Get a consultation

Blog

  • All
  • Expertise
  • Success stories
  • Releases
December 4th, 2024 12 MIN READ
Non-obvious strategies for profit growth: the impact of AI demand forecasting on a company’s P&L
Modern businesses deal with huge amounts of data that have complex and often non-obvious relationships between different industries and areas of the company’s activities. Therefore, traditional analytical tools are becoming less and less effective. Without the latest approaches to analysis and forecasting, it is difficult to understand the underlying factors that affect profitability and where the causes of potential losses lie. Solutions using AI-based algorithms come to the rescue, more accurately recognizing hidden correlations and allowing a better understanding of the connections between large data flows. Such solutions automate routine tasks and give businesses a new level of operational efficiency. AI algorithms free up resources for strategic planning, which has a positive impact on management decisions and expenses. As a result, the company receives improved sales profitability, supply chain optimization, reduced storage costs, workforce planning, etc., as well as the ability to quickly respond to market changes. All this ultimately affects the P&L indicators. Therefore, we suggest sorting out how exactly accurate AI forecasting can optimize costs and increase the profitability of your business.

Why is forecasting accuracy a strategic challenge not only for Supply Chain but also for other parts of your business?

Consulting company Grand View Research predicts that the global AI demand forecasting market size will grow by 10.3% every year. Forecasting is usually associated with Supply Chain, as accurate forecasts require a clear understanding of future needs for optimal inventory, production, and delivery management. However, this key function has much broader business implications and extends beyond the supply chain, impacting the company’s P&L. Forecasting accuracy is essential for finance departments, as it allows for better control and planning of expenses and revenues, as well as minimizing risks associated with market volatility:
  1. Budgeting and Spending Planning: Accurate demand and sales forecasts provide the finance department with clear guidelines for more informed planning. Knowing which products or services will be in greatest demand and in what volumes allows finance specialists to more effectively allocate resources between departments and projects, adapting the budget to actual needs. This helps avoid unnecessary spending or insufficient funds to support priority areas. By aligning forecasts with the company’s budget decisions, the finance department gets a clear picture of the business’ needs, which helps to plan expenses rationally.
  2. Capital Investments: Accurate demand forecasts give the business an understanding of when additional resources may be needed to meet growing demand. This allows it to more effectively plan investments in new equipment, capacity expansions, or other capital projects. This way, the business has the data to avoid the risks of underinvestment or overspending.
  3. Cash Flow Stability: The cash flow of a business depends on the ability to anticipate seasonal fluctuations in demand or changes in supply costs. When forecasts are accurate, the finance department can plan payments and investments to avoid cash flow gaps and maintain cash flow stability.
  4. Minimizing financial risks: Accuracy in forecasts helps reduce risks that arise from sudden changes in the market, such as a shortage of raw materials or a sudden drop in demand. The finance department, thanks to forecasts, can create reserves or plan alternative sources of income in case of changes. For example, according to Pro Consulting, due to the crisis in Sri Lanka and complicated logistics as a result of russia’s full-scale invasion of Ukraine in 2022, prices for unpackaged tea leaves have increased significantly – by more than 40%. Using modern AI solutions for forecasting, tea packaging companies in Ukraine can predict the growth or decline in demand in advance, as well as select optimal suppliers of raw materials. This helps make strategic decisions on creating reserves, reorienting imports to more stable markets, or concluding long-term contracts with suppliers at fixed prices. In addition, with the help of accurate forecasts, a company can develop strategies for diversifying sources or finding an alternative. For example, replacing some of the raw materials with local suppliers, which will reduce the risks of supply chain disruptions. For example, domestic companies can quickly introduce new positions of herbal, phyto and medicinal teas grown in Ukraine. Forecasting will help meet the trendy consumer demand for a healthy lifestyle or strengthening the immune system in the pre-winter period. This example shows the logical connection between demand forecasting, the market’s dependence on imports, crisis conditions and the ability of companies to adapt their business processes.
  5. Impact on profit and loss (P&L): Accurate forecasting of demand and expenses helps to avoid excess production and reduces the cost of storing unsold goods. This has a positive effect on profitability indicators, reduces operating costs and ensures stable profits.
In addition to financial departments, the marketing department can significantly improve the development of its promotional strategies. For example, the SMART Demand Forecast solution has a scenario forecasting functionality that allows businesses to assess how demand will change under different conditions. For example, the solution helps to model what will happen if you reduce the price, launch or change promotion, or invest more in marketing activities. After all, without additional stimulation of demand, it will be possible to fulfill the sales plan by only 70%. This approach helps the company choose the most profitable scenario that best meets business goals. Based on this, the company can make decisions on the feasibility of additional expenses or investments and adapt its resources and strategies in advance. In addition, demand forecasting affects the work of cross-functional teams, because the results of such calculations affect human resource planning. Thanks to accurate forecasts, the HR department can plan the recruitment and attraction of employees for periods of expected demand growth. This allows you to avoid both a personnel shortage and an excess of human resources, which has a positive effect on work efficiency and cost optimization. And if you are still convinced that forecasting only concerns the Supply Chain and bypasses other departments of your company, we suggest you visually understand how forecasting accuracy has a direct impact on P&L indicators. In the table below, we analyzed the key components of income and expenses, in particular: revenue, cost of goods, operating expenses, marketing and administrative expenses, and taxes. Each item contains both planned and actual indicators, supplemented by comments on the reasons for deviations. The last column shows the share of influence of accurate forecasting on these indicators, which could ensure a more accurate match between forecasts and actual results.
P&L COMPONENTSPLANNED INDICATORSCOMMENTS ON PLANNED INDICATORSACTUAL INDICATORSCOMMENTS ON ACTUAL INDICATORSFORECASTING ACCURACY IMPACT
Income1,100,000.00Profit was planned based on standard monthly demand without taking into account major market changes or additional promotions.1,200,000.00An unplanned promotional campaign lowered product prices, which resulted in a temporary increase in revenue, but this increase did not offset the additional expenses.100% impact
Other income 50,000.00A small income from the sale of services or goods of another category that is not the core business of the company was planned. 55,000.00"A slight increase in additional income from accessory sales due to increased customer traffic due to the promotion. The slight deviation is explained by less stable service revenue, which depends on current market demand."Partial impact
Cost of goods sold 400,000.00The cost of goods was calculated taking into account stable prices for materials and the expected production volume. 442,000.00"Due to increased demand for products during the promotion, production costs increased, as the company was forced to purchase additional materials at higher prices. Increased demand led to higher production costs, which demonstrates the dependence of cost on demand."100% impact
Operating expenses 180,000.00Operating expenses were planned based on fixed expenses for office, warehouse space and basic needs of the company. 215,000.00Operating expenses exceeded the planned ones due to additional warehousing and logistics costs during the promotion. Due to increased demand for products, some costs increased. This shows that operating expenses can also be variable depending on demand.100% impact
Labor expenses 120,000.00Salary expenses were planned in accordance with the current company staffing and payment schedule. 127,000.00Wages increased due to the need to attract additional personnel and the overtime required to process orders.100% impact
Marketing expenses 40,000.00Marketing expenses included only basic advertising expenses without additional campaigns. 70,000.00Marketing expenses increased significantly due to unplanned advertising expenses for the promotion, which led to an increase in expenses overall.100% impact
Administrative expenses 20,000.00Administrative expenses for current needs were projected to be minimal. 25,000.00Administrative expenses increased slightly due to additional costs to support the workflow.100% impact
Depreciation 30,000.00Depreciation was planned based on the gradual wear and tear of fixed assets. 30,000.00Depreciation remained at the planned level, unchanged.Minimal impact
Interest expense 20,000.00Loan servicing expenses were stable and projected in accordance with current conditions. 21,000.00A slight increase in interest expense related to a delay in settlements due to high cash turnover during the promotion.Partial impact
Taxes 40,000.00Expected tax expenses calculated at a standard rate based on projected revenues. 45,000.00Tax liabilities increased as higher revenues entail a higher tax burden.100% impact
Net Profit 300,000.00Projected net income taking into account planned revenues and expenses without significant promotions or changes. 280,000.00Net income decreased as higher expenses on the promotion and product discounts offset revenue from increased sales, preventing the company from achieving the planned profit level.100% impact
As we can see, the actual net income was lower than planned due to unexpected expenses, market changes and unplanned marketing campaigns, as well as logistics costs. Operating expenses and the cost of goods had a significant impact on the final financial result. The ability to quickly obtain an accurate demand forecast could significantly reduce the risk of deviations, as shown in the last column. After all, this would provide better predictability and greater efficiency in the distribution of expenses, which would increase the company’s profitability. So, if you want to have additional leverage over your company’s P&L indicators, SMART Demand Forecast will provide these opportunities for you with its following benefits:
Automation and speed of forecast updates AI algorithms allow for quick analysis of large volumes of data, enabling prompt update of supply recommendations, reduction of lost sales and avoidance of freezing of funds in stock.
Consideration of long-term and short-term factors The solution takes into account both long-term (economic, social and cultural) and short-term factors (temporary promotions, special offers, etc.). This ensures multi-level forecasting and accurate tailored forecasts for business.
Flexibility in forecasting at different levels of aggregation The system allows forecasting at different levels of aggregation – from national to individual regions, cities and locations. This ensures accurate adaptation of forecasts to the specifics of each level for business needs. In addition, the system supports analysis on different time scales – from short-term forecasts for operational planning to long-term ones for strategic decisions.
Customer behavior analysis The solution analyzes sales data for each store separately, taking into account individual trends and consumer preferences.
Avoiding shortages and surpluses Accurate forecasts for each store help avoid product shortages and reduce surpluses, thus optimizing warehouse stocks
Taking into account unique location specifics The solution’s algorithms adapt forecasts to individual store characteristics (area, traffic, demand), helping to build the most efficient business processes.
Supply chain optimization Thanks to accurate forecasting for each point, companies can organize the supply chain more efficiently. This helps reduce transportation costs and ensures continuous supply of exactly those products that are most needed in a particular store.
Building a single intelligent IT ecosystem SMART Demand Forecast is a modern solution based on Microsoft technologies that uses AI algorithms for highly accurate forecasting. Integration with other software products of the Microsoft ecosystem ensures data integrity and consistency of key business processes. This approach helps to create a powerful IT ecosystem that adapts to the specific needs of the company and provides ample opportunities for its development and scaling.
A McKinsey & Company study found that forecasting using AI algorithms can reduce supply chain management errors by 20-50%. At the same time, there is a 65% reduction in lost sales and unavailability of goods. In addition, storage costs can be reduced by 5-10%, and administrative costs by 25-40%. See for yourself how forecasting accuracy can affect your business profits. To do this, use special calculator from SMART business. This tool will help you estimate how much money your company will receive by increasing forecasting accuracy using SMART Demand Forecast. Don’t miss the opportunity to improve your strategic planning and increase future profitability now:
September 3rd, 2024 15 MIN READ
How the implementation of SMART Demand Forecast improved the accuracy of demand forecasting at McDonald’s Georgia
McDonald's Georgia is a fast-food restaurant chain with 23 locations. The company employs more than 1,500 employees, serving about 35,000 visitors daily.
SMART Demand Forecast implementation
McDonald’s Georgia has been maintaining a high level of service and customer satisfaction for 25 years, making a significant contribution to the reputation of one of the largest and most famous fast-food chains in the world. Thanks to the Experience of the Future initiative, aimed at improving the visitor experience by implementing the latest technologies, McDonald’s Georgia provides them with quality service. That’s why visitors always want to come back here to once again make sure that “I’m really loving it!”. Digital boards with products and offers, self-order terminals, table service, a modern mobile app – a lot has changed since the opening of the first restaurant in Tbilisi in 1999. At the same time, McDonald’s Georgia has managed to preserve the fundamental values of the establishment and give it its own identity. It is the world-famous hospitality of Sakartvelo that distinguishes this chain in its segment. The company continues to implement new methods to improve service. A significant step for McDonald’s Georgia was the implementation of SMART Demand Forecast. This comprehensive tool is aimed at increasing the accuracy of demand forecasting thanks to innovative technologies based on machine learning algorithms and artificial intelligence. This solution is a proprietary development of SMART business, which has been promoting digital business transformation in more than 60 countries for 15 years. And today we talk about the challenges, process and benefits of implementing this solution in a new success story with McDonald’s Georgia.

