Demand Planning and S&OP in Fast-Moving Consumer Goods (FMCG): 5 Critical Topics

In brief: Five topics determine the success of planning in fast-moving consumer goods (FMCG) companies: separating the promotion effect from baseline demand, new products and portfolio changes, seasonality and shelf life, channel and distributor stocks, and linking the monthly S&OP process with weekly execution (S&OE). This article covers each topic with examples, explains the metrics to be monitored, and touches on the technologies used. The institutions to which training and/or consulting services are provided in the fast-moving consumer goods sector are listed later in the article.

What Makes Planning Difficult in Fast-Moving Consumer Goods?

In fast-moving consumer goods, a broad product portfolio, frequent campaigns, short shelf life, multi-channel distribution (modern retail, traditional trade, e-commerce, food service), and the expectation of short lead times are managed at the same time. The portfolio can reach hundreds of product codes, and each product is sold across multiple channels under different price and campaign conditions.

For this reason, planning deviations come back in two directions: on one side, shelf gaps and lost sales; on the other, stock that is approaching expiry, sold at a discount, or destroyed. For example, if a beverage producer cannot produce enough during an unexpected heat wave, shelves empty out and the sale is permanently lost; in the post-campaign period, the overproduced product approaches its expiry date and is sold at a discount or destroyed.

In fast-moving consumer goods, demand planning is not just a statistical forecasting task; it is a management process in which sales, marketing, production, and finance agree on the same number. Below, definitions, examples, and practical points to watch are given for the five critical topics of this process.

1. How Is the Promotion Effect Separated from Baseline Demand?

Historical sales data is cleansed of promotion periods to obtain baseline demand; the additional effect of the promotion (uplift) is forecast separately and added to the plan as a separate layer. Models built with uncleansed data take campaign peaks for "normal demand" and produce excess stock in periods without campaigns.

For example, when a product with weekly baseline demand of 1,000 cases is put on promotion for one week, it can rise to 2,600 cases and then fall to 800 and 850 cases in the following two weeks (figures are for illustration purposes). The average of these three weeks is about 1,400 cases. If the data is given to the model as it is, a forecast of around 1,400 cases is also produced for a week without a campaign; yet the actual baseline demand is 1,000 cases.

• Cannibalization: A promoted product pulls sales from other products of the same brand or category; the category total does not increase as much as the product total. For example, when the 1-liter pack of a brand goes on discount, sales of its 500-milliliter pack may fall.
• Post-campaign dip: Because customers and channels stock up, sales fall below baseline demand after the campaign. The dip is pronounced for products that can be stockpiled at home, such as pasta and detergent, and limited for perishable products.
• Differences by campaign type: Price discounts, multi-buy offers, display, and advertising campaigns have different effects on demand; the type of each campaign should be recorded separately in the campaign library.
• Post-campaign analysis: For each campaign, forecast and actual are compared; the results are collected in a promotion library and become input for forecasting later campaigns.

How the effect can be calculated by campaign type is covered in detail in the article How to Forecast Promotion Impact: Campaign Types and Uplift Calculation.

2. How Are New Products and Portfolio Changes Planned?

For products without history, the launch curves of similar (analogous) products, the target channel distribution, and the listing calendar are used as the basis. For example, for a new flavor added to an existing brand, the first 12-week sales curve of the brand's previously launched varieties can be used as the initial forecast. Two points require attention during the launch period: the calendar of retail chain listings (sales do not start until the product reaches the shelves) and the distribution fill effect (because the quantity first loaded into the channel is higher than consumption, shipments appear above actual demand in the first weeks).

Systematic optimism is common in new product forecasts; for this reason, forecast and actual are monitored closely in the first weeks after launch, and the curve is quickly updated according to actual data.

For products to be phased out of the portfolio, the stock run-down plan is made together with the launch plan of the new product. For example, for a product with renewed packaging, if the run-down of the old pack and the introduction of the new pack are not handled on the same calendar, the old and new products cannibalize each other's sales on the same shelf, and old-pack stock has to be sold at a discount.

As the number of products grows, it is not possible to give the same attention to every product. With ABC-XYZ segmentation, planning effort is directed to high-volume and volatile products: ABC classifies by volume or value, and XYZ by the variability of demand. Products with high volume and high variability become the focus of the planner's effort, while products with low volume and stable demand are largely managed with automatic forecasting.

3. Why Are Seasonality, Special Days, and Shelf Life Critical?

Summer heat, the school period, Ramadan, and religious holidays change demand markedly in fast-moving consumer goods. Ice cream, beverages, and sun care products rise in summer; soup, hot drinks, and cold remedies in winter; confectionery, chocolate, and some food groups before holidays. Such fluctuations are modeled as seasonality; external variables such as air temperature improve forecasts especially for beverages and ice cream.

