Business Planning and Forecasting

Explore data cleansing to manage level, trend, seasonality, and other components for reliable forecasting, including handling cyclical fluctuations, noise and unexplained variations.

Explore the trend component, including linear and exponential trends, and distinguish seasonal effects using moving averages to improve forecasting and evaluate promotions on historical data.

Seasonality involves repetitive, predictable sales movement around trends driven by weather, festival holidays, and back-to-school, with additive and multiplicative models using seasonal indices to guide forecasts and stock.

Explore cyclical variation and leading indicators to forecast business cycles, apply dynamic regression, and distinguish leading from lagging indicators to anticipate recessions and demand shifts.

Analyze how trading day variations and calendar effects shape sales, inventory, and forecasting by accounting for weekends, holidays, and seasonality across stores.

Use longer data lengths to reveal seasonality, trend, and noise and improve forecast accuracy; with only a few months, forecasts flatten, so aim for six to seven years of data.

Explore how the product hierarchy—from item to prototype, line, class, and family—drives short-term sku movements. Link planning levels to regions, stores, and long-term forecasting to capture patterns and seasonality.

Explore the naive forecast method by relying on the last data point, with no trend or seasonality, and see how one-month lag forecasts mirror actual demand during COVID times.

Explore the exponential smoothing family of models, including 12 variations with level and seasonality. Account for random noise, emphasize recent data, and estimate level, trend, and seasonality.

Explore how smoothing constants alpha, beta, and gamma adapt during a product life cycle, emphasizing level, seasonality, and trend to balance responsiveness and memory.

Explore exponential smoothing concepts, including base level with seasonality and multiplicative seasonality, and the model's pros and cons; learn implementation techniques and how to choose the right model for data.

Explore extensions of exponential smoothing to handle intermittent demand, where zeros occur, using an intelligent model that forecasts non-zero demand and estimates time between demands for better inventory management.

Learn Croston’s intermittent data model through a practical example, deriving average demand and demand frequency to assess inventory needs, safety stock, and the limits of exponential forecasts.

Apply the discrete data model to forecast small unit sales with intermittent demand in B2B settings. Use the negative binomial distribution to estimate counts and confidence with low data volumes.

The custom component model extends long-horizon forecasting by enabling adjustable starting points, trends, and seasonality, and accounts for future events, as shown with telecom data and promotions.

Explore the event model to isolate and quantify the impact of promotions on demand, remove event effects from forecasts, and improve planning for promotions and seasonality.

Explore ARIMA, the auto regressive integrated moving average model, a forecasting tool. Understand its components and advantages, and how it differs from exponential moving models, with computations handled by computers.

Explore arima, a time series model that combines auto regressive, integrated, and moving average components using its own history for forecasting, with Box-Jenkins rules on white noise and stationary.

Explore the components of ARIMA, including auto regressive and integrated moving average, and how past data, forecast errors, and differencing yield reliable sales forecasts.

Learn to read ARIMA models, distinguish non seasonal and seasonal components, and interpret autoregressive, integrated, and moving-average parts, including seasonal indicators for forecasting.

ARIMA provides strong forecast accuracy for stable time series and competes with exponential smoothing, but cannot easily incorporate external parameters and needs longer data histories.

Compare exponential smoothing and ARIMA, understand level, trend, and seasonality components, and learn to use both as forecasting tools for diverse sales data.

Choose regression or ARIMA based on data needs; ARIMA is a univariate model that uses history and cannot exploit leading indicators, while regression is multivariate and can incorporate external variables.

Explore ARIMA examples and computations, illustrating 0-0-0, 1-0-0, and 0-1-0 models, and explain how constants, alpha, and beta drive July–August forecasts.

Learn how an automated forecasting solution selects ARIMA and exponential smoothing models, detects seasonality, and optimizes forecast errors for historical data.

We establish a baseline or statistical forecast from data, add inputs, and form the final plan; then measure deviations, analyze causes, and iterate to improve error and reduce risk.

In business planning and forecasting, assess model fit by comparing actual versus fitted data using r-squared and adjusted r-squared. Higher values show stronger relationships; near zero indicates weak fit.

Identify and monitor forecast bias over time by comparing forecast to actuals, assess strength with tracking signal and absolute error, and adjust sales and supply chain planning.

Learn how mean absolute deviation measures forecast accuracy by averaging the absolute differences between actual and forecast, a unit-based method for business planning and forecasting.

learn how mean absolute percentage error measures forecast accuracy by using the actual as the denominator, calculating the absolute percentage error, and averaging these errors for high-volume data.

Explore how mean absolute deviation (MAD) and mean absolute percentage error (MAPE) compare forecast accuracy across products, addressing unit differences, averaging pitfalls, and error aggregation.

Learn to quantify value add in forecasting by comparing baseline forecasts to system overrides, measure accuracy, track who adds value, and use unit-consistent metrics to drive process improvements.

Classify forecasting errors into buckets, analyze underlying causes, and focus on high-value items to improve forecasts. Use dashboards to visualize bias, lag, and value added for ongoing refinement.

Explore new product forecasting using statistical techniques, an algorithm, and machine learning to improve planning and efficiency in product introductions.