Prerequisites of searching for an implementation partner and how the demand was forecasted before the integration of SMART Demand Forecast

The need to implement a new solution for demand forecasting at McDonald’s Georgia was caused by several reasons. But before we talk about them, for a better understanding, remember how crowded it is in one McDonald’s restaurant and how free it is in another, located literally a block away. At the same time, on other days the situation can be the other way around. For visitors, this is a matter of just a few extra minutes, since McDonald’s establishments are known for their fast service even at peak times. But from the restaurant’s standpoint, such fluctuations in demand cause more serious challenges. The McDonald’s Georgia team allowed us to have a look at their forecasting “kitchen” and shared how the process was organized before the implementation of SMART Demand Forecast and why the old approach needed to be transformed. The company had a well-established demand forecasting process, but at the level of the entire chain. With this approach, it was difficult to take into account all the necessary factors with granularity to any McDonald’s Georgia restaurant. Accordingly, this did not provide the necessary detailing and led to the frequent need for the team to instantly respond to a shortage or excess of components for the products. In addition, this entailed unpredictable logistics costs and difficulties in inventory management. Let’s add here peak periods, mass events in the city and the proximity of retail outlets to them, as well as various promotional factors. As a result, due to unpredictable increased demand, some restaurants were forced to attract more personnel, while in other establishments, on the contrary, there was an excess of man-hours. The above factors were triggers that gave an understanding that it is more effective to make forecasts not for the entire chain at once, but for each restaurant. McDonald’s Georgia also wanted to reduce the forecast intervals. Before collaborating with SMART business, forecasts were made for the next 3 months with aggregation at the level of each month. Such a wide forecast interval did not take into account shorter-term changes in demand, such as pronounced seasonal dependence, local events or changes in the competitive environment. The company representatives noted that since the forecast accuracy indicator was calculated for the entire chain, this did not reflect the real situation in terms of individual restaurants. Let’s assume that in summer, at the level of the entire McDonald’s Georgia chain, the demand for cold drinks is predicted to increase by 20%. This forecast is based on historical data for the past few years and general trends. In the restaurant in the coastal city of Batumi, the forecast is justified due to the large number of tourists and vacationers who want to cool down on hot days. But in the central areas of Tbilisi, especially near office centers, the demand for cold drinks may increase by only 10%, since most visitors prefer coffee to perk up. Accordingly, the Tbilisi location is faced with an excess of cold beverage stock.

Discrepancies between forecasts and actual demand at the end points of sale were negatively affecting inventory management, which increased our costs. At the same time, manual forecasting did not allow us to effectively take into account all factors, even with an understanding of all causal relationships. Having analyzed the market, we realized that we needed an AI-based solution that could process large volumes of data and automatically take into account multiple factors.

  • Giorgi Asatiani Head of BI & Data Analytics, BI & Data Analytics Department McDonald's Georgia

Requirements for a new forecasting system and why the solution from SMART business was chosen

There were several key needs that McDonald’s Georgia sought to meet through future implementation:
  • The new solution had to provide forecasting with aggregation down to the weekly level. Such an approach would allow for more accurate consideration of short-term changes in demand, such as seasonal fluctuations, local events and other factors that can have a significant impact on sales in a particular week.
  • Granulation down to product and end location was also important for the customer. Since, as noted above, each restaurant in the chain may have a different level of demand for specific products. Forecasting at the level of each restaurant and individual products allows us to take these differences into account. This would provide more accurate forecasts and help avoid situations where one restaurant lacks a product, while another has a surplus. As a result, granulation increases the efficiency of inventory management and reduces the company’s costs.
  • McDonald's Georgia was looking for a solution that would allow them to switch to a centralized forecasting system. In this way, the company wanted to ensure consistency of data and forecasts across the entire restaurant chain to avoid discrepancies and errors that arise during manual forecasting.

When choosing a new demand forecasting system, we found out that SMART Demand Forecast can forecast both regular and promotional sales. We were especially interested in the system's ability to work with analogs and anomalies. The vendor outlined the implementation process in detail, which included launching a pilot project with the ability to assess the accuracy of forecasts at the initial stage. In addition, a clearly designed product development map gave us confidence in long-term cooperation, which together was the decisive factor in choosing a solution from SMART business

  • Giorgi Asatiani Head of BI & Data Analytics, BI & Data Analytics Department McDonald's Georgia
Learn more

How was SMART Demand Forecast implemented for McDonald’s Georgia?

