S&OP: From Monthly Meeting to Profit Engine

What Is S&OP? A Definition That Actually Matters

Sales and Operations Planning (S&OP) is a cross-functional decision cycle — typically monthly — that synchronizes demand signals, supply capacity, inventory positioning, and financial targets into a single operating plan. It is not a forecasting exercise. It is not a reporting ritual. It is the mechanism by which your company decides where to deploy capital over the next 3 to 18 months.

That distinction matters. Most teams treat S&OP as the meeting where departments align their numbers. High-performing supply chains treat it as the moment where the business commits to a plan — and holds people accountable to it.

The core promise: one number everyone owns

The whole architecture of S&OP collapses into one idea: one number. Not sales' number. Not supply's number. Not finance's budget-anchored number. One one-number forecast powered by Celes that every function has reviewed, challenged, and committed to — the single baseline that drives purchasing decisions, replenishment orders, promotional buys, and capacity calls simultaneously.

When that number doesn't exist — when each function defends its own version — the plan is fiction. And fiction is expensive.

Why S&OP lives at the VP level, not the planner level

Planners can build the data. Only a VP can resolve the trade-offs. When demand wants to carry 60 days of cover on a promotional SKU and finance wants 30, that's not a forecasting disagreement — that's a capital allocation decision. S&OP exists precisely because those decisions require authority, not consensus-by-exhaustion.

---

The Five Steps of the Monthly S&OP Cycle

The canonical S&OP process has five steps. Most companies execute all five. Most companies still get the output wrong — because the steps are structurally sound but informationally broken.

Step 1: Data gathering — garbage in, garbage out

The cycle opens with pulling historical sales, current inventory positions, open purchase orders, and promotional calendars. When this data lives in four systems with different update frequencies, the demand review that follows is built on a foundation of noise. A broken output here looks like a demand team spending 40% of its time reconciling data instead of interpreting it.

Step 2: Demand review — building the unconstrained plan

The demand review produces an unconstrained forecast: what the business would sell if supply were infinite. This is where statistical models, market intelligence, and commercial judgment combine. The broken version is a spreadsheet where sales managers override the model upward — because their bonus depends on volume — without documenting the rationale.

Step 3: Supply review — stress-testing capacity and replenishment

Supply takes the unconstrained demand plan and checks it against real constraints: lead times, supplier capacity, DC throughput, working capital limits. The broken version is a supply team that systematically deflates the demand signal to protect operational comfort, creating a gap before the conversation even starts.

Step 4: Pre-S&OP — closing the gaps before the boardroom

Pre-S&OP is where the operational teams — demand, supply, finance, commercial — reconcile their numbers before the executive meeting. Gaps that can be resolved at this level should never reach the boardroom. A broken pre-S&OP produces a list of open issues instead of a list of decisions — and dumps political disputes onto executives who have 60 minutes and no operational context.

Step 5: The executive S&OP meeting — decisions, not debates

The executive meeting should ratify one plan and make the handful of calls that require VP-level authority: approve a promotional inventory build, authorize an emergency import, release or freeze a SKU rationalization. If the meeting is still debating whether the forecast is right, the process upstream failed.

---

S&OP vs S&OE: Same Goal, Different Clocks

S&OP operates on a monthly cadence with a horizon of 3 to 18 months. Sales and Operations Execution (S&OE) operates weekly or daily, managing short-horizon deviations from the approved plan — a supplier that missed a shipment, a promotional spike that beat forecast by 40%, a weather event that crushed footfall in 12 stores.

S&OE does not replace S&OP. It protects the plan between cycles.

Where S&OP ends and S&OE begins

S&OP sets the baseline: volume commitments, inventory targets, supplier purchase cadence. S&OE asks: are we tracking to that plan this week? If not, what do we correct right now — without waiting for next month's cycle?

Why retailers need both: the stockout problem that monthly cycles miss

A monthly S&OP cycle cannot see a stockout that develops on a Tuesday in a specific store because a replenishment order was delayed 48 hours. That gap is a S&OE problem — a daily-cadence problem. For a retailer running 40,000 SKUs across 500 stores, the difference between a plan and execution is measured in thousands of micro-decisions per day that no monthly meeting can govern.