Explore why firms introduce new products amid fierce competition and evolving customer needs, and examine forecasting challenges, from lack of sales data to unpredictable adoption and life cycles.

Classify new products into six major categories—new to the world, new to the firm, repositioning, additions to existing product lines, improvements and revisions to existing products, and cost reductions.

Explore new-to-the-world products that create new markets and lack direct comparisons. Learn how firms research market acceptance and the challenges of forecasting accuracy for innovative launches.

Explore new to the firm products, expanding offerings via acquisitions or in-house development, while planning demand with surveys, SWOT analysis, and Porter five forces amid imitation risk.

Explore product repositioning to target new markets and uses, with examples like baking soda and a spring as a blood thinner, and the need for market studies before forecasting.

Examine line extensions that add to existing product lines to grow business and market share, using Toyota’s Lexus launch and Porsche, Unilever, and fmcg examples driven by customer feedback.

Examine how firms revise products to stay ahead of competition through rapid r&d and new features, such as rfid, with examples like iPhone 10, Ivory soap, and Tide powder.

Identify cost-reduction strategies across the six categories of new product classifications by replacing existing products with lower-cost designs or production that maintain similar performance.

Explore forecasting for six product types, using analogy for variations and base diffusion for new-to-world products to estimate market size, adoption, and life cycle.

Forecasting by analogy estimates demand for a new product by relating it to a similar item, mirroring its trend, seasonality, and level. Teams collaborate to deploy an easy approach.

Identify a similar product based on its characteristics, select its historical data, align seasonality with the launch date, and specify the launch total using sales and marketing input.

Pick a similar product from many options, as in handbags, by analyzing customer behavior with inputs from the sales team and existing product data to forecast the new product's behavior.

Use by analogy to select the forecast period, leveraging the history of a similar product and favoring recent data when consumer behavior shifts in technology, while considering similar product forecasts.

Align seasonality in data by matching monthly peaks and troughs to improve forecasting for new product launches, ensuring the launch month aligns with similar products.

Specify the launch total by merging business data with maths, use surveys to gauge market acceptance, then distribute the six-month forecast based on a similar product's sales trend and seasonality.

Apply forecast techniques using a practice sheet with criteria, data from a similar product, and launch scenarios, aided by the cloud-based tool Plan our mind to compute forecasts.

Use k-means clustering to find similar products for forecasting by analogy, selecting attributes, scaling, weighting, and using the elbow method with centroids.

Learn the four steps of forecasting by analogy and use Kane's clustering method to narrow SKU options and select a similar product.

Identify product attributes, scale them, assign attribute weights, determine the number of homogeneous clusters, and initialize centroids before applying the k means clustering algorithm.

Use k means clustering to group similar products by chosen attributes, forming clusters via Euclidean distance. Identify the natural group for a new product to simplify forecasting by analogy.

Identify product attributes that influence sales decisions, such as color, size, price, and category. Consider geography and culture to form customer clusters, using input from sales teams.

Learn how to convert categorical values to numeric for k-means clustering, map similar colors to nearby numbers, and standardize attributes by subtracting the mean and dividing by the standard deviation.

Assign weights to attributes in clustering to prioritize category, price, size, and color according to business requirements and customer preferences, enabling category-based grouping of products like handbags.

Identify the optimal number of clusters in k-means with the elbow method by examining the sum of squared errors across k values and choosing the elbow point.

Initialize two randomly assigned centroids for five clusters in k means clustering, assign observations by minimum Euclidean distance to centroids, and iteratively refine centroids.

Learn how Euclidean distance calculates the straight-line separation between observations and cluster centroids in k means, guiding minimum-distance cluster assignments in Cartesian coordinates.

Learn k-means clustering by initializing centroids, iteratively minimizing the sum of squared distances to items, forming three clusters, and grouping similar product sets while removing outliers.

Use k-means clustering to group similar products and pick the best options by Euclidean distance to the new item, then monitor forecast accuracy.

Forecast sales for new to the world products using the diffusion-based base model, which captures first purchase and adoption by innovators and imitators across a lifecycle curve with three parameters.

Explain the Bass diffusion model parameters, innovation P and imitation Q, and how they forecast adoption using the remaining potential market M, current adopters, and time D.

Identify the Bass diffusion model's core assumptions—binary adoption, a constant maximum potential number of buyers, no repeat purchases, independent word-of-mouth, and no substitutes—and their impact on forecasting new-to-world products.

Explore forecasting strategies for new product launches, focusing on cannibalization, transition periods, and current inventory to plan sales and manage stock.

Analyze cannibalization scenarios and manage sales plans across drink sizes. 200 ml and 400 ml variants drive 80 percent of sales; assess the impact of a new 250 ml option.

Learn how a new 250-milliliter product cannibalizes 60 percent of the 200-milliliter sales, reallocating to the new item while keeping overall market sales unchanged.

examine cannibalization as a 300 ml drink shifts 2500 units from 200 ml and 1000 units from 400 ml, totaling 3500 units, with 100 ml unchanged and no category growth.

Analyze cannibalization when launching 450 ml and 500 ml drinks, reallocating 50% of 400 ml sales and capturing additional sales from a new market; observe revised product contributions.

Explore transition period scenarios and plan sales by accounting for the transition during new product launches, enhancing forecasting and sales strategy.

Analyze a one-month transition where a new 250 millilitres drink cannibalizes the 200 millilitres beverage, transferring 5,000 monthly units to the new size and achieving 100 percent contribution afterward.

Explain the four-month transition where a new 250 ml product cannibalizes 60% of the 200 ml sales at 15% per month, illustrating cannibalization effects and monthly sales shifts.

Analyze cannibalization during a four-month transition where a 250 ml product shifts sales from 200 ml and 300 ml, with monthly cannibalization rates.

Explore a four-month transition where a new 250 ml product cannibalizes 100% of 200 ml and 33% of 400 ml sales, guiding demand planning and monthly cannibalization rates.

Explore current inventory scenarios and how stock levels impact cash flow. Guide liquidation or relocation to promising sales locations for phased-out items.

Plan phase-out and phase-in by analyzing cannibalization, forecasting, and existing stock; shift from 200 ml to 250 ml with 5000 unit inventory, exhausting in one month.

Adjust forecasts and launch plans when on-hand inventory exceeds the monthly forecast, as shown by phasing out 200 ml for 250 ml with 50% cannibalization.

Analyze how to plan sales when current inventory exceeds monthly forecasts during a one-month transition from 200 ml to 250 ml drinks, accounting for cannibalization and seasonal forecasts.

Plan sales and adjust forecasts across a four-month transition for phasing out a 200-milliliter item and launching 300-milliliter offerings, accounting for 33 percent cannibalization and differing monthly inventories.

Adjust the 200 milliliters forecast as it is phased out and transferred to 300 milliliters, then allocate the five-month demand while managing opening inventory and contributions by product.

Analyze how a new 400 ml drink cannibalizes 33.3% of sales and drives monthly forecast shifts, allocating 8.3% per month to 300 ml and 400 ml as the total stabilizes.

Explore how introducing a 300 ml item reshapes cannibalization across 200 ml and 400 ml products, with month-by-month forecasts and evolving contribution shares.

Explore real-life business scenarios that arise when launching a new product and learn how to forecast demand and plan effectively with strategic approaches.

Forecast demand for one-to-many sku scenarios by mapping old sku sales to new components, such as phone, charger, and earphones, and allocate sales proportionally to each new sku.

Examine many-to-one sku forecasting where one product replaces multiple color variants, then forecast each variant from historical data and aggregate to the new model.

Learn how to forecast a merged sku by valuing bundled products, allocating costs from the dominant item's historic sales, and accounting for price sensitivity and cannibalization.

Map old Eskew codes to a stable Eskew name to preserve historical trends and enable forecast continuity when data is short during phase-in and phase-out.

Explain how new products launch across geographies using phased pilots to detect patterns and plan demand, while accounting for regional differences and proactive forecasting.

Explore the components of inventory management, including the key supply barometers, safety stock calculations, stocking strategies, and supply data management, plus item classification and replenishment to improve inventory efficiency.

Examine the inventory components, including pipeline stock during lead time, safety stock, and cyclical stock, and how reorder level drives replenishment cycles.

Examine supply parameters shaping decisions, from display stock, seasonal stock, safety stock, and lead time variance. Balance min and max order quantities, shelf life and costs to maximize sales.

Learn how safety stock serves as a buffer to protect service levels amid demand fluctuations and forecast errors, balancing supply, cash flow, and inventory risks.

Balance safety stock as a buffer to minimize inventory costs while maximizing customer service level during uncertain times.

Calculate safety stock to balance demand variability and supplier uncertainty by applying standard deviation, risk factors, and service level to lead time.

Explore a practical safety stock calculation that links demand, standard deviation, and lead time to service level targets. Learn how supplier performance and forecast uncertainty shape inventory levels and kpis.

Analyze lead time and usage variability to balance safety stock and service levels, considering supplier variability, average daily usage, and demand variability.

Explore how forecast accuracy and lead time shape inventory coverage, safety stock, and stockout risk, with practical coverage calculations and optimization tips.

Explain how demand variability drives safety stock decisions, using static safety stock for relatively stable demand and dynamic safety stock with two levels for periods of high fluctuation.

Explore stocking strategies for balancing demand with inventory through monthly versus dynamic weekly replenishment, including safety stock, reorder points, and coverage planning to optimize service levels and costs.

Categorize products into four types and apply selective inventory strategies to manage item counts, protect cash flow, and focus on high-value products.

Explore item classification strategies for inventory planning, including ABC analysis and unit-cost based approaches. Assess movement, criticality, and supply sources to safeguard cash flow and service levels.

Improve inventory and fulfillment by tightening forecasts, raising safety stock above 25 percent, and reducing lead time variability. Rationalize SKUs, eliminate obsolete stock, and balance procurement, transportation, and inventory costs.

Explore replenishment calculation by linking stock on hand, safety stock, and lead time to set order points, forecast demand, and optimize replenishment strategies for inventory and service level.