SMART Demand Forecast was implemented into the business processes of McDonald’s Georgia in three stages: diagnostics, modeling and implementation.
  • Diagnostics First, the SMART business team analyzed in detail how the customer’s business processes were organized “as is”. At the next stage, a clear vision of how forecasting would work at McDonald’s Georgia with SMART Demand Forecast was defined. The value of this step was that SMART business offered the customer a modern approach and tools for building accurate forecasts without the need to change the company’s existing business processes.
Next, a process diagram with a detailed flow was built for a clear vision by the McDonald’s Georgia team of future steps and the possibilities of planning their further actions in accordance with the plan. Additionally, roles and responsible persons were assigned, terms were defined, and priorities were set. All this provided a structured approach to introducing changes and guaranteed that all project participants understood their tasks and areas of responsibility. Thanks to this, the implementation was as smooth and productive as possible. At the diagnostic stage, a unified data structure was approved by the customer, which will be used by SMART Demand Forecast to build demand forecasts. It is important to note that for solutions based on ML & AI algorithms, a unified data structure is the key to achieving efficient model operation and high forecast accuracy at the output. The SMART business project team analyzed the customer’s key business processes and came to McDonald’s Georgia with a ready-made template. The customer only had to collect and transfer the necessary data to the vendor. In addition, SMART business developed technical documentation describing all aspects of the implementation and use of the solution by McDonald’s Georgia.

For us, the value is that the vendor independently determined what data is needed to build a high-quality demand forecast, focusing on its own expertise and the specifics of our business processes. Accordingly, we, for our part, simply provided information according to the universal data structure (sales, product hierarchy, promotional campaigns, external factors, etc.). And the SMART business team adapted them and built ML models that are used in the system and take into account cannibalization aspects and additional calculation factors that significantly influenced demand.

  • Giorgi Asatiani Head of BI & Data Analytics, BI & Data Analytics Department McDonald's Georgia
Indeed, the SMART business team analyzed and highlighted those points of influence on the forecast that the customer may have been unaware of. This is a painstaking and multi-stage process consisting of a thorough audit and observation, briefing key employees, using advanced methods of analyzing large volumes of data and machine learning algorithms. The team uses data visualization tools for a more transparent understanding of the results obtained. Thus, at the diagnostic stage, it was possible to identify many local challenges of McDonald’s Georgia that needed to be taken into account when implementing the solution. We would like to highlight the most interesting of them:
  • Anomalies – that is, unusual or unexpected spikes or drops in sales that stand out from standard patterns. At McDonald’s, this could be a bus passing by with hungry tourists who visit a specific restaurant. In this case, this event is not systemic and, accordingly, it is not taken into account in standard data and is labelled as an anomaly. SMART Demand Forecast identifies such incomprehensible changes and cleans the data so that they are not involved in forecasting.
  • Dependence of individual SKUs on seasonality – the customer has many different product items that have a seasonal dependency. For example, demand for McFlurry usually increases in the summer when it is hot. Therefore, the system automatically calculates special seasonal factors and coefficients.
  • Forecasting demand for new products introduced during promotional campaigns – the difficulty is that new items do not have historical sales data. But SMART Demand Forecast uses an integrated approach that takes into account multiple factors, such as similar products, common sales patterns, seasonal trends, and other factors. This approach differs significantly from traditional manual forecasting, which is often limited in terms of processing large volumes of data. The automated solution allowed for faster and more accurate data processing, providing more accurate forecasts for new products.
It is important to note that if the customer has a non-standard request that needs to be taken into account in forecasts, SMART Demand Forecast supports manual demand forecasting. This feature allows them to quickly enter additional clarifying data and flexibly respond to changes in the conditions of market volatility. At the same time, the possibility of automating any specific needs at the customer’s request was discussed as a nice to have for the future. The final step of diagnostics was preparing for the transition to the next stage – modeling.
  • Modeling The main task of this stage in McDonald’s Georgia was to train the artificial intelligence model so that its algorithms were as accurately oriented to the needs of the business as possible. Since the “fit once = fit every time” approach does not work for different companies, even from the same industry, the SMART business team customized the model for McDonald’s Georgia:

First, based on basic factors affecting demand, such as the day of the week, sales data for previous months, product prices, etc., an initial model was built that provided a certain level of forecast accuracy. Then we moved on to optimizing the model by adding new factors: cannibalization, seasonal coefficients, elasticity factors, etc. We observed what improved and what did not. The main thing here is not to overdo it and not to overload the system with unnecessary things, because an overtrained model can make false forecasts. Our team has deep knowledge and all the necessary tools to determine the importance of the influence of each factor. This allowed us to filter out the least significant factors, reduce the “noise” of the data and improve the accuracy of forecasts for each individual McDonald’s Georgia restaurant.

  • Artem Stepanov Product Owner SMART Demand Forecast

What did McDonald’s Georgia gain from implementing SMART Demand Forecast?

After identifying the key factors, the model began to produce stable forecasting results. An optimal number of factors was formed, by which McDonald’s Georgia can receive the most accurate indicators with minimal acceptable fluctuations. The success of the project was determined by achieving a forecast accuracy of 70 to 80%:
  1. on a horizon of 4 calendar weeks from the start date of the planning period,
  2. for each calendar week,
  3. for each of the 21 declared restaurants of the network.
The project started in October 2023 and was fully implemented in July 2024. The success of the implementation was evident already at the pilot project stage. It started in the summer, when there are many different seasonal factors and, in general, the demand of the McDonald’s Georgia chain grows the most. As a result, it was possible to achieve the following target indicators that even exceeded expectations:
  • 83% accuracy of sales forecasting for each restaurant based on weekly data aggregation for a period of 4 weeks.
  • 80% accuracy of sales forecasting for each restaurant based on weekly data aggregation for a period of 12 weeks.
  • Average deviation of up to 5% from forecast fulfillment, which is the norm among the global business community.

We are very pleased with the results of the SMART Demand Forecast implementation. Thanks to this project, we were able to significantly improve the accuracy of demand forecasting in each of our restaurants. And when we received the first results, we clearly understood that the return on this investment will be quick. We are grateful for the high professionalism and well-coordinated work of the entire SMART business team, which helps us feel and understand our business processes at a new level.

  • Giorgi Asatiani Head of BI & Data Analytics, BI & Data Analytics Department McDonald's Georgia
But even such high indicators are not the limit for forecasting improvement. McDonald’s Georgia together with SMART business are going to continue to develop the functionality of SMART Demand Forecast, which will cover omnichannel forecasting.

Implementation of solutions with artificial intelligence requires constant cooperation, analysis of the steps taken and close communication between the vendor and the customer, which in tandem brings high results, which we managed to achieve with McDonald’s Georgia. Our teams were always in touch, responded promptly to the issues, which positively affected the productivity and success of this project and laid a solid foundation for further cooperation.