Because Ramadan and the holidays shift about 11 days earlier every year, seasonality coefficients calculated with monthly periods are misleading; these periods should be modeled as a separate event variable in weekly data. Otherwise, the holiday effect is attributed to a different month every year, and the model systematically mis-forecasts the demand of the months concerned.

For products with short shelf life, safety stock cannot be determined solely by the target service level; the remaining shelf life of the product and the minimum shelf life accepted by the customer also set the upper limit of stock. For example, if a product with a total shelf life of six months must be delivered to the customer with at least four months of remaining shelf life, the product cannot stay in the warehouse for more than two months (figures are for illustration purposes). The first-expired-first-out (FEFO) rule in the warehouse must also be evaluated on the planning side, together with the production batch size; because if the minimum batch size is higher than demand, the surplus turns directly into obsolescence risk.

The general framework for modeling seasonality is covered in the article Time Series Analysis for Demand Forecasting: Stationarity, Trend, and Seasonality.

4. How Do Channel and Distributor Stocks Affect the Forecast?

The sales invoiced by the manufacturer (sell-in) and the sales made to the consumer (sell-out) are not the same thing. Loadings made to distributors and the traditional channel under end-of-period target pressure lead to a drop in orders in the following period and to fluctuations that grow along the supply chain (the bullwhip effect).

For example, when distributors are loaded 30 percent above normal in the last week of the month to reach the sales target (figures are for illustration purposes), sell-in looks high, but distributor stocks swell. In the following month, orders fall. If the manufacturer interprets this drop as a decline in demand and cuts production, shelves can empty a few weeks later even though consumer demand has not changed.

Where possible, sell-out and channel stock data should be included in the forecast; the forecast hierarchy should be built by channel, customer, and product. Since sell-out data is mostly unavailable in the traditional channel, distributor stock reports and sales representative reports are used as proxy data.

5. How Are S&OP and S&OE Linked?

What Is Weekly Execution (S&OE)?

S&OE (Sales and Operations Execution, in short Weekly Execution) is the frequent meeting cycle in which feasible decisions are taken, at the product and location level, for the coming days and weeks, based on the actual situation at hand. S&OP creates the plan; S&OE manages the execution of the approved plan in the field. The meeting is usually weekly, daily in some organizations; its horizon does not exceed a few weeks. Sales operations, demand planning, production planning, logistics, and purchasing units attend.

The kinds of decisions S&OE covers can be seen more clearly with examples:

• Campaign deviation: In the Monday meeting, it is seen that sell-out in the first two days of the campaign that started this week came in 30 percent above plan (figures are for illustration purposes). In the meeting, it is decided to bring the product forward on the production line, to allocate the available stock to campaign customers (allocation), and to partially postpone the orders of other customers.
• Material delay: When the packaging supplier delays delivery by one week, which product's production will be brought forward and which customer will receive a short shipment is determined in this meeting.
• Unexpected change in demand: When a heat wave increases beverage demand, the weekly stock distribution and production priority are updated.
• Line breakdown: If a line stops for two days, it is decided to shift production to other lines or to later days and to reschedule the affected orders.
• Stock approaching expiry: For a batch whose shelf life is getting shorter, quick sales to the channel or a targeted campaign is proposed.

The main distinctions summarizing the difference between S&OP and S&OE are as follows. S&OP meets monthly and makes volume, capacity, inventory, and financial decisions at the product group level over a 12-18 month horizon; the decision is made by management. S&OE meets weekly and makes allocation, priority, and exception decisions at the product and location level over a horizon of a few weeks; the decision is made by the operations team.

How Are S&OP and S&OE Linked to Each Other?

The monthly S&OP cycle sets the framework: approved volume, capacity, inventory targets, and the campaign calendar. S&OE works within this framework; it does not change the framework on its own. Four elements are needed for the link to be established soundly.

• Tolerance band: The deviation band that S&OE can manage with its own authority is defined in writing (for example, ±10% of plan). Deviations within this band are resolved in the S&OE meeting.
• Escalation rule: If the deviation exceeds the band and S&OE's authority (overtime, allocation, priority change) is not enough, the issue is taken to S&OP. Decisions such as capacity increases, additional subcontracted production, or shifting the timing of a campaign are made at the S&OP level.
• Feedback: Recurring deviations in S&OE (for example, the campaign effect being consistently under-forecast) return to the demand assumptions and the campaign library in S&OP.
• One data set and one number: The plan approved in S&OP and the plan used in S&OE must be kept in the same data source; plans run in separate spreadsheets soon diverge from each other.