A stockout is not a logistical inconvenience. It is revenue that never existed — because the customer went to another store or switched brands permanently.

---

Why S&OP Breaks: The Multiple-Forecast Problem

Here is the structural failure mode that nobody names correctly.

Sales inflates the forecast to protect service levels and hit revenue targets. Supply deflates it to protect capacity and avoid being blamed for overstock. Finance anchors to the budget number from six months ago, regardless of what the market is doing. The result: three numbers, zero accountability, and a meeting that produces a negotiated average rather than a committed plan.

The political economy of forecast bias

Each function's bias is rational from its own perspective. Sales gets penalized for stockouts. Supply gets penalized for excess inventory. Finance gets penalized for missing budget. No one is penalized for the collective cost of misalignment — because no one measures it. That asymmetry is the root cause, not a lack of process discipline.

How siloed numbers destroy working capital

When three departments run three forecasts, purchase orders reflect a blend of political compromise, not market reality. The result is inventory in the wrong locations — too much in the DC, too little at point of sale — and a working capital position that looks acceptable in aggregate but is structurally misallocated.

The hidden cost: trapped inventory vs. stockouts

The two failure modes appear to be opposites. They are the same disease. Overstock in one location coexists with stockouts in another because the supply signal never unified. You are simultaneously over-invested and under-available.

---

The Financial Stakes: A Working-Capital Calculation You Can Run Today

Formula: cost of excess inventory

Cost of excess inventory =

Excess inventory ($) × Cost of capital (%) × Days of excess cover / 365

Suppose your cost of capital is 14% annually — not unusual in a high-rate Latin American market — and you are carrying 17 days of excess cover on a category with $8M in average inventory. That is:

$8M × 14% × 17 / 365 = ~$52,000 per month in financial drag — from one category, from one root cause.

Formula: revenue at risk from stockouts

Revenue at risk =

Baseline daily sales ($) × Stockout rate (%) × Days out of stock × (1 − recovery rate)

Recovery rate accounts for customers who return when the product is back. For a fast-moving consumer brand, recovery is often low — the substitution happens in the aisle, not later.

Suppose your DC holds 45 days of cover when 28 is optimal — what does that gap cost?

Take a retailer with $200M in annual COGS across replenished categories. Average inventory at 45 days of cover: $24.7M. Optimal inventory at 28 days of cover: $15.3M. The gap is $9.4M in trapped capital.

At a 14% cost of capital, that gap costs $1.3M per year in pure financial drag — before accounting for shrink, obsolescence, or markdown risk on slow-moving SKUs. That number is CFO-ready. Walk it into your next budget review.

---

What a Great S&OP Meeting Actually Looks Like

A high-functioning executive S&OP meeting has one version of the plan on screen. Not multiple tabs. Not competing slides from different functions. One number, one view, one conversation.

Agenda design: 60 minutes, five decisions

The meeting should not last more than 60 minutes if pre-S&OP did its job. A working agenda:

• Performance review (10 min): How did last month track against plan? Where were the gaps and what drove them?

• Demand outlook (15 min): What does the next 3-month horizon look like? What assumptions have changed?

• Supply & inventory positioning (15 min): Can supply deliver against the demand plan? Where are the constraints?

• Gap resolution (10 min): The two or three open items from pre-S&OP that require executive authority.

• Decisions and owners (10 min): Every open item gets a name and a date. No exceptions.

The one-number forecast as the meeting's foundation

The meeting only works if the number entering the room is not a negotiation. It needs to be a statistically grounded baseline that each function can challenge with data — not override with authority. When the baseline is credible, the conversation shifts from "whose number is right" to "what trade-offs do we accept."

How Latin American retailers like Celes clients use the meeting to align promotional buys with replenishment cycles

Promotional events in Latin America — Buen Fin in Mexico, CyberLunes across the region, local festivities by country — compress enormous demand into short windows with long lead times behind them. Aligning the S&OP cycle to pre-authorize inventory builds for those windows — at the SKU and store level — is the difference between capturing the event and running out by noon on day one.