  • Artem Stepanov Product Owner SMART Demand Forecast
The mission of SMART business is to help entrepreneurs make the future more understandable and transparent to reduce risks and be ready for any challenges. The company prides itself on contributing to the success of its customers around the world and empowering them to achieve their key goals.
July 24th, 2024 7 MIN READ
Release 4 0 SMART ENG
Release 4.0. SMART Demand Forecast: building a forecast is now even more convenient
The state-of-the-art demand forecasting system SMART Demand Forecast uses multi-factor models, complex algorithms and analyzes huge amounts of data to produce accurate results. Therefore, we are constantly working to ensure that you can make accurate forecasts easily and quickly, thus leaving complex processes to the system. Changes in the interface, improvement of models, addition of new features – these are all constant improvements for better results for your business. The SMART Demand Forecast update will make the forecasting process even more convenient and accurate. New features and improvements are aimed at improving the user experience, simplifying the setup and optimizing forecasting algorithms. With these updates, users will be able to quickly and easily get accurate forecasts that will help them make more informed and effective business decisions. In this article we will look at new possibilities for working with the system. Improved work with anomalies: manual correction and three viewing modes The ability to manually adjust anomaly smoothing has been added to the Anomaly Processing page. You can make changes to anomalies at all available product and business levels, after which other dependent values ​​are recalculated. In addition, you can now switch between different modes in the main panel and view information for the selected period: week, month or year. Breaking a data set into smaller pieces makes it easier to view and search. The new approach to working with anomalies improves metrics such as Time to Interactive (TTI) and First Contentful Paint (FCP), thereby ensuring convenient work with large arrays of anomalies. Manual adjustment of the forecast result You can adjust the forecast result manually on the Modeling page. The adjustment applies at all available product and business levels. The system automatically distributes changes made to the lowest forecast level. This ability to manage the forecast increases the level of relevance of the data in accordance with any changes in the company or market. Automatic recognition of the best model In the system you can see the optimal model, which SMART Demand Forecast will automatically recommend based on its validation period metrics. You can find the selected model in the list with a corresponding mark that will distinguish it from others. Thus, you save time on determining the optimal option among those proposed. Now the analysis will not take too much time, and you will be able to build a forecast faster and more efficiently. Ability to adjust the conditions for promotional campaigns Now you can change the settings of promotional campaigns, such as start and end dates, product, store, promo type, regular and promotional price and discount – directly in the system interface. Settings are saved in the system after the entered values ​​are verified and implemented. This approach saves significant time making changes because you no longer have to re-upload CSV files with updated data. Now planning promotional campaigns becomes more flexible, and therefore more effective. New filters and sorting for the anomaly table On the Anomaly Processing page, you can quickly filter and sort by store and product parameters. Filtering helps group data, making it more structured and easier to analyze. DS Settings for model training A separate space is allocated for training models, where you can select groups of factors on the DS Settings tab. Analysts and Data Science teams have another tool to help them achieve greater forecast accuracy. This reduces the time for experimentation and increases the efficiency of adapting the system to specific business processes. Ability to select date and time format For convenient use of the system, you can configure the desired time zone to synchronize the time on the device or in the region where you work. To set the time, go to User Settings > Profile Settings and select the desired format. Verification of analogs for products and stores with insufficient history You can obtain information about how correctly an analog from another category is set for a specific product or store. This automatic check minimizes errors and ensures forecast quality. Thanks to this functionality, the accuracy and reliability of setting analogs increases. Improved operation of switches and filters in tables and modal windows When making a request with new filtering conditions, all previous requests are canceled, leaving only the relevant parameters in operation. Eliminating redundant requests improves system speed and helps specialists work with filters. Now you can quickly change parameters and get updated results without wasting time manually canceling requests. Improved performance of the Isolation Forest algorithm You can use additional factors from the Universal Data Structure to search for anomalies. The system has a switch for using additional factors when searching for anomalies using the Isolation Forest algorithm, which can be enabled. This option will allow you to more accurately and flexibly identify genuine anomalies in sales, taking into account various factors such as price, product and store characteristics, etc. Simplified work with group elements Now you can collapse and expand group interface elements on different pages of the system. The settings are saved automatically, and you can use them the next time you log in. This helps you organize the interface according to your needs, quickly find important information and work with the data you need. A new “Weather Factors” volume building block Weather factors are now available for model training and scoring. All the system needs is data on the geolocation of points of sale. These factors include precipitation, temperature, humidity and other weather characteristics. The new feature can significantly improve the quality of forecasting seasonal goods. Baseline forecasting model For training and evaluation of forecast building, a new Baseline model is now available, which works based on linear regression. The model has a basic set of factors from the UDS and is not burdened with the calculation of additional factors, which allows it to train and produce results many times faster than traditional models in the system. This approach is especially relevant for building trends – a quick result with a certain number of parameters, which is a classic method of initial analytics when planning long-term campaigns. A new approach to using Azure Spot Instances SMART Demand Forecast can now significantly save your Azure Databricks resources involved in the complex processes of training and evaluating forecasting models. This approach consists of keeping computing resources in a specific shared resource group, where prices are significantly lower due to the small chance of transferring these resources to more privileged users. Processes in the system do not suffer from this transfer in any way. This allows you to save time on training and evaluation processes and get better functionality without increasing costs. Improved quality of the ML core code of the solution As part of the current release, work to improve the code of the Databricks framework and Azure Machine Learning have been tested and implemented. Automatic pipelines for styling code in accordance with the Pep-8 standard have been implemented. Work was carried out to improve the efficiency of code and CI/CD processes. The system has become even more stable and reliable in operation. Solution optimization in terms of back-end Changes from the Domain Driven Design and status processing section, microservice for access to Azure Data Factory has been implemented. These improvements are of an optimization nature and, as a result, improve the user experience when working with the system.
February 13th, 2024 13 MIN READ
Forecasting in the supply chain: How to achieve the perfect balance of supply and demand
If the supply chain is the path of goods from raw materials to the end consumer, then the accuracy of the consumer demand forecast will determine how efficiently this process will be accomplished. This means that all chain participants, from manufacturing companies to retailers, will be able to use their resources to maximize profits and optimize costs. In fact, the demand forecast is the beginning of the entire chain. It determines the workload and "work schedule" of all its links. It is also the beginning of production planning, and it is the basis on which distributors and retail stores rely when purchasing and transporting goods. Any mistakes made at this stage can be very costly. A shortage of manufactured or purchased products when there is demand means lost sales or additional costs for urgent replenishment of stocks. According to RetailDive, businesses lose about $1 trillion a year due to the inability to meet customer demand on time. Overstocking is an equally annoying phenomenon. Excessive inventory requires additional funds to store goods that are not in demand and, in the case of goods with a limited shelf life, to write them off. You can easily calculate how an accurate forecast will affect your company's sales using our interactive calculator. Therefore, innovative companies are massively implementing modern forecasting solutions based on artificial intelligence; one of them is SMART Demand Forecast. These solutions allow companies to make the forecast as accurate as possible, considering a large amount of diverse data and identifying non-obvious patterns. According to a study conducted by McKinsey & Company, 20% of companies are already using the latest technologies that plan and forecast demand in the supply chain using AI. And 60% of organizations have already planned to implement such technologies. The rest still rely on expert analysis, considering trends in sales history and insights from business analysts. However, experience shows that such calculations are often significantly less accurate than the results provided by tools with properly selected and configured artificial intelligence algorithms. There are different types of demand forecasting in supply chain management: quantitative, qualitative, etc. But best practices usually include AI algorithms and advanced analytics.