For example, a campaign in June is placed on the calendar in the March S&OP meeting, and capacity and materials are planned accordingly. In June, the weekly S&OE meeting updates allocation and production priorities according to the actual course of the campaign. If campaign demand deviates markedly from plan and exceeds S&OE's authority, the information is carried to the July S&OP meeting and the plan for subsequent periods is redetermined. The decision logic of the meetings is covered separately in the article Which Decision in Which Meeting in the S&OP Cycle? Calendar, Decision Logic, and Scenarios.

In fast-moving consumer goods, the monthly plan ages quickly with weekly campaign changes;
the link between S&OP and S&OE keeps the plan valid in the field.

Which Metrics Should Be Monitored?

The metrics to be monitored for each of the five critical topics are covered below, together with the relevant planning area.

Promotion effect (Demand Planning, S&OP): Forecast and actual in campaign weeks are compared separately. The weighted mean absolute percentage error (WMAPE) shows how much the forecast is wrong, and bias shows in which direction it is wrong. A bias that is consistently positive in campaigns indicates that the campaign effect is systematically overstated.

New products and portfolio (Demand Planning, Inventory Management): Forecast error and bias in the launch period (for example, the first 8-12 weeks) are monitored separately. The amount of stock remaining after a portfolio exit and the time to run it down show the success of the exit plan.

Seasonality and shelf life (Demand Planning, Production Planning, Inventory Management): The share of stock approaching expiry and destroyed is monitored together with the service level. The shelf-gap rate in peak weeks shows the adequacy of the seasonal forecast and production readiness.

Channel stocks (Demand Planning, S&OP): The difference between sell-in and sell-out and channel stock days (how many days the channel's stock will last at the current sales rate) are monitored. A rise in channel stock days gives advance warning that orders are running ahead of consumption and that orders will fall in the next period.

S&OP-S&OE link (S&OP/IBP, Production Planning): Plan adherence (planned versus actual production and shipments), on-time in-full delivery (OTIF), and the order fulfillment rate are monitored. In addition, the number of issues escalated from S&OE to S&OP shows whether the tolerance band has been set correctly.

Forecast error should be measured at the level and lag at which the decision is made; for example, if the production plan is frozen two weeks ahead, the forecast from two weeks earlier is used. Measurement methods are covered in the article How to Measure and Interpret Forecast Error: Bias and Percentage Error Metrics.

Institutions We Work With in Fast-Moving Consumer Goods

Some of the institutions to which we provide training and/or consulting services in the fast-moving consumer goods sector are listed below. For all sectors, you can visit our References page.

Akgıda, 2022

Alfarm, 2011

Anadolu Efes, 2019, 2021

Banvit, 2010

Barilla Gıda, 2011–2022

Baydöner, 2015

Bereket Döner, 2013

BTA Catering, 2010–2022

Cola Turka, 2016

Crown Bevcan, 2023

Duru Bulgur, 2022

Eker Süt Ürünleri, 2024, 2025

Elita Gıda, 2024, 2025

Essity, 2022

Eti Gıda, 2010

Flormar, 2017

Giz Kozmetik, 2014

Haribo, 2017

Hero Baby, 2015

Imperial Tobacco, 2016

İpek Kağıt, 2010–2015

Kellogg's, 2023

Kılıç Deniz Ürünleri, 2010, 2015

Lezita, 2017

Lidersan, 2016

Lila Kağıt, 2018–2024

Makbul Kuruyemiş, 2021, 2022

Migros, 2021

Milhans Gıda, 2011

Nivea Türkiye, 2019

Penguen Gıda, 2010

Penti, 2014

Rebul Kozmetik, 2021

Red Bull Türkiye, 2023, 2025

Sanat Ambalaj, 2023

Sunar Grup, 2025

Superfresh, 2013

Şölen Çikolata, 2010

Tamek Sagra, 2025

Ulduz Şokolad, 2018

Unifo Gıda, 2010

Wefood, 2026

Yemtar, 2010

“It allows you to look at the demand forecasting process from a higher level. Since examples from different sectors are given, you also get a chance for benchmarking. It explains with examples how you can make accurate and logical forecasts using statistical studies.”

Ender Erdoğmuş, Production Planning Manager, Anadolu Efes (Demand Planning, 2021)

“Thank you for the complete and effective training, thank you again for providing us with the opportunity to notice our areas of improvement.”

Meral Zor, Planning Manager, Akgıda (S&OP, 2022)

Technology and Digitalization: Digital Planning in Fast-Moving Consumer Goods

In fast-moving consumer goods, campaigns, channel stocks, and short shelf life cause the plan to age very quickly. Digitalization allows signals from the field to be reflected in the plan within hours instead of days.