---

How AI Changes the One-Number Forecast at SKU × Store × Day

The traditional S&OP baseline is built in monthly buckets, aggregated at the category or brand level, then allocated down by rules. That allocation is where the error compounds. AI-generated demand signals work in the opposite direction: granular first, aggregated second.

From monthly buckets to daily replenishment signals

When the forecast is generated at SKU × store × day, the S&OP plan stops being a volume commitment and becomes a replenishment schedule. Every purchase order is traceable to a demand signal. Every inventory decision has a specific store and date attached to it. The plan is not a target — it is a set of operational instructions.

How Celes generates the one-number forecast that feeds S&OP

AI-driven demand planning from Celes generates forecasts at the SKU × store × day level — incorporating seasonality, promotional calendars, weather signals, and local demand patterns — to produce a single, statistically grounded number that replaces the negotiated consensus. Mi Comisariato in Ecuador runs perishables forecasting at exactly this granularity, where the cost of a one-day overstock is not a working capital issue — it is a shrink event. Olímpica in Colombia reduced stockouts by 85% operating with Celes replenishment signals feeding their execution cycle. SanaSana reduced total inventory by 37% while maintaining availability. These are not forecast accuracy achievements. They are capital deployment outcomes.

The link between forecast granularity and purchase order accuracy

When the forecast is monthly and aggregated, the purchase order is a guess. When the forecast is daily and granular, the purchase order is a calculation. Pycca in Ecuador saw a 490% improvement in Forecast Value Added in long-lead-time categories — the exact segment where purchase order accuracy determines whether you are over-inventoried for 90 days or out of stock at launch.

---

Is Your S&OP Cycle Managing Inventory — or Just Documenting It?

Here is a diagnostic worth running before your next S&OP cycle.

Ask yourself three questions:

1. When the meeting ends, does each function walk out with a different number than the one they walked in with? If not, you ran a reporting session, not a planning process. Real S&OP changes commitments.

2. Can you trace a purchase order from your last cycle directly to a specific demand signal, store, and week? If your orders are driven by gut feel or category-level allocations, your S&OP plan is not actually governing your replenishment.

3. If sales and supply disagreed on the forecast last month, who resolved it — and with what data? If the answer is "the VP split the difference," you don't have a one-number forecast. You have a politically smoothed average that no function owns.

The monthly S&OP meeting is not the problem. The problem is the information environment it operates in. When the forecast entering the room is a negotiated compromise built from siloed systems, the most disciplined process in the world will ratify the wrong plan.

Fix the number first. The meeting follows.

---

Frequently Asked Questions

Q: What is S&OP in plain terms?

S&OP (Sales and Operations Planning) is a monthly cross-functional process that aligns demand forecasts, supply capacity, inventory positioning, and financial targets into a single operating plan. The output is one number — owned by all functions — that drives purchasing, replenishment, and capital allocation decisions for the next 3 to 18 months.

Q: What is the difference between S&OP and S&OE?

S&OP is strategic and monthly, setting the plan for the medium-term horizon. S&OE (Sales and Operations Execution) is tactical and weekly or daily, managing short-term deviations from the approved plan. S&OE does not replace S&OP — it executes and protects the plan between monthly cycles.

Q: Why do most S&OP processes fail?

The most common failure mode is the multiple-forecast problem: sales, supply, and finance each maintain their own forecast driven by their own incentives. The result is a negotiated average that no function truly owns, which translates into purchase orders that misallocate inventory — excess in some locations, stockouts in others.

Q: What does a one-number forecast mean in S&OP?

A one-number forecast is a single, statistically grounded demand baseline — reviewed and accepted by all functions — that replaces department-level forecasts in the S&OP process. When the number is credible and granular (ideally at SKU × store × day), the executive meeting shifts from debating whose forecast is right to making capital deployment trade-offs.

Q: How does AI improve the S&OP process for retailers?

AI generates demand forecasts at the SKU × store × day level, eliminating the aggregation error that compounds when monthly category-level plans are allocated down by rules. For retailers like Olímpica and SanaSana, AI-driven demand signals from Celes translated directly into measurable outcomes: 85% reduction in stockouts and 37% reduction in total inventory, respectively — without sacrificing product availability.