Benefits of accurate demand forecasting in supply chain management

Forecasting tools like SMART Demand Forecast can greatly reduce all mentioned risks and increase supply chain management efficiency. Let's take a closer look at the benefits that businesses get from an accurate demand forecast. Minimization of lost sales As the above statistics show, companies lose significant amounts of money due to their inability to meet consumer demand. This happens if analysts underestimate or fail to consider the factors or hidden patterns that led to a more intense spike in demand for certain groups of goods than expected.  Consequently, the supply of products was not organized properly. Organizations find themselves in a situation where they are simply unable to sell a popular product due to its unavailability. Or they are forced to make additional efforts and funds to replenish stocks as soon as possible, overpaying for the urgency of order fulfillment. Increase in customer loyalty  If a customer or buyer periodically has difficulties purchasing the products they need, this can be a serious threat to the image of the supplier or store and the chain. Poor service will lead to the fact that you will at least lose customer loyalty, or in most cases, lose them completely. At the same time, attracting new customers requires much more money than retaining existing ones. In addition, if you don't have the goods when there is demand, you will lose profit. An accurate forecast helps you maintain the required level of inventory at all times, especially during promotional campaigns. Ability to respond quickly to fluctuations in demand One of the difficulties in forecasting supply and demand is the changing nature of consumer behavior, which can be influenced by many factors, such as hot trends, competitors' activities, and even the weather. It is clear that even artificial intelligence cannot take into account unpredictable factors, as it processes only the data provided to it. However, it will definitely be able to take into account all the necessary factors for a more accurate and timely forecast of demand fluctuations, which will help manage the supply chain more efficiently and flexibly. Reduction of logistics costs  Inaccurate forecasting of demand in the supply chain also leads to unnecessary logistics costs. For example, if a forecast for a chain of stores is based entirely on analysts' estimates and assumptions, it means that:
  • the forecast will often be less accurate and more pessimistic than realistic;
  • it will usually be made for the entire network, with a further approximate distribution between different stores.
As a result, some stores will usually have an excess of products, while others will have a shortage. To overcome this imbalance, companies have to redistribute the volume of goods between outlets, which, in turn, leads to additional logistics costs. These costs increase production costs and "eat" margins. Modern tools make it possible to forecast demand for specific items for each store separately. This helps ensure optimal stock levels in all stores of the chain. Reduction of warehouse costs  An erroneously overestimated demand forecast leads to the fact that some products remain unsold. These products must be stored in warehouses and become the "frozen capital" of the organization, resulting in additional costs for renting warehouse space. Minimization of the number of written-off products An excess of unsold goods with a limited shelf life often leads to the fact that, at some point, they must be written off. That means that all the money invested in their purchase, delivery, and storage is simply wasted. In addition, the write-off process itself involves certain costs. In order to at least partially avoid write-offs, companies are forced to sell products significantly cheaper while also spending an additional budget on marketing activities. More effective promotions The use of supply and demand forecasting tools can make promotions more effective and manageable. But it requires systems that can calculate both regular and promotional sales volumes. SMART Demand Forecast is one of these systems. For example, they help you determine the optimal discount for certain items. Let's assume you plan to reduce the price of a product by 30%. The system may find that with such a discount, demand will be so high that suppliers will not be able to meet it. Meanwhile, with a 15% discount, you will sell fewer items but earn more due to the higher price. Based on the system's forecast, the supply chain manager will be able to make a reasonable decision on how many products to purchase, depending on the optimal discount. In addition, tools that forecast promotions can take into account the so-called cannibalization, which is very important. When you set a discount for a certain product, the higher demand for it "eats" sales of similar non-promotional products. Some companies forecast only regular sales, and these forecasts are always wrong because they do not take into account the cannibalization factor. Systems like SMART Demand Forecast automatically and accurately reduce the demand forecast for non-discounted products and increase it for similar discounted products. More effective pricing policy An accurate forecast of demand fluctuations for various items also helps to set optimal prices for them. For example, if the system "sees" that there is a demand spike for a certain product, it gives a reason to set a higher price, and vice versa. In addition, a forecasting solution will help determine what price will maximize sales. Forecasting features for manufacturers and distributors Demand forecasting is the beginning of the supply chain for all participants, including manufacturers and distributors. Production planning begins with the forecast. It is used to calculate the required amount of raw materials and production capacity to meet the expected demand. It considers the stock level of products and the production processes that have already been launched. And while such companies usually make a profit from sales to distributors, the so-called primary sales, it is important to take into account secondary sales for a correct forecast: how much of your products the distributor will deliver to retailers. For example, if a distributor plans to run a promotion for certain products in stores, it means that the distributor will buy more of those products from the manufacturer to meet the higher retailer demand. It may also turn out that the distributor purchased less of your products than they shipped to stores during a certain accounting period. The ratio of goods purchased and sold by distributors will help you determine how much product will be purchased in the next accounting period. Distributors engaged in secondary sales make their own forecasts based on insider information from retailers (planned store promotions, their sales forecasts, etc.). But since such information is not always available, distributors are forced to keep a safety reserve, which is determined considering the deviations from previous forecasts. It is crucial for them to maintain a stock that, on the one hand, is not excessive and, on the other hand, ensures the constant availability of the necessary goods. If at some point the distributor does not have enough products for the retailer, the retailer will address another supplier or sign a direct contract with the manufacturer. Thus, the distributor simply drops out of the supply chain.