Enterprise Resource Planning (ERP) is the enterprise transaction system that records orders, inventory, purchasing, production, and finance transactions in a common database; it is the master data and transaction source for planning. It is the source of order, shipment, inventory, and price data and forms the basis of the sell‑in based plan.

Electronic Data Interchange (EDI) is the transfer of documents such as orders, call-offs, shipping notices, and invoices between companies in a standard format without human intervention. It enables sell‑out and inventory data to be received regularly and automatically from distributors and retail chains.

Demand Sensing is an approach that quickly reflects current signals such as point-of-sale, order, and external data in the short-term forecast. It updates the short-term forecast daily or weekly using point-of-sale and order data.

Artificial Intelligence (AI) and Machine Learning (ML) are methods that learn patterns from data to produce forecasts, classifications, and recommendations; machine learning is the sub-field of artificial intelligence that learns from data. By modeling promotion, price, weather, and special-day effects together, they separate baseline demand from the campaign effect.

Internet of Things (IoT) is the infrastructure in which sensors on machines, vehicles, warehouses, and products transmit data automatically over a network; location, temperature, status, and consumption data are collected without human input. Connected coolers, vehicle telematics, and cold chain sensors monitor point-of-sale stock and product freshness.

Advanced Planning and Scheduling (APS) is a planning system based on optimization and heuristic methods that produces feasible plans and schedules by considering capacity, material, and priority constraints at the same time. On filling and packaging lines, it matches campaign volumes with capacity; for products with short shelf life, it plans the production sequence.

Supply Chain Control Tower is a monitoring structure that combines order, inventory, production, and shipment data on a single screen to show exceptions and risks early. For S&OE, it monitors inventory, shipment, and campaign exceptions on a single screen.

In fast-moving consumer goods, artificial intelligence comes into play most in short-term forecasting and in separating the promotion effect. For example, demand sensing uses point-of-sale data from the first days of a campaign to update the forecast daily or weekly and presents it to the S&OE meeting as an exception list. Medium-term S&OP decisions, on the other hand, are still decisions made jointly by sales, marketing, and operations.

Linking these two layers, that is, taking the corrections coming from demand sensing to the S&OE meeting, allows the technology to turn into results. Monitoring structures such as the control tower also shorten the meeting by collecting on a single screen the exceptions to be discussed in the S&OE meeting (delayed shipments, insufficient stock during a campaign, batches approaching expiry).

Demand sensing catches the signal, and the S&OE meeting decides accordingly;
without one, the other does not turn into a plan.

Frequently Asked Questions

What is the most common mistake in demand planning in fast-moving consumer goods?

Building a statistical model without separating promotion periods from baseline demand. In this case, campaign peaks are perceived as normal demand, and excess stock builds up in periods without campaigns.

What is S&OE and how does it differ from S&OP?

S&OE is the weekly execution meeting in which allocation, production priority, and exception decisions are taken at the product and location level for the coming days and weeks. S&OP meets monthly and makes volume, capacity, and stocking decisions over a 12-18 month horizon. S&OE works within the plan approved in S&OP and carries deviations that exceed its authority to S&OP.

Should sell-in or sell-out be used in demand forecasting?

The production and shipment plan is made on sell-in; however, in order to see actual consumer demand and channel stocks, sell-out data should be included in the forecast in every channel where possible.

What is demand sensing in fast-moving consumer goods?

It is the rapid reflection of current signals such as point-of-sale, order, and distributor data in the short-term forecast. This forecast, updated daily or weekly with artificial intelligence methods, is used to determine replenishment and production priorities; it does not replace the medium-term S&OP plan.

What services does Optimum Planlama offer to fast-moving consumer goods companies?

It offers organization-specific trainings on demand planning, S&OP/IBP, inventory management, and production planning; and consulting on demand planning, S&OP/IBP, and planning maturity assessment.

References

• Wallace, T. F., & Stahl, R. A. (2008). Sales & Operations Planning: The How-To Handbook (3rd ed.). T. F. Wallace & Company.
• Chopra, S., & Meindl, P. (2019). Supply Chain Management: Strategy, Planning, and Operation (7th ed.). Pearson.
• Blattberg, R. C., & Neslin, S. A. (1990). Sales Promotion: Concepts, Methods, and Strategies. Prentice Hall.
• Gilliland, M. (2010). The Business Forecasting Deal. Wiley.
• Hyndman, R. J., & Athanasopoulos, G. (2021). Forecasting: Principles and Practice (3rd ed.). OTexts.

Optimum Planlama Services for Planning in Fast-Moving Consumer Goods

For planning teams and sales and marketing teams in fast-moving consumer goods to speak the same language, we recommend first running an organization-specific workshop.

Request a quote for training or consulting tailored to your organization.