S&OP Guide: What It Is, the 5-Step Process and How It Differs from IBP

What S&OP is

S&OP stands for Sales and Operations Planning. It is the monthly process through which a company agrees on a single demand and supply plan for the coming months, backed by sales, operations and finance, with decisions approved by leadership.

The short definition hides three ideas worth separating:

• It is a process, not a meeting. It has a fixed calendar, defined inputs for each step and outputs that feed the next. The executive meeting is only the last of five steps.
• It produces one number. The demand plan leadership approves is the one purchasing, production, logistics and finance use. There is no sales version, operations version and budget version.
• It ends in decisions. If the cycle closes without someone deciding something (approve the plan, accept a gap, authorize spending), it was a report, not S&OP.

The concept was formalized by Oliver Wight in the 1980s and has since been adopted by most manufacturing and distribution companies. What has changed is the tooling: today statistical forecasting and machine learning do much of the work that used to be done by hand.

What problems it solves

Without S&OP, each area plans with its own version of the future. Sales promises what it wants to sell, operations produces what it believes will sell and finance budgets what it needs to happen. The three versions coexist until they collide in the warehouse.

The symptoms are always the same:

• Stock-outs on what moves and overstock on what does not. Total inventory can look reasonable while the mix is wrong.
• Permanent firefighting. Purchasing and production live putting out fires because nobody told them about the promotion or the new customer.
• A forecast nobody believes. Sales inflates it to secure product, operations cuts it to avoid inventory, and nobody defends the final number.
• Budget and operating plan disconnected. Finance finds out at close that sales did not arrive or that working capital went into inventory.

S&OP does not remove uncertainty. What it does is make the whole company work with the same uncertainty, measure it and decide on it once a month.

The monthly cycle in 5 steps

The classic process has five steps repeated every month. Names vary by author; the logic is always the same: first the data, then demand, then supply, then money and finally the decision.

Step 1. Data and portfolio review (days 1 to 3)

The previous month is closed: actual sales, inventory, open orders, returns. The product master is updated with launches, discontinuations and substitutions. The forecast error of the previous cycle is computed.

Output: a clean database and the forecast accuracy report. If this step fails, the rest of the cycle argues over the wrong numbers.

Step 2. Demand review (days 4 to 8)

The planner generates the statistical forecast per item from the clean history. Sales and marketing enrich it with what the data does not know: planned promotions, new customers, price changes, tenders. The result is the unconstrained demand plan, that is, what the market would buy if we could supply it.

Output: demand plan by family and item for the planning horizon (usually 3 to 18 months), with an owner who defends it and documented assumptions.

Step 3. Supply review (days 9 to 12)

Operations, purchasing and logistics check the demand plan against production capacity, supplier lead times and available inventory. Gaps appear: what cannot be supplied on time, what would require overtime or early inventory.

Output: supply plan with the gaps identified and two or three options for each, with their cost.

Step 4. Reconciliation or pre-S&OP (days 13 to 16)

Finance translates the scenarios into money: revenue, margin, working capital, obsolescence risk. Area managers resolve what they can among themselves and prepare for leadership only the decisions that exceed them.

Output: a short document with the recommended plan, the open gaps and the decisions requested from leadership, each with options and financial impact.

Step 5. Executive S&OP meeting (days 17 to 20)

Leadership reviews the plan, decides on the gaps and allocates resources. It lasts an hour if the previous steps were done well. Decisions are recorded with owner and date.

Output: the approved plan, which from that moment is the only number for purchasing, production, logistics and finance. The cycle starts again.

Roles and calendar

Role	Who it usually is	Responsibility in the cycle
Process owner	Supply chain or planning manager	Coordinates the calendar, prepares materials, facilitates meetings
Demand planner	Planning analyst	Generates the statistical forecast, consolidates sales adjustments, measures error
Demand owner	Commercial manager	Defends the demand plan and its assumptions
Supply owner	Operations or purchasing manager	Presents capacity, gaps and options
Finance	Controller or finance manager	Values scenarios, reconciles with budget
Sponsor	General manager	Chairs the executive meeting and makes the decisions

The calendar is set once and does not move. If the executive meeting is the third Thursday of every month, it always is. Processes that get rescheduled "because this month is complicated" die within six months.