How artificial intelligence helps to build accurate demand forecasts

AI-based solutions for demand forecasting in the supply chain greatly simplify analytical calculations and make them more accurate. Let's take a closer look at the benefits of using such tools. Labor cost reduction If a company does not use modern tools for demand forecasting, it is forced to recruit a large staff of analysts. But even an entire analytical department requires a lot of time to consolidate information, perform complex calculations, process large data sets, and identify various dependencies and trends. Using an AI-based demand forecasting solution, an analyst only needs to upload correct and comprehensive data to the system, start the calculation process, and get a ready-made forecast. The only thing left to do is to demonstrate the result to the management, explain why this result was achieved, and suggest measures that would bring the forecast closer to the organization's business goals. Thus, the analyst gets rid of a large amount of troublesome and complex work and has more time to actively participate in decision-making. Instead of being stuck in an endless routine as a "computer operator," they become a forecasting expert. Minimization of human error Supply and demand forecasting solutions also minimize the possibility of human error, which can cause significant losses for businesses. The analyst does not have to make any calculations. All the necessary formulas and algorithms are already built into the system. The user understands how the forecast is calculated, but the process is automatic, fast, and guaranteed to be correct. A single database Thanks to tools like SMART Demand Forecast, all information is stored in a single, unified database. In companies that don't use such forecasting solutions, analysts usually have to combine information from disparate systems. They need to get master data from one system, a sales report from another, data on promotional campaigns from a third, and so on. Using SMART Demand Forecast, all departments and all employees work in a unified system, in a single format, in a single interface. The benefits of this are especially noticeable when a certain employee falls out of the workflow for some reason, for example, by going on vacation, taking a sick leave, or resigning. Another employee can easily continue to work with the information that the previous employee entered into the unified system. Without the unification provided by modern forecasting solutions, it can be very difficult for one employee to understand the calculations of another, thus jeopardizing the continuity of business processes. Consideration of all necessary factors One of the main difficulties in demand forecasting in the supply chain is that many factors, both internal and external, must be considered. This is quite difficult to do, even if a team of experienced experts is working on the forecast. Artificial intelligence solutions can do this quickly and accurately if all the necessary data has been entered into the system. The list of factors that can be taken into account by artificial intelligence is quite long. Here are just some of them: - research results; - data on competitors; - macroeconomic trends; - supplier's promotions; - own promotions; - cannibalization; - data from social networks; - weather forecast; - data from POS terminals; - events in the country... Artificial intelligence is able to take these and other factors into account, compare them, and find patterns, interdependencies, and hidden trends that are not obvious to analysts. Thus, the forecast will be more accurate and well-justified. Chain reaction It is statistically proven that the extra costs associated with poor forecasting in the supply chain on average make up 2% of the cost of sales. If the analyst makes a mistake, the company's losses from such a forecast can be much higher. Systems using artificial intelligence and machine learning technologies, such as SMART Demand Forecast, are able to minimize these losses by making more accurate calculations. That's why the vast majority of companies in the world are either already using such solutions or preparing to implement them. Would you like to learn more about how accurate demand forecasting in the supply chain will affect sales in your company? Request a personalized presentation.
December 6th, 2023 9 MIN READ
Release 3 MAIN ENG
Building a forecast has become even easier with the new SMART Demand Forecast release 3.0
Improving forecasting accuracy plays a crucial role in contributing to strategic business planning by helping to understand market trends, customer demand, and customer needs. This allows you to effectively manage inventory and logistics, minimizing unnecessary costs and risks of shortages. Modern forecasting systems, such as SMART Demand Forecast, use multivariate models and analyze a huge amount of data to predict changes in demand. To do this more effectively, the system must be constantly improved. In this update, we have done a lot of work to optimize all system processes and the interface to ensure easy use of the solution, data download, table viewing, and analytics. We also added new key functionalities:
  • data anomaly detection and processing;
  • the ability to decompose the forecast into key components;
  • forecast calculation factors: external factors (natural phenomena, force majeure, exchange rate fluctuations, etc.) and competitors' influence;
  • forecasting model – XGBoost;
  • Power BI reports and much more.
Read more about all system updates in the article below. The definition of technical statuses and their display in the interface has been updated Changes to the interface and statuses make it easy to navigate the system and quickly adapt it to changes in company processes. This makes SMART Demand Forecast a more flexible solution and helps new users learn, as the program is intuitive and meets modern information presentation standards. Dynamic adaptation to changes in the number of product levels has been added The workflows associated with building a demand forecast become easier due to the flexibility in managing products in the assortment. The system is now able to quickly adapt to changes in the product structure, making it much more convenient to use in everyday work. The update allows you to easily scale and customize the system to meet your current needs with less time and human resources. Improved visualization of system processes and data display Thanks to optimized algorithms and improved data processing, process visualization has become faster and smoother. Now the information on the screen reflects what is happening during system operation more accurately. It is easier for the user to see the stages of data processing, so interaction with the solution will be as comfortable and understandable as possible, even for new employees. Optimized recovery in case of connection loss For the SMART Demand Forecast team, one of the most important areas is providing customers with a stable performance of the solution. The system now monitors processes in real time, so if the Internet connection suddenly goes down, updated recovery strategies are automatically launched. The work process is now smoother, despite external factors. Now you don't have to worry about losing the data you've been working with. This ensures not only high-quality work in technical terms but also reduces employee stress and minimizes unplanned interruptions in work. The ability to process anomalies has been added The issues of finding and processing anomalous data are always in the first place for systems that work with large amounts of data and use artificial intelligence. SMART Demand Forecast is no exception, but thanks to the accumulated expertise, the ability to detect anomalies in historical sales, smooth them, and correct them has been introduced. Synchronous scrolling of anomaly tables and graphs has been added When working in SMART Demand Forecast, you can simultaneously scroll and compare information in tables and anomaly graphs. This feature provides a more convenient and efficient way to analyze information that will affect the demand forecast. Now users can identify important patterns and anomalies in the data faster and form an analytical basis for further decisions in supply chain management. The ability to decompose the forecast into component parts has been added The key elements that make up the total volume of the forecast are the blocks of component parts. Their decomposition helps to more accurately predict future trends, identify key drivers of change, and develop more effective resource management and planning strategies. The report "Component parts blocks" was created in the system. It allows you to quickly analyze which factors influenced the forecast up or down, make strategic decisions on deliveries and promotions, etc. based on these factors. A new interface for importing and exporting analog tables has been implemented Continuous improvement of the system's appearance is one of the important areas of development for SMART Demand Forecast. Therefore, this system update includes the integration of a new interface for importing and exporting analog tables. Thanks to the simple and clear display of information, users can interact with the system faster. This reduces the amount of time and effort required to perform operational tasks. A new range component has been added for anomaly search settings With this feature, users can easily set the contamination level for the Isolation Forest search method through an intuitive interface, simplifying the process of identifying anomalies in data. This increases the productivity of the system and enables the team to prepare high-quality analytical reports. CI/CD processes have been optimized As a part of the 3.0 release, CI/CD processes have been modified, which significantly increases the speed of deployment and quality control of product delivery during development, testing, and delivery to the client. Although this is not obvious to users, it is an important aspect of the work, as it helps to quickly and efficiently implement new features and continuously improve the system in the future. The ability to use the calendar during modeling has been expanded The updated interface of the SMART Demand Forecast calendar ensures its effective use, allowing you to quickly obtain the necessary information and view data at different levels of aggregation. In addition, the ability to quickly select a date to open the planning period has been added. The approach to forecasting for new products has been updated. New products in the assortment make it difficult to build a demand forecast. The standard logic of providing analogs for forecasting works, however, there are situations when you need to set a large number of analogs for hundreds or thousands of products. Now this is possible without specifying an analog but based on average sales in the category. A new machine learning model, XGBoost, has been added XGBoost is one of the most successful machine learning models (extreme gradient boosting) and the winner of many Kaggle competitions in the field of forecasting and regression tasks. By adding this model to SMART Demand Forecast, the company gets the opportunity to experiment with different forecasting tools and choose the most effective ones. TFT model performance has been optimized The TFT (Temporal Fusion Transformation) model implemented in release 2.0 has been refactored and improved in terms of performance and accuracy. First of all, we improved its accuracy, which makes it possible to produce better forecasts and thus increase business performance. New factors have been implemented: Bass diffusion, influence of competitors and external factors The Bass diffusion factor is based on research conducted for the previous release. It helps to identify patterns for products with a short history. Experiments have shown that this factor is very useful for forecasting all products, as it helps to take a more comprehensive approach to building a forecast. It is also possible to take into account such important factors as the influence of competitors if the relevant data is available. We continue to work on taking into account external factors of demand formation. It is now possible to take into account economic factors.
  • A new analytical data model has been created In release 3.0, we optimized and created a new concept of analytical reporting:
    • The existing Power BI reports were adapted; The "Compensated Sales" report allows the user to analyze the facts of shortages of goods in certain stores and see how much revenue was lost. It helps to reduce losses in the future. The "Sales Analysis" report allows the user to work with the actual sales history, compare different periods, and see trends. The "Forecast Analysis" report compares actual sales data with the final forecast for previous calendar periods and evaluates its accuracy.
    • There are 5 new Power BI reports that have been added: The "ABC-XYZ Analysis" report displays metrics by product and store categorization, depending on sales volumes and volatility. The "Stock balances analysis" report shows how long certain products will last in certain stores given the forecasted demand. This allows you to make decisions about planning the supply of goods to stores and determining the optimal Safety Stock. The "Product Availability Analysis" report shows for which products in which stores there were shortages during the historical period, which allows you to identify shortages in supplies for more efficient planning. The "Validation Analysis" report has been updated. Now it allows you to analyze how well the validation data set of the forecast coincides with actual sales. This allows you to evaluate the accuracy of the forecast results and make a decision on the further use of the trained model. The "Final Forecast Analysis" report allows you to compare the scenarios of the created forecasts with each other and make a decision on finalization by selecting one of the scenarios.
    • a unified design concept was developed;
    • the main metrics for analysis were agreed upon.
Now it is possible to analyze and visualize both the current state of the business (actual sales and inventory) and the results of the DF application (visualization of the forecast and its accuracy metrics). For more information about SMART business solutions and services, please call +38 (044) 585-35-50 or submit your request.
September 28th, 2023 11 MIN READ
The role of forecasting in demand planning: Strategies and methodologies
Demand is the driver of any business. However, when the perception of the level of consumer demand is far from reality, the business will face one of two troubles. The first is the cost of storing stock or, even worse, writing off unsold products. The second is the loss of income and, at the same time, customers due to a lack of goods on the shelves. Both are quite annoying. Especially now that there are modern solutions on the market based on machine learning and artificial intelligence technologies that allow you to accurately predict demand fluctuations and help businesses make quick management decisions based on up-to-date data. Companies are increasingly using such tools to gain a clear competitive advantage. As a result, the market for demand forecasting and planning software continues to grow. While in 2021 its volume amounted to USD 3.5 billion, in 2028 it is expected to reach USD 6.8 billion, which is almost double. If you would like to learn more about how demand forecasting tools can help you prevent write-offs, please request a personalized presentation.

What key task does demand planning solve?

Demand planning is the starting point in supply chain and business management, affecting almost all KPIs of an organization. Like a nerve impulse, it sends a signal to all links in the chain, adjusting their workload in accordance with the expected sales volume. Procurement, supply, production, logistics, warehouse operations—everything is adjusted to ensure that there are enough goods to meet consumer demand while preventing excess finished goods. The key here is to maintain a perfect balance when there is enough stock, but this sufficiency does not turn into excess. Maintaining such a balance is complicated by the fact that the market situation is constantly, and sometimes rapidly, changing. It can be influenced by macroeconomic factors and many other factors, including industry trends, competitors' activities, the political situation, natural disasters, and high-profile events—down to a single post by an opinion leader on social media. All of this affects the already volatile and unpredictable consumer buying behavior. Of course, it is impossible to take everything into account, but ideally, businesses should be able to react to these changes—or better yet, be ahead of at least those that can be predicted. That is why the accuracy of the forecast is particularly important in demand planning. It will determine whether the company will be able to provide optimal volumes of certain categories of goods in the right locations at the right time.

How can inaccurate demand forecasting affect financial performance?

An inaccurate demand forecast can have significant negative financial consequences for an organization, especially if the deviations are significant. For example, in cases where analysts fail to take into account factors that will lead to a significant increase in sales, this may at some point lead to the fact that the company will not be able to sell some of its goods because there simply will not be enough of them, or they will not be delivered in time at the right time and place. Thus, the organization will lose the opportunity to make a profit from the sale of its products at the peak of its demand, and customers will be disappointed by the inability to make the purchases they need. And there is no telling what is worse. Customer loyalty is very easy to lose, but it takes a lot of time and marketing efforts to regain it or attract new customers. Even if a company manages to fill product gaps by moving products from other locations, it still comes at the cost of additional logistics costs. Thus, underestimating demand when planning sales can be costly. However, overestimation also has unfortunate consequences. Companies have to spend money on storing unsold products, paying for additional warehouse space, and the accumulation of excess inventory blocks working capital. In addition, organizations must reduce the price to sell the stock as soon as possible, and if the expiration date is over, they must write it off altogether. Product write-offs are the most annoying and common consequence of inaccurate forecasts in retail. It is clear that the consequences for companies are particularly painful when they make serious miscalculations in demand planning. Minor mistakes lead to less noticeable problems, but when they happen on a regular basis, it still significantly harms the business in the long run. Therefore, organizations that want to optimize their processes as much as possible and eliminate any causes of unnecessary costs are implementing modern demand forecasting tools. In today's competitive environment, accurate forecasting and, as a result, efficient inventory management can be a decisive competitive advantage.

