaping, supply expansion, inventory, capital and risk.

What the Executive S&OP Meeting Should Actually Decide

The executive meeting should not re-analyze every SKU or debate basic data quality. Those issues belong earlier in the cycle. The executive agenda should focus on choices that require cross-functional authority.

Typical decisions include:

• which customers or products receive scarce supply;
• whether to add overtime, subcontracting or capacity;
• whether to increase or reduce inventory ahead of a demand event;
• whether to delay or accelerate a product launch;
• whether pricing or promotion should be changed to shape demand;
• which financial target takes priority when revenue, margin, service and cash conflict;
• which risks justify contingency capacity or alternate supply;
• which assumptions trigger a scenario response rather than routine replanning.

Every decision should have an owner, due date and a clear change to the operating plan.

Worked Example: Promotion Demand Exceeds Supply Capacity

A consumer-products business is planning a national promotion for a high-margin product family. The statistical baseline indicates 80,000 units of monthly demand. Sales and marketing expect the promotion to add 30,000 units based on retailer commitments and prior event behavior, producing a consensus demand plan of 110,000 units.

The supply review shows that normal internal capacity can support only 95,000 units for the promotion month. Three options are prepared:

Option	Operating choice	Trade-off
A	Run normal capacity and allocate 95,000 units	Lowest incremental cost, but lost sales and customer-service risk
B	Add overtime and premium material supply to reach 105,000	Higher cost, better service and revenue capture
C	Pre-build 15,000 units in the prior month and support the full 110,000	Working-capital and obsolescence exposure if promotional uplift underperforms

Finance translates each option into revenue, gross margin, inventory and cash effects. The executive team selects the option that best fits the company’s margin, customer and liquidity priorities. The forecast was never “wrong” because capacity was lower. The planning process exposed a decision the organization had to make.

Design the Planning Hierarchy Before You Automate It

One of the least visible but most important design choices in demand planning is the level at which the business forecasts, reviews and approves demand. A retailer may execute at SKU-location level, a manufacturer may plan capacity at product-family level, and finance may budget at region or business-unit level. Those levels must reconcile without forcing every meeting into the lowest level of detail.

A practical hierarchy often includes:

• Execution level: SKU-location, customer-item or another detailed level used for replenishment, production or allocation.
• Planning level: product family, market, channel or region where patterns are stable enough to review and decisions are meaningful.
• Executive level: business unit, portfolio or financial view where trade-offs can be linked to revenue, margin and capital.

Aggregation can improve signal quality because random item-level noise partially offsets at higher levels. But too much aggregation can hide mix risk. A family may be on plan overall while one high-margin item is short and one low-margin item is overstocked. The planning hierarchy should therefore allow managers to move between aggregate decisions and the detailed exceptions that explain them.

The same logic applies to time. Near-term execution may require daily or weekly detail, while S&OP may review monthly buckets across a 12-to-24-month horizon. The right design reflects the company’s lead times, capacity decisions and commercial cycle rather than copying a generic software template.

Use Time Fences to Protect Execution from Planning Noise

Plans become less flexible as they move closer to execution. A forecast change twelve months out may be cheap to absorb. The same change one week before shipment may trigger overtime, premium freight, supplier expedites or customer failure.

That is why mature planning systems use time fences or decision zones. The names differ by organization, but the principle is simple:

• Flexible horizon: capacity, sourcing and inventory decisions can still be changed with limited cost.
• Managed-change horizon: changes are possible but require explicit evaluation of cost, service and operational disruption.
• Frozen or execution horizon: only material exceptions should alter the committed plan.

Demand planners should still record new information inside the near-term horizon, but the business needs rules for whether that information changes execution immediately, creates an exception decision, or is carried into the next cycle. Without those rules, every late sales update can destabilize production and logistics.

New-Product Demand Planning Needs Assumptions, Not False Precision

New products are a special planning problem because there is little or no direct sales history. A model can still help, but the process should make assumptions visible rather than generate an apparently precise number from weak evidence.

Useful inputs include:

• analogous products or prior launches;
• expected distribution, customer reach or channel availability;
• launch timing and ramp assumptions;
• price and promotion plans;
• cannibalization of existing products;
• sales pipeline or customer commitments;
• supply availability and launch constraints;
• scenario ranges for adoption speed.

The review should distinguish the commercial ambition from the demand assumption. If marketing needs 50,000 units to justify the launch economics but evidence supports a base demand plan of 35,000 with an upside scenario of 55,000, those are three different pieces of information. S&OP should decide how much inventory or flexible capacity is justified against that uncertainty.

As actual demand arrives, the launch forecast should be re-estimated quickly. Early-warning indicators can include distribution build, initial sell-through, repeat purchase, pipeline conversion, digital traffic or other category-specific signals. The purpose is to learn faster, not to defend the original forecast.

Define Planning Governance and Decision Rights

S&OP fails when every function participates but no one is accountable. A clear governance model should separate data preparation, recommendation and decision authority.

Role	Primary responsibility	What should not happen
Demand planning	Baseline forecast, exceptions, bias/error analysis, consensus-demand process	Changing the forecast simply to match supply
Sales / marketing	Customer, promotion, pricing and market assumptions	Submitting unexplained optimism as fact
Supply / operations	Capacity, material, inventory and logistics feasibility	Hiding constraints inside a lower demand number
Finance	Revenue, margin, inventory, cash and plan reconciliation	Treating the budget as the demand forecast
Executive sponsor	Resolve cross-functional trade-offs and approve the operating plan	Turning the meeting into passive reporting

Decision rights matter most when the plan is uncomfortable. If demand exceeds capacity, who can authorize overtime? If finance wants lower inventory but service risk rises, who decides the acceptable trade-off? If sales requests a late promotion, who owns the incremental supply and logistics cost? S&OP should answer those questions before the crisis arrives.