S&OP KPIs

S&OP without indicators is a conversation of opinions. The minimum KPIs are those that measure the three promises of the process: forecast better, hold the right inventory and serve the customer.

• Forecast accuracy: MAPE or WAPE by family and item, and bias (whether the error always goes the same way). Measured with the cycle lag: the forecast made a month ago against this month\'s sales.
• Inventory health: coverage in days, turns, excess and obsolescence.
• Service: fill rate and complete on-time orders (OTIF).
• Plan adherence: how close what was executed (purchasing, production) was to what was approved. This one says whether the process is respected.

We have an article dedicated to the formulas and how to organize them in a dashboard: demand planning KPIs.

S&OP, S&OE and IBP: how they differ

The three terms appear together and get confused. They differ by horizon and by who decides.

	S&OE	S&OP	IBP
Full name	Sales and Operations Execution	Sales and Operations Planning	Integrated Business Planning
Horizon	0 to 3 months, weekly	3 to 18 months, monthly	12 to 36 months, monthly or quarterly
Question	Are we hitting this month\'s plan?	Which plan do we approve?	Does the plan meet strategy and budget?
Unit	Units and orders	Units and money	Money, scenarios, portfolio
Who decides	Operations and sales	General management	Executive committee

S&OE is short-term control: reallocating inventory between warehouses, expediting an order, serving an urgent request. It does not change the plan, it executes it. IBP is the evolution of S&OP once the process is mature: the horizon lengthens, finance moves to the center and portfolio, capacity and investment scenarios are planned. In practice, almost every company first needs an S&OP that closes cycles with decisions. Jumping to IBP without that is renaming the meeting.

Tools: Excel or software

S&OP does not require software to start. It requires clean data, a forecast that can be trusted and calendar discipline. What does change with size is where the forecast lives:

• Well-organized Excel works for small catalogs and a single planner, as long as the forecast is not a moving average dragged by hand.
• ForecastSolve is the next step without leaving Excel: an add-in that runs 23 models per item, classifies demand and freezes every cycle to compare plan against actuals.
• DemandSolve is for when several planners, sales and finance have to work on the same plan, with roles, approvals and optional ERP integration.

The practical rule: the tool must save every cycle as a snapshot. If March\'s forecast disappears when you recalculate in April, you cannot measure error or adherence, and the process does not learn.

Mistakes that make S&OP fail

1. Starting with the executive meeting. Without steps 1 to 4, leadership gets a list of problems instead of prepared decisions. By the third meeting they stop coming.
2. Discussing items instead of families. The executive meeting works at product family level and in money. SKU detail is resolved in steps 2 and 3.
3. A forecast nobody measures. If the error is not computed every month, sales keeps inflating and operations keeps cutting, and nobody knows who was right.
4. Confusing the demand plan with the budget. The budget is a financial target set once a year. The demand plan is the best estimate of what will happen, updated every month. When the second is forced to match the first, the only useful information is lost.
5. No owner and no fixed calendar. The process gets rescheduled, materials arrive on meeting day and decisions are not recorded.
6. Too many KPIs. Ten well-defined indicators are worth more than forty nobody understands. Start with accuracy, coverage, fill rate and adherence.
7. Waiting for the perfect software. Processes that wait for the system implementation never start. Start with what you have and change tools when the process asks for it.

How to start: the first cycle in 90 days

A realistic plan for a mid-sized company with no process today:

1. Weeks 1 to 3: diagnosis and data. Inventory of data sources, product master cleanup, first measurement of current forecast accuracy (even if it is a moving average, to have a baseline). Definition of families, horizon and calendar.
2. Weeks 4 to 6: forecast and KPIs. Set up the statistical forecast per item and the minimum dashboard: accuracy, coverage, fill rate. Agree definitions with each area.
3. Weeks 7 to 9: first full cycle. Run the five steps with the final calendar, even if materials are imperfect. Record the decisions.
4. Weeks 10 to 13: second cycle and adjustments. Measure the error of the first forecast, fix what did not work in the meetings, lock the process.

From the third cycle the process starts to sustain itself, as long as the sponsor keeps chairing the executive meeting. If you would rather do it with support, at DataSolve we implement exactly this: S&OP consulting, from diagnosis through the first cycles.