What benefits does a business get from accurate demand planning?

Forewarned is forearmed. Being aware of future demand fluctuations in advance allows companies to prepare for them, which means they can manage their resources more rationally and implement optimal product strategies. Let us look at the main advantages of an accurate forecast in demand planning.
Sales maximization Timely knowledge of demand surges allows organizations to make the most of them by preparing the necessary amount of goods for sale and avoiding shortages. This leads to an increase in sales of an average of 15%.
Reducing the cost of storing surplus goods On the other hand, anticipating downturns in demand helps to avoid excess inventory and thus the cost of storing it in warehouses.
Reducing the volume of written-off products Maintaining optimal inventory levels and avoiding overstocks also ensures that fewer products are written off due to expiration.
Improved service Accurate demand forecasting allows you to maintain the constant availability of products that customers need without unnecessary logistics and warehouse costs. Customer satisfaction increases by an average of 20%.
More accurate data for business decisions A demand forecast is one of the starting points for making management decisions. The more accurate the forecast, the more successful the decisions will be.
More reasonable pricing policy The level of demand is one of the key factors that should be taken into account in the pricing strategy. Accurate demand planning makes it possible to set prices that are optimal for the market, which is one of the important prerequisites for successful sales.
More efficient marketing Marketing activities yield the best results when they take into account future fluctuations in demand for certain product groups, taking into account cannibalization. Thus, the effectiveness of promotions increases by 15%.
In general, an accurate forecast can increase business profitability by 25%. Of course, this figure varies from company to company. You can use a calculator to understand how much profit your organization can make from accurate demand forecasting.

What are the main methods of demand forecasting?

Traditionally, forecasting methods are divided into two large groups according to the level of formalization: intuitive and formal or statistical. The former, despite their name, are based not so much on irrational feelings as on the insights and experience of experts. That is why they are also called "experts." They are divided into individual (when the forecast is made by individual specialists) and collective (when forecasting takes place during one-time or serial discussions). Typically, intuitive methods are used as a supplement to formal methods and in situations of high uncertainty when there is a lack of data for analysis. For example, when a company has just entered the market or is opening a new business line. The advantage of these methods is that they are more versatile and can cover a wider range of factors than formal methods. The disadvantage is the subjectivity and accuracy of forecasts. Formal methods are based on a mathematical approach and the ancient truth that the best way to understand the future is to study the past. These methods are used when a sufficient amount of historical data is available for analysis, primarily in relation to sales. By studying the trends of past periods, analysts extrapolate them into the future. In doing so, they use complex mathematical formulas and models that consider various factors and interdependencies that may affect demand. The advantage of this approach is its objectivity and the fact that it produces results in quantitative terms. The disadvantage is that it is not suitable for identifying complex nonlinear patterns and that many factors are left out of the equations.

How can IT solutions improve demand forecasting?

Modern demand forecasting tools using AI and machine learning, such as SMART Demand Forecast, can be used as an effective alternative to traditional intuitive and formal methods or complement them. In any case, they provide businesses with advantages that were not available before. At the very least, they significantly simplify and speed up the forecasting process by taking over complex analysis and calculations—something that required long and painstaking work of an entire analytical department. Moreover, as practice shows, such software is usually more efficient than traditional statistical methods. One McKinsey Digital study showed that demand forecasting using artificial intelligence can reduce supply chain administration costs by 25–40%, reduce lost sales due to out-of-stocks by 65%, and reduce inventory by 20–50%. How is this efficiency achieved? AI and machine learning software combine the precision of traditional statistical analysis with the ability to take into account a large number of factors inherent in expert analysis. These tools process entire arrays of both structured and unstructured data from various sources, and at the same time, they usually produce results with greater accuracy than experts because they are able to spot hidden patterns. In addition to the sales history of different types of goods in different periods, the following may be taken into account:
  • macroeconomic indicators
  • market trends
  • competitors' activities;
  • survey results;
  • own marketing activities;
  • information from POS terminals and IoT sensors;
  • data from suppliers and distributors;
  • weather forecasts;
  • information from social networks, etc.
Machine learning algorithms are able to process all these data sets and find both linear and non-linear correlations between them, "see" patterns and trends, including non-obvious ones that a human analyst may not notice. As a result, a predictive model is generated, which will be used for demand planning. Moreover, the model can learn and improve over time, and the more data is added to the system, the more accurate the forecasts become. It is worth noting that much will depend on choosing the right algorithms. However, this is the task of forecasting solution providers. If you choose SMART Demand Forecast, our experts will help you with the right choice, testing, and further use of the algorithms that will provide the most accurate forecast based on the specifics of your business. It is worth emphasizing that the fundamental advantage of AI-based demand forecasting tools compared to traditional statistical methods is their ability to take into account not only the past (historical sales data) but also many current factors. Moreover, information can be fed into the system in almost real time, such as data from POS terminals, IoT sensors, and signals from websites and social networks. This makes the forecasting process continuous and more flexible, demand planning more adaptable to the changing circumstances of today, and the supply chain more closely aligned with real consumer demand.