Use AI and Planning Technology as Decision Support, Not Governance

Modern planning platforms can improve statistical forecasting, detect anomalies, automate hierarchy calculations and make scenario comparisons faster. Those capabilities are useful, but they do not remove the need for process design and decision accountability.

Technology should ideally help the planning team:

• generate repeatable baselines at the right hierarchy;
• identify material exceptions rather than force manual review of every item;
• track who changed a forecast and why;
• compare forecast versions, bias and stability;
• translate volume scenarios into inventory, capacity and financial effects;
• surface constraint and service-risk exceptions early;
• maintain one controlled set of assumptions across functions.

It should not create an opaque “AI forecast” that nobody understands, nor should a system workflow substitute for executive trade-offs. The best planning technology reduces analytical and administrative effort so people can spend more time challenging assumptions and making decisions.

Common Demand Planning and S&OP Failure Modes

1. Treating sales targets as demand forecasts

A target states what the business wants to achieve. A demand forecast estimates what customers are likely to request under current assumptions. Confusing the two creates systematic optimism and destroys trust in the plan.

2. Hiding constraints inside the demand number

Reducing demand to match supply capacity conceals lost opportunity and makes it impossible to evaluate whether removing the constraint is worthwhile.

3. Excessive manual overrides

If every item is manually changed every cycle, the statistical baseline becomes irrelevant and accountability disappears. Overrides should be exception-based and evidence-backed.

4. One accuracy metric for every demand pattern

Stable, intermittent, seasonal and new-product demand behave differently. The same metric and target should not be applied blindly across the entire portfolio.

5. No financial reconciliation

A unit-balanced plan can still be financially unattractive. Revenue mix, margin, inventory, working capital and incremental supply cost belong inside S&OP.

6. Executive review without decisions

If the meeting only receives updates, S&OP becomes reporting theatre. The process should surface decisions that require executive authority.

7. Constant replanning inside execution horizons

Late forecast changes can destabilize procurement, production and logistics after commitments have been made. Organizations need clear time fences and change-control rules.

A 90-Day Demand Planning and S&OP Implementation Roadmap

Days 1-30: Establish the planning fact base

• Map current forecasts, sales inputs, supply plans and financial plans.
• Agree the planning hierarchy: item, location, customer, product family and financial aggregation.
• Define planning horizon, calendar and decision time fences.
• Baseline forecast error, bias, service, inventory and premium-expedite indicators.
• Identify where multiple conflicting numbers are currently used.

Days 31-60: Build the cross-functional cycle

• Create the demand-review process and override rules.
• Define the consensus demand-plan owner.
• Create a supply-review template for capacity, material and supplier constraints.
• Add financial reconciliation to revenue, margin, inventory and cash.
• Define escalation rules for unresolved gaps.

Days 61-90: Run and stabilize the first cycles

• Run the full planning calendar with named owners and deadlines.
• Keep detailed data correction outside the executive meeting.
• Record executive decisions and changes to the operating plan.
• Review forecast overrides and planning exceptions after actuals arrive.
• Remove reports or meetings that do not influence a decision.

Demand Planning and S&OP Management Checklist

Frequently Asked Questions

What is the difference between demand forecasting and demand planning?

Forecasting estimates future demand. Demand planning combines that forecast with commercial knowledge, product lifecycle information, events, overrides and agreed assumptions to produce the demand view used by the wider business.

What is the difference between demand planning and S&OP?

Demand planning produces the expected unconstrained demand plan. S&OP compares that demand with supply, capacity and financial constraints, develops options, and enables executives to approve one feasible operating plan.

Should S&OP use one number?

The organization needs one approved operating plan for execution, but that does not mean uncertainty should be hidden. Scenario ranges, assumptions and risks can remain visible around the approved plan. “One plan” means aligned execution, not false certainty.

Is MAPE the best forecast-accuracy measure?

No single measure is best for every demand pattern. MAPE is easy to understand but performs poorly when actual demand is zero or very small. WAPE can be useful at aggregate level, while bias and stability should be monitored separately. The right scorecard depends on the product mix and the decisions the forecast supports.

How often should S&OP run?

A monthly cycle is common because it balances strategic decision-making with operational responsiveness, and Oracle’s standard planning-cycle design is organized around recurring product, demand, supply, financial and executive review stages. Businesses with shorter cycles can supplement monthly S&OP with weekly exception management without turning the executive process into daily scheduling.

Where does scenario planning fit into S&OP?

Scenario planning is useful when uncertainty is structural rather than routine, such as regulatory change, new technology, geopolitical disruption or highly uncertain market adoption. It can sit around the base S&OP plan by defining alternative assumptions, trigger points and contingent actions.

Primary References

• Oracle: What Is Sales and Operations Planning (S&OP)? – current definition of S&OP as an integrated demand, supply and financial planning process for executive decision-making.
• Oracle: Overview of Planning Cycles – product, demand, supply, financial and executive review stages.
• Oracle 26B: Executive Reviews – current S&OP executive-review measures and decision visibility.
• ASCM: How to Become a Demand Planner (2026) – current demand-planning roles, consensus demand plan and performance measures.
• APQC: End-to-End Sales and Operations Planning Process and Measures – 2025 process map and shared KPI framework.
• IBF: Beyond Forecast Accuracy – Summer 2026 discussion of forecast error, bias, stability and business context.