Preparing for Black Friday with demand forecasting

4 days ago

4 min read

Preparing for Black Friday with Demand Forecasting

Ahead of Black Friday, one of the most important decisions for retailers is how much inventory to prepare for each product, and when that inventory will be needed. Demand forecasting provides a quantitative basis for these decisions. Its business value becomes tangible when patterns identified in historical data are combined with this year’s campaign plans and the market knowledge of commercial teams, then consolidated into a shared demand plan.

The business challenge behind the seasonal peak

During a promotional period, both stockouts and excess inventory at the end of the season represent significant risks. One can lead to lost sales, while the other ties up working capital and may require further markdowns. The task, therefore, is to determine by product and sales channel when, where, and what level of demand the business should prepare for.

In PwC’s 2025 research in Hungary, 44% of online shoppers reported making purchases during Black Friday promotions. According to the study, 30% of Christmas gifts had already been purchased during this period. For retailers, this means that the November campaign and December inventory requirements should be planned in relation to one another. A successful promotion may partly bring forward purchases that would otherwise have taken place later. If planning fails to account for this effect, inventory levels may remain too high after the campaign. Understanding seasonality therefore requires examining both the period before and after the peak.

Why last year’s Black Friday figures are not enough

Consider a retailer that ran a promotion on a popular product for a single weekend last year. This year, the campaign runs for two weeks, uses a different discount level, and includes more stores. Simply scaling up last year’s sales figures does not account for these changes separately. This is an illustrative example, but it clearly shows why the statistical baseline needs to be evaluated against the current commercial plan.

Recurring seasonal patterns, calendar events, and the specific circumstances of a promotion need to be interpreted separately. In forecasting methodology, holidays and special events can also be treated as explanatory factors in their own right. How these are handled in a specific system, whether through a particular model, configuration, or planner adjustment, depends on the solution being used.

Professional judgement is equally important during data cleansing. An outlier may indicate an error, but it may also reflect the genuine impact of a highly successful campaign. Automatically replacing such values without understanding their cause can remove valuable information. It is therefore advisable to retain the original data, verify the reason for any adjustment, and determine whether the event that caused the deviation could occur again.

How forecasting supports seasonal operations

A more robust demand plan primarily improves the quality of business decisions. In inventory planning, it helps identify which products require greater attention and where an incorrect assumption could create significant risk. A high-volume product with uncertain demand may justify a different review frequency from a product with stable sales.

For procurement, the timing of expected demand is also critical. A forecast becomes an executable purchasing plan when it is combined with available inventory, goods already in transit, and supplier lead times. Warehouse and distribution capacity planning can rely on the same aligned demand baseline.

This connection also creates value in financial planning by making it clearer which inventory decisions result in what level of capital commitment. Achieving a positive business outcome still requires the company to act on the forecast. The impact of a better forecast also depends on purchasing rules and execution.

Optasoft’s approach starts from the actual planning process

The AIMMS-based Demand Forecast solution offered by Optasoft provides a statistical foundation for demand planning. Data cleansing suggestions can be reviewed, multiple forecasting methods are available, and results can be analysed at different levels of aggregation. The available methods include Holt–Winters for seasonal patterns and Croston for intermittent demand.

The central business question is how this calculation connects to the company’s own decision-making processes. The planner starts from the data, sales contributes market information, and management looks at the overall picture. The planning process should enable each stakeholder to see the relevant level of detail within the same demand plan, while making the impact of adjustments traceable.

Within the Demand Plan process, the statistical plan can be adjusted based on business information, and changes can be recalculated across the related hierarchy. During implementation, we define the functional scope, access rights, and planning governance that best fit the company. This allows sales expectations to become an integrated part of the shared planning process.

Optasoft’s consulting and development work supports this alignment with business processes. Key questions include the level at which decisions are made, who is authorised to modify the plan, which data they work with, and how the resulting plan is transferred to procurement or production. Application configuration and any required custom development are aligned with these operational needs.

Ultimately, a forecast describes demand that must be purchased or produced, then stored and delivered. This connection gives planning its real business significance. Black Friday is therefore a good opportunity to review whether the different functions within the company are truly working from the same demand plan.