Plans for success

For any company, demand planning is both a starting point and a benchmark. It determines the extent to which the organization's activities are aligned with customer buying behavior and how effectively the business will respond to market demands. Therefore, it is crucial for companies to use effective demand forecasting tools. Solutions using artificial intelligence and machine learning algorithms demonstrate the greatest proven effectiveness and are replacing traditional tools. If you would like to learn more about how to implement the best practices of demand forecasting in your organization, please request a personal presentation.
June 28th, 2023 6 MIN READ
sdh release 2.0 main eng
New capabilities of SMART Demand Forecast. Release 2.0
A successful supply chain starts with accurate demand forecasting. Supply chain managers constantly challenge themselves to determine the optimal order quantity that meets customer needs, helps to avoid overstocking and write-offs, and ensures a high level of service. According to McKinsey & Company, companies can increase marginal returns by an average of 2-5% and reduce inventory levels by 20-25% by improving forecast accuracy by an average of 10-20%. To ensure high quality forecasting, solutions like SMART Demand Forecast are used, which are based on machine learning and artificial intelligence algorithms. The SMART business team works daily to improve the functionality of the demand forecasting system to provide you with the best user experience. In this update, we introduced a clustering algorithm for stores and goods for a convenient selection of analogs, modified the promo activity forecasting processes, added a new forecasting model, TFT (Temporal Fusion Transformation) and much more. Read more about all system updates in the article. A TFT (Temporal Fusion Transformation) prediction model has been added to the advanced subscription. The implemented TFT (Temporal Fusion Transformation) forecasting model will allow to obtain a higher accuracy of the forecast for long-term periods, which in turn will have a significant impact on the performance of the customer’s business. This will allow your business to be more reactive and adaptive to changing market conditions. Clustering algorithm for stores and goods for the selection of analogs has been implemented. The clustering algorithm for store and goods analogs in the system saves time and effort for users, providing them with convenient functionality for quickly finding and comparing the best options. Now you can group similar products and stores into clusters and find similar ones by sales patterns and characteristics. DevOps developments have been synchronized and updated. The priority of the team is to provide customers with high-quality and stable operation of the system, as well as reducing the influence of the human factor on the operation of the solution, therefore DevOps processes are an integral part of product quality control. This implemented development will help to improve the system and implement the software more efficiently and faster. A unified approach to forecasting at higher levels of aggregation by date, business and product levels has been implemented. A unified approach for building demand forecasts at higher levels of aggregation has been introduced in SMART Demand Forecast. The update simplifies and standardizes the forecasting process. In turn, this helps to improve the quality of the forecast and directly affects the planning and management decision-making processes. PoC for the approach to forecasting demand for new products using Bass diffusion has been conducted. When working with new products and stores, forecasting accuracy plays an important role and contributes to building a profitable strategy for their development. The implemented PoC approach using Bass diffusion allows the system to make a qualitative forecast and use this information for strategic planning. Optimization work and parallelization of the scoring process were performed, which made it possible to speed up forecast preparation by 2.5 times. A parallel scoring process has been implemented in SMART Demand Forecast. This means that all system tasks are divided into smaller subtasks that can be executed simultaneously on different computing resources. The update was implemented by distributing work between different processors, as well as using parallel computing algorithms. This will allow users of the system to perform more computations simultaneously and quickly receive accurate and efficient solutions that improve the competitiveness and bottom line of the business. PoC work for the functionality of choosing the optimal set of complex scenarios has been conducted. The ability to analyze complex combinations of promotional campaigns, events and parameters that can affect the demand forecast built by the system has been added. Users can choose the best scenario and make more accurate management decisions, use resources efficiently and optimize processes in the system. All of these combine to improve productivity and reduce costs. Modification of promo activity forecasting processes has been performed. The ability to forecast more than one promotion at one time level has been added to SMART Demand Forecast, which provides system users with greater flexibility in introducing promotion activity variations into the system, assessing their impact on demand and, accordingly, additional opportunities to improve the company’s marketing activities. Technical and process project documentation has been added. The Environment Policy document has been drawn up. The availability of documentation is an important aspect that directly affects the work of users in the system, as well as the speed and quality of development. That is why in this release the technical and process project documentation has been added and the Environment Policy has been drawn up. This is important because: First, technical and process documentation helps users to better understand the system, its functionality, rules of use and related processes. This contributes to more efficient use of the system and reduces misunderstandings. Secondly, the availability of technical documentation simplifies the development of new system functions, modification and extensions. Developers have access to important information about the architecture and components of the system, which makes their work easier. Thirdly, the availability of documentation helps the system comply with regulatory requirements and standards, which can be important when working with partners or customers who have their own requirements for the environment and standards for its protection. Report localization in Power BI has been implemented. SMART Demand Forecast is now configured to automatically adapt reports to match the language requirements of users. The use of language packages allows analytics to be more understandable and usable for different audiences. The Data Health Check report has been created. Quick detection of critical errors or data incompleteness is now possible with the new Data Health Check report. Users can get a visual overview of the state of their data and timely identify and fix issues related to the accuracy and reliability of the data used by the system in the process of training a mathematical model and preparing a forecast. For more information about SMART business solutions and services, please call +38 (044) 585-35-50 or submit your request.
June 26th, 2023 3 MIN READ
df 1200x630 3
Accurate demand forecasts for right business decisions
Consumer demand is a volatile factor, especially in today’s growing uncertainty. The behavior of buyers is influenced by a lot of factors, sometimes non-obvious and unexpected ones: from the season and market trends to the activities of competitors and the political situation. At the same time, sales efficiency directly depends on the accuracy of demand forecasting. Miscalculations in this matter can lead to a number of unpleasant consequences. Overestimation of demand can lead to additional costs for storing excess goods in warehouses and to product write-offs. Underestimation results in a shortage of goods on the shelves, problems with logistics and suppliers, loss of part of the profit, as well as decreases the level of service and customer loyalty. In order to avoid these hassles in a highly volatile business environment, it is critical to have a tool for forecasting as accurately as possible. SMART Demand Forecast is exactly such a tool. It works on the basis of machine learning and artificial intelligence and is able to take into account and process factors that affect demand from sales history to external factors. Thus, it enables extremely accurate operational planning in uncertain conditions. In this video, you can see how the tool works. As you can see, it consists of three functional blocks: 1. Work with analogs This block is designed to work with goods or stores that do not have enough history and data for further forecasting. In this case, the forecast is based on the history of analogs. The system automatically generates a list of analogs, or analogs can be configured manually. 2. Modelling The functionality of the block is designed to start or restart the process of forecasting regular and promotional sales, exporting forecast results, as well as managing promotional campaigns. All the necessary parameters are pre-configured, including the forecast horizon and period, points of sale, products, data on promotions, promotional campaigns, etc. 3. Analytics This block allows you to analyze the history of sales, the quality of promo sets related to forecasting and compensated sales. The use of the tool allows companies to accurately anticipate fluctuations in demand and adjust product strategies in accordance with them. And thus always maintain optimal availability of goods, fully satisfying demand with minimal logistics and storage costs. With this tool at hand, executives can make quick, informed decisions about sourcing, pricing, campaigns and promotions, maximizing sales and reducing operating costs. Find out more about how SMART Demand Forecast can help you improve your business efficiency – book a personalized presentation.
January 27th, 2023 5 MIN READ
obkladinka en
All about the SMART Demand Forecast System Update

The functioning of the entire supply chain begins with demand forecasting. How much to order to satisfy consumers, and how to avoid freezing inventory, prevent write-offs, and ensure proper customer service are questions that supply chain managers ask themselves every day when building forecasts for each SKU. To ensure their high accuracy, there are solutions such as SMART Demand Forecast, which are based on machine learning and artificial intelligence algorithms.

The SMART business team works daily to improve the SMART Demand Forecast functionality. Our focus is always on improving the financial performance of our customers’ business and providing the best user experience, so in this release, we have developed the table personalization functionality, optimized the promotional campaign import process and cross-browser compatibility, accelerated the interface and Data Science processes, and implemented the technical system optimization. Read the article to learn more about all system updates.

Product documentation has been updated

The Security and Environment Policy documents, that describe the operation and interaction of environments, standards and approaches to the security component of the product, have been finalized. The documents regulate the interaction of SMART Demand Forecast with the data of our customers.

The project management structure has been updated. We are meticulous in creating and keeping product technical documentation up to date in order to provide customers with the opportunity to familiarize themselves with documents regarding business processes, solution architecture, terminology, etc.

Table personalization functionality has been developed

The SMART business team has added table personalization functionality with the ability to select the number of displayed rows, customize the list and column sequence.

This customization allows users to change the appearance of the tables while working with the demand forecasting system, which in turn greatly improves the user experience.

In other words, if the nuances of your work require customization of the interface, you can move or hide the information as needed.

The process of importing promo campaigns has been optimized and the loading speed has been increased

We always care about ensuring fast and reliable operation of the system. Therefore, we have worked hard to reduce the waiting time when importing promo campaigns and their subsequent validation in SMART Demand Forecast.

Business rules and process of basic Data Health Check have been developed

When integrating data into the system, it will automatically check the downloaded data for critical errors that will interfere with the main function of the solution: preparing a forecast. In this release, we have added a check for:

  • metadata
  • data integrity
  • compliance with the reference structure

System interface performance has been improved

To prevent repeated calculations, saving the results of executed functions has been implemented in the system. Thus, there was a significant increase in interface performance when loading system pages.

Cross-browser compatibility has been ensured

The correct display of the interface and work of the system functionality, regardless of the installed browser, has been set up.

The functionality of forecasting to the higher aggregation levels has been developed

Previously, the processes for forecasting at the lowest level of aggregation were implemented in SMART Demand Forecast. To cover a wider range of business tasks, the team worked out approaches to generating a model that allows forecasting at higher levels of time, product and business aggregation. Weekly and monthly aggregation levels have already been implemented in the system.

The concept of Time Series forecasting by models has been validated

The system works with one model for forecasting, LGBM. To develop a multi-model approach for different business requirements, a Deep Learning model has been developed for time series forecasting. Work on the implementation of this concept into the system will be included in release 2.0.

Accelerated Data Science processes In release 1.2, a number of frameworks and approaches were tested to reduce the execution time of data science processes, which will affect the cost of commercial operation of the SMART Demand Forecast system. The process of parallel scoring of several complex scenarios was considered, which in turn will speed up the provision of user forecast results.

The system has been technically optimized

An important aspect of the IT solution is to provide users with stable and fast system operation, given the constant processing and analysis of large amounts of data, so the SMART business team carried out a technical optimization of SMART Demand Forecast, including:

  • Adaptation of ICacheValidator methods
  • Development of the process of incremental loading of data into the system database
  • Testing and modification of measurement unit conversion processes.
  • Optimization of performance for executing stored procedures in the application’s database
  • Automation of the Feature Selection module operation

The factor selection module has been optimized according to the input conditions. Two running modes have appeared: manual and automatic. Manual mode is required for the pilot stage and finer tuning of the solution for the customer. Automatic mode is necessary for comprehensive use with minimal human time spending.

  • DS processes code has been refactored

The SMART business team refactored and optimized the step of generating a planned data set and importing the library in Databricks notebooks, and also reduced the size of the Docker image. Thanks to this, the code became cleaner and more understandable.

For more information about SMART business solutions and services, please call +38 (044) 585-35-50, or submit your request here.

Order a presentation to increase
forecast accuracy as soon as tomorrow!

      Thank you for your inquiry
      we will contact you shortly