Automating Demand Planning for high prediction accuracy is all the buzz these days! While you want to make the best use of technology by searching for the right demand planning software, you should also learn everything relevant before you make the move. In this blog, we attempted to address the most frequently asked questions from our customers as they contemplate automating demand planning.

‍What is the difference between demand sensing and demand forecasting?

‍Demand Forecasting is a theoretical and statistical way of predicting consumer demand based on historical sales data. Typically, forecasts are done over 18-24 month cycles but may vary based on product category and industry.

Demand Sensing on the other hand is the new age method of utilizing AI/ML, IoT, and other advanced technologies to provide more accurate, real-time demand predictions. Demand Sensing takes Demand forecasting to the next level by considering short-term trends, incorporating baseline data adjusted for external and internal factors to create an ever-evolving real-time demand forecast. Demand Sensing provides higher accuracy compared to demand forecasting by leveraging a high level of data granularity to analyze daily demand information as close as possible to the end customer and immediately detect changes in demand behavior.

‍
Which factors do you include to achieve accurate demand forecasts?

Kronoscope follows a bottom up approach to sense demand changes at the most granular (SKU) level to achieve unmatched demand prediction accuracy. These predictions are made from supply chain nodes that are closer to the customer and rolled up to the top nodes like the mother warehouse.

The Baseline predictions deal with SKU, Channel, Location level demand predictions which take into account Historical sales trends, Seasonality, Cyclicity and Outlier Corrections. In the next stage, the Factor adjusted demand predictions are made by including factors like Changes in pricing, Marketing/Promotional Campaigns, Holiday & Weather. The system analyzes the impact of these factors on demand and adjusts the baseline predictions accordingly. These prediction numbers can still be edited/changed by users based on their judgment or business knowledge. The updated prediction numbers are then finally pushed to the Inventory Planning module.

How do you predict demand for newly launched products?

Predicting demand accurately and planning inventory for newly launched products or stores can get tricky as there’s no historical data available. Our solution simplifies this by intelligently associating new products or new stores with existing ones that have similar attributes. Eventually, the system learns continuously and trains itself to predict demand and plan inventory for new products/stores independently.

How do you help measure the impact of promotions on demand uplift?

The Promotions Planning feature of Kronoscope allows you to systematically record any promotions or marketing events that you are planning to run. You can create the event thereby specifying the event type, duration and event segment. This helps in capturing the real time impact of promotions on demand levels. The system then learns from how demand looks during these events. This is then factored in to predict the promotional uplift in demand when similar promotions/marketing events happen in the future. This way, demand forecasting and inventory planning for promotions becomes much easier and more systematic.

Solving Promotional Uplift Challenges With Kronoscope →

We don’t change prices very often, how will you be able to establish price sensitivity?  

There could be few SKUs for which you may not have experimented with prices before and you would want to do it now. There could also be new SKUs that don’t have a history of price points. In this it might be confusing to set a new price. In this case we establish price sensitivity by using dimension hierarchy. We take into account the price elasticity of a cluster of products that have similar attributes. The system intelligently learns from similar SKUs and recommends optimum price points. This way, it becomes easier to run price discounts to meet revenue targets even for SKUs whose prices haven’t been changed in the past.

How do we establish price sensitivity using Dimension Hierarchy?

For example our dimension hierarchy is [SKU_ID, Sub_Category] if we have a SKU Active White Toothpaste in Branch - Los Angeles and we have never changed the price of it. The system checks price sensitivity at SKU_ID level across all the branches. If price changes are not available at SKU_ID level as well then it will check at Sub category level (at all toothpaste level) and apply that sensitivity on the current SKU - Branch combination.
‍

Measuring The Impact Of Strategic Pricing with the help of Price Elasticity and Price Optimization →

‍

Automating Demand Planning for high prediction accuracy is all the buzz these days! While you want to make the best use of technology by searching for the right demand planning software, you should also learn everything relevant before you make the move. In this blog, we attempted to address the most frequently asked questions from our customers as they contemplate automating demand planning.

‍What is the difference between demand sensing and demand forecasting?

‍Demand Forecasting is a theoretical and statistical way of predicting consumer demand based on historical sales data. Typically, forecasts are done over 18-24 month cycles but may vary based on product category and industry.

Demand Sensing on the other hand is the new age method of utilizing AI/ML, IoT, and other advanced technologies to provide more accurate, real-time demand predictions. Demand Sensing takes Demand forecasting to the next level by considering short-term trends, incorporating baseline data adjusted for external and internal factors to create an ever-evolving real-time demand forecast. Demand Sensing provides higher accuracy compared to demand forecasting by leveraging a high level of data granularity to analyze daily demand information as close as possible to the end customer and immediately detect changes in demand behavior.

‍
Which factors do you include to achieve accurate demand forecasts?

Kronoscope follows a bottom up approach to sense demand changes at the most granular (SKU) level to achieve unmatched demand prediction accuracy. These predictions are made from supply chain nodes that are closer to the customer and rolled up to the top nodes like the mother warehouse.

The Baseline predictions deal with SKU, Channel, Location level demand predictions which take into account Historical sales trends, Seasonality, Cyclicity and Outlier Corrections. In the next stage, the Factor adjusted demand predictions are made by including factors like Changes in pricing, Marketing/Promotional Campaigns, Holiday & Weather. The system analyzes the impact of these factors on demand and adjusts the baseline predictions accordingly. These prediction numbers can still be edited/changed by users based on their judgment or business knowledge. The updated prediction numbers are then finally pushed to the Inventory Planning module.

How do you predict demand for newly launched products?

Predicting demand accurately and planning inventory for newly launched products or stores can get tricky as there’s no historical data available. Our solution simplifies this by intelligently associating new products or new stores with existing ones that have similar attributes. Eventually, the system learns continuously and trains itself to predict demand and plan inventory for new products/stores independently.

How do you help measure the impact of promotions on demand uplift?

The Promotions Planning feature of Kronoscope allows you to systematically record any promotions or marketing events that you are planning to run. You can create the event thereby specifying the event type, duration and event segment. This helps in capturing the real time impact of promotions on demand levels. The system then learns from how demand looks during these events. This is then factored in to predict the promotional uplift in demand when similar promotions/marketing events happen in the future. This way, demand forecasting and inventory planning for promotions becomes much easier and more systematic.

Solving Promotional Uplift Challenges With Kronoscope →

We don’t change prices very often, how will you be able to establish price sensitivity?  

There could be few SKUs for which you may not have experimented with prices before and you would want to do it now. There could also be new SKUs that don’t have a history of price points. In this it might be confusing to set a new price. In this case we establish price sensitivity by using dimension hierarchy. We take into account the price elasticity of a cluster of products that have similar attributes. The system intelligently learns from similar SKUs and recommends optimum price points. This way, it becomes easier to run price discounts to meet revenue targets even for SKUs whose prices haven’t been changed in the past.

How do we establish price sensitivity using Dimension Hierarchy?

For example our dimension hierarchy is [SKU_ID, Sub_Category] if we have a SKU Active White Toothpaste in Branch - Los Angeles and we have never changed the price of it. The system checks price sensitivity at SKU_ID level across all the branches. If price changes are not available at SKU_ID level as well then it will check at Sub category level (at all toothpaste level) and apply that sensitivity on the current SKU - Branch combination.
‍

Measuring The Impact Of Strategic Pricing with the help of Price Elasticity and Price Optimization →

‍

Automating Demand Planning for high prediction accuracy is all the buzz these days! While you want to make the best use of technology by searching for the right demand planning software, you should also learn everything relevant before you make the move. In this blog, we attempted to address the most frequently asked questions from our customers as they contemplate automating demand planning.

‍What is the difference between demand sensing and demand forecasting?

‍Demand Forecasting is a theoretical and statistical way of predicting consumer demand based on historical sales data. Typically, forecasts are done over 18-24 month cycles but may vary based on product category and industry.

Demand Sensing on the other hand is the new age method of utilizing AI/ML, IoT, and other advanced technologies to provide more accurate, real-time demand predictions. Demand Sensing takes Demand forecasting to the next level by considering short-term trends, incorporating baseline data adjusted for external and internal factors to create an ever-evolving real-time demand forecast. Demand Sensing provides higher accuracy compared to demand forecasting by leveraging a high level of data granularity to analyze daily demand information as close as possible to the end customer and immediately detect changes in demand behavior.

‍
Which factors do you include to achieve accurate demand forecasts?

Kronoscope follows a bottom up approach to sense demand changes at the most granular (SKU) level to achieve unmatched demand prediction accuracy. These predictions are made from supply chain nodes that are closer to the customer and rolled up to the top nodes like the mother warehouse.

The Baseline predictions deal with SKