Demand planning goes beyond simply forecasting product needs; it’s about ensuring your business meets customer demands with precision, efficiency, and cost-effectiveness. Latest demand planning technology addresses key challenges like forecast accuracy, inventory management, and market responsiveness. In this blog, we will introduce critical demand planning trends, including data-driven insights, probabilistic forecasting, consensus planning, predictive analytics, scenario modeling, real-time visibility, and multilevel forecasting. These trends will help you stay ahead of the curve, optimize your supply chain, reduce costs, and enhance customer satisfaction, positioning your business for long-term success.

Data-Driven Insights

Advanced analytics, machine learning, and artificial intelligence (AI) are becoming integral to demand planning. Technologies like Smart UP&O allow businesses to analyze complex data sets, identify patterns, and make more accurate predictions. This shift towards data-driven insights helps businesses respond quickly to market changes, minimizing stockouts and reducing excess inventory.

Probabilistic Forecasting

Probabilistic forecasting focuses on predicting a range of possible outcomes rather than a single figure. This trend is particularly important for managing uncertainty and risk in demand planning. It helps businesses prepare for various demand scenarios, enhancing inventory management and reducing the likelihood of stockouts or overstocking​.

Consensus Forecasting

Modern manufacturing is moving towards an integrated approach where departments and stakeholders collaborate more closely. Collaborative forecasting involves sharing insights across the supply chain, from suppliers to distributors and internal teams. This approach breaks down silos and ensures that everyone is working towards a common goal, leading to a more synchronized and efficient supply chain​.

Predictive and Prescriptive Analytics

Predictive analytics forecasts future outcomes based on historical data and trends, helping businesses anticipate demand fluctuations. For example, Smart Demand Planner (SDP) automates forecasting to adjust inventory and production levels accordingly​.

Prescriptive analytics goes further by offering actionable recommendations. Smart Inventory Planning and Optimization (IP&O), for instance, prescribes optimal inventory policies based on service levels, costs, and risks​. ogether, these tools enable proactive decision-making, allowing companies to both predict and optimize their responses to future challenges.

Scenario Modeling

Scenario modeling is becoming a key part of demand planning, enabling businesses to simulate different scenarios and assess their impact on operations. This method helps companies create adaptable strategies to effectively handle uncertainties. Smart IP&O enhances this capability by offering What If Scenarios that allow users to test different inventory policies before implementation. By adjusting variables like service levels or order quantities, businesses can visualize the effects on costs and service levels, empowering them to select the optimal strategy for minimizing risks and controlling costs​​.

Real-Time Visibility

As supply chains become more global and interconnected, real-time visibility into inventory and supply chain activities is crucial. Enhanced collaboration with suppliers and distributors, combined with real-time data, enables businesses to make quicker, more informed decisions. This helps optimize inventory levels, reduce lead times, and improve overall supply chain resilience​.

Multilevel Forecasting

This involves forecasting at different levels of the product hierarchy, such as individual items, product families, or even entire product lines. Multilevel forecasting is vital for businesses with complex product portfolios, as it ensures that forecasts are accurate at both the micro and macro levels​.

Demand planning is a decisive aspect of modern supply chain management, offering businesses the ability to enhance operational efficiency, reduce costs, and better meet customer demands. Leveraging advanced platforms like Smart IP&O significantly improves forecasting accuracy and inventory management, enabling swift responses to market fluctuations. Automated statistical forecasting, combined with capabilities like hierarchy forecasting and forecast overrides, ensures that forecasts are accurate and adaptable, leading to more realistic planning decisions. Additionally, with tools like scenario modeling, businesses can explore various demand scenarios across their product hierarchy, facilitating informed decision-making by providing insights into potential outcomes and risks. This approach allows businesses to anticipate the impact of policy changes, make better decisions, and ultimately optimize their inventory and overall supply chain management, staying ahead of key trends in the process.

In this blog, we will analyze the automatic forecasting for time series demand projections, focusing on key techniques, challenges, and best practices. There are multiple methods to predict future demand for an item, and this becomes complex when dealing with thousands of items, each requiring a different forecasting technique due to their unique demand patterns. Some items have stable demand, others trend upwards or downwards, and some exhibit seasonality. Selecting the right method for each item can be overwhelming. Here, we’ll explore how automatic forecasting simplifies this process.

Automatic forecasting becomes fundamental in managing large-scale demand projections. With thousands of items, manually selecting a forecasting method for each is impractical. Automatic forecasting uses software to make these decisions, ensuring accuracy and efficiency in the forecasting process. It’s importance lies in its ability to handle complex, large-scale forecasting needs efficiently. It eliminates the need for manual selection, saving time and reducing errors. This approach is particularly beneficial in environments with diverse demand patterns, where each item may require a different forecasting method.

Key Considerations for Effective Forecasting

1. Challenges of Manual Forecasting:
   • Infeasibility: Manually choosing forecasting methods for thousands of items is unmanageable.
   • Inconsistency: Human error can lead to inconsistent and inaccurate forecasts.
2. Criteria for Method Selection:
   • Error Measurement: The primary criterion for selecting a forecasting method is the typical forecast error, defined as the difference between predicted and actual values. This error is averaged over the forecast horizon (e.g., monthly forecasts over a year).
   • Holdout Analysis: This technique simulates the process of waiting for a year to elapse by hiding some historical data, making forecasts, and then revealing the hidden data to compute errors. This helps in choosing the best method in real-time.
3. Forecasting Tournament:
   • Method Comparison: Different methods compete to forecast each item, with the method producing the lowest average error winning.
   • Parameter Tuning: Each method is tested with various parameters to find the optimal settings. For example, simple exponential smoothing may be tried with different weighting factors.

The Algorithms Behind Effective Automatic Forecasting

Automatic forecasting is highly computational but feasible with modern technology. The process involves:

• Data Segmentation: Dividing historical data into segments helps manage and leverage different aspects of historical data for more accurate forecasting. For instance, for a product with seasonal demand, data might be segmented by seasons to capture season-specific trends and patterns. This segmentation allows forecasters to make and test forecasts more effectively.
• Repeated Simulations: Using sliding simulations involves repeatedly testing and refining forecasts over different periods. This method validates the accuracy of forecasting methods by applying them to different segments of data. An example is the sliding window method, where a fixed-size window moves across the time series data, generating forecasts for each position to evaluate performance.
• Parameter Optimization: Parameter optimization involves trying multiple variants of each forecasting method to find the best-performing one. By adjusting parameters, such as the smoothing factor in exponential smoothing methods or the number of past observations in ARIMA models, forecasters can fine-tune models to improve performance.

For instance, in our software, we allow various forecasting methods to compete for the best performance on a given item.  Knowledge of Automatic forecasting immediately carries over to Simple Moving Average, linear moving average, Single Exponential Smoothing, Double Exponential Smoothing, Winters’ Exponential Smoothing, and Promo forecasting. This competition ensures that the most suitable method is selected based on empirical evidence, not subjective judgment. The tournament winner is the closest method to predicting new data values from old. Accuracy is measured by average absolute error (that is, the average error, ignoring any minus signs). The average is computed over a set of forecasts, each using a portion of the data, in a process known as sliding simulation, which we have explained previously in a previous blog.

Methods used in Automatic forecasting

Normally, there are six extrapolative forecasting methods competing in the Automatic forecasting tournament:

• Simple moving average
• Linear moving average
• Single exponential smoothing
• Double exponential smoothing
• Additive version of Winters’ exponential smoothing
• Multiplicative version of Winters’ exponential smoothing

The latter two methods are appropriate for seasonal series; however, they are automatically excluded from the tournament if there are fewer than two full seasonal cycles of data (for example, fewer than 24 periods of monthly data or eight periods of quarterly data). These six classical, smoothing-based methods have proven themselves to be easy to understand, easy to compute and accurate. You can exclude any of these methods from the tournament if you have a preference for some of the competitors and not others.

Automatic forecasting for time series data is essential for managing large-scale demand projections efficiently and accurately. Businesses can achieve better forecast accuracy and streamline their planning processes by automating the selection of forecasting methods and utilizing techniques like holdout analysis and forecasting tournaments. Embracing these advanced forecasting techniques ensures that businesses stay ahead in dynamic market environments, making informed decisions based on reliable data projections.

Companies that depend on spreadsheets for demand planning, forecasting, and inventory management are often constrained by the spreadsheet’s inherent limitations. This post examines the drawbacks of traditional inventory management approaches caused by spreadsheets and their associated costs, contrasting these with the significant benefits gained from embracing state-of-the-art planning technologies.

Spreadsheets, while flexible for their infinite customizability, are fundamentally manual in nature requiring significant data management, human input, and oversight. This increases the risk of errors, from simple data entry mistakes to complex formula errors, that cause cascading effects that adversely impact forecasts.  Additionally, despite advances in collaborative features that enable multiple users to interact with a common sheet, spreadsheet-based processes are often siloed. The holder of the spreadsheet holds the data.  When this happens, many sources of data truth begin to emerge.  Without the trust of an agreed-upon, pristine, and automatically updated source of data, organizations don’t have the necessary foundation from which predictive modeling, forecasting, and analytics can be built.

In contrast, advanced planning systems like Smart IP&O are designed to overcome these limitations. Such systems are built to automatically ingest data via API or files from ERP and EAM systems, transform that data using built in ETL tools, and can process large volumes of data efficiently.  This enables businesses to manage complex inventory and forecasting tasks with greater accuracy and less manual effort because the data collection, aggregation, and transformation is already done. Transitioning to advanced planning systems is key for optimizing resources for several reasons.

Spreadsheets also have a scaling problem. The bigger the business grows, the greater the number of spreadsheets, workbooks, and formulas becomes.  The result is a tightly wound and rigid set of interdependencies that become unwieldy and inefficient.  Users will struggle to handle the increased load and complexity with slow processing times and an inability to manage large datasets and face challenges collaborating across teams and departments.

On the other hand, advanced planning systems for inventory optimization, demand planning, and inventory management are scalable, designed to grow with the business and adapt to its changing needs. This scalability ensures that companies can continue to manage their inventory and forecasting effectively, regardless of the size or complexity of their operations. By transitioning to systems like Smart IP&O, companies can not only improve the accuracy of their inventory management and forecasting but also gain a competitive edge in the market by being more responsive to changes in demand and more efficient in their operations.

Benefits of Jumping in: An electric utility company struggled to maintain service parts availability without overstocking for over 250,000-part numbers across a diverse network of power generation and distribution facilities. It replaced their twenty-year-old legacy planning process that made heavy use of spreadsheets with Smart IP&O and a real-time integration to their EAM system.  Before Smart, they were only able to modify Min/Max and Safety Stock levels infrequently.  When they did, it was nearly always because a problem occurred that triggered the review.  The methods used to change the stocking parameters relied heavily on gut feel and averages of the historical usage.   The Utility leveraged Smart’s what-if scenarios to create digital twins of alternate stocking policies and simulated how each scenario would perform across key performance indicators such as inventory value, service levels, fill rates, and shortage costs.  The software pinpointed targeted Min/Max increases and decreases that were deployed to their EAM system, driving optimal replenishments of their spare parts.  The result:  A significant inventory reduction of $9 million that freed up cash and valuable warehouse space while sustaining 99%+ target service levels.

Managing Forecast Accuracy: Forecast error is an inevitable part of inventory management, but most businesses don’t track it.  As Peter Drucker said, “You can’t improve what you don’t measure.”  A global high-tech manufacturing company utilizing a spreadsheet-based forecast process had to manually create its baseline forecasts and forecast accuracy reporting.  Given the planners’ workload and siloed processes, they just didn’t update their reports very often, and when they did, the results had to be manually distributed.  The business didn’t have a way of knowing just how accurate a given forecast was and couldn’t cite their actual errors by group of part with any confidence.  They also didn’t know whether their forecasts were outperforming a control method.  After Smart IP&O went live, the Demand Planning module automated this for them. Smart Demand Planner now automatically reforecasts their demand each planning cycle utilizing ML methods and saves accuracy reports for every part x location.  Any overrides that are applied to the forecasts can now be auto-compared to the baseline to measure forecast value add – i.e., whether the additional effort to make those changes improved the accuracy.  Now that the ability to automate the baseline statistical forecasting and produce accuracy reports is in place, this business has solid footing from which to improve their forecast process and resulting forecast accuracy.

Get it Right and Keep it Right:  Another customer in the aftermarket parts business has used Smart’s forecasting solutions since 2005 – nearly 20 years!  They were faced with challenges forecasting intermittently demanded parts sold to support their auto aftermarket business. By replacing their spreadsheet-based approach and manual uploads to SAP with statistical forecasts of demand and safety stock from SmartForecasts, they were able to significantly reduce backorders and lost sales, with fill rates improving from 93% to 96% within just three months.  The key to their success was leveraging Smart’s patented method for forecasting intermittent demand – The “Smart-Willemain” bootstrap method generated accurate estimates of the cumulative demand over the lead time that helped ensure better visibility of the possible demands.

Connecting Forecasts to the Inventory Plan: Advanced planning systems support forecast-based inventory management, which is a proactive approach that relies on demand forecasts and simulations to predict possible outcomes and their associated probabilities.  This data is used to determine optimal inventory levels.  Scenario-based or probabilistic forecasting contrasts with the more reactive nature of spreadsheet-based methods. A longtime customer in the fabric business, previously dealt with overstocks and stockouts due to intermittent demand for thousands of SKUs. They had no way of knowing what their stock-out risks were and so couldn’t proactively modify policies to mitigate risk other than making very rough-cut assumptions that tended to overstock grossly.  They adopted Smart Software’s demand and inventory planning software to generate simulations of demand that identified optimal Minimum On-Hand values and order quantities, maintaining product availability for immediate shipping, highlighting the advantages of a forecast-based inventory management approach.

Better Collaboration:  Sharing forecasts with key suppliers helps to ensure supply.  Kratos Space, part of Kratos Defense & Security Solutions, Inc., leveraged Smart forecasts to provide their Contract Manufacturers with better insights on future demand.  They used the forecasts to make commitments on future buys that enabled the CM to reduce material costs and lead times for engineered-to-order systems. This collaboration demonstrates how advanced forecasting techniques can lead to significant supply chain collaboration that yields efficiencies and cost savings for both parties.

​

For some of our customers, weather has a significant influence on demand. Extreme short-term weather events like fires, droughts, hot spells, and so forth can have a significant near-term influence on demand.

There are two ways to factor weather into a demand forecast: indirectly and directly. The indirect route is easier using the scenario-based approach of Smart Demand Planner. The direct approach requires a tailored special project requiring additional data and hand-crafted modeling.

Indirect Accounting for Weather

The standard model built into Smart Demand Planner (SDP) accommodates weather effects in four ways:

1. If the world is steadily getting warmer/colder/drier/wetter in ways that impact your sales, SDP detects these trends automatically and incorporates them into the demand scenarios it generates.
2. If your business has a regular rhythm in which certain days of the week or certain months of the year have consistently higher or lower-than-average demand, SDP also automatically detects this seasonality and incorporates it into its demand scenarios.
3. Often it is the cussed randomness of weather that interferes with forecast accuracy. We often refer to this effect as “noise”. Noise is a catch-all term that incorporates all kinds of random trouble. Besides weather, a geopolitical flareup, the surprise failure of a regional bank, or a ship getting stuck in the Suez Canal can and have added surprises to product demand. SDP assesses the volatility of demand and reproduces it in its demand scenarios.
4. Management overrides. Most of the time, customers let SDP churn away to automatically generate tens of thousands of demand scenarios. But if users feel the need to touch up specific forecasts using their insider knowledge, SDP can make that happen through management overrides.

Direct Accounting for Weather

Sometimes a user will be able to articulate subject matter expertise linking factors outside their company (such as interest rates or raw materials costs or technology trends) to their own aggregate sales. In these situations, Smart Software can arrange for one-off special projects that provide alternative (“causal”) models to supplement our standard statistical forecasting models. Contact your Smart Software representative to discuss a possible causal modeling project.

Meanwhile, don’t forget your umbrella.

For too many companies, a critical piece of data fact-finding ― the measurement of demand uncertainty ― is handled by simple but inaccurate rules of thumb.  For example, demand planners will often compute safety stock by a user-defined multiple of the forecast or historical average.  Or they may configure their ERP to order more when on hand inventory gets to 2 x the average demand over the lead time for important items and 1.5 x for less important ones. This is a huge mistake with costly consequences.

The choice of multiple ends up being a guessing game.  This is because no human being can compute exactly how much inventory to stock considering all the uncertainties.  Multiples of the average lead time demand are simple to use but you can never know whether the multiple used is too large or too small until it is too late.  And once you know, all the information has changed, so you must guess again and then wait and see how the latest guess turns out.  With each new day, you have new demand, new details on lead times, and the costs may have changed.  Yesterday’s guess, no more matter how educated is no longer relevant today.  Proper inventory planning should be void of inventory and forecast guesswork.  Decisions must be made with incomplete information but guessing is not the way to go.

Knowing how much to buffer requires a fact-based statistical analysis that can accurately answer questions such as:

• How much extra stock is needed to improve service levels by 5%
• What the hit to on-time delivery will be if inventory is reduced by 5%
• What service level target is most profitable.
• How will the stockout risk be impacted by the random lead times we face.

Intuition can’t answer these questions, doesn’t scale across thousands of parts, and is often wrong.  Data, probability math and modern software are much more effective. Winging it is not the path to sustained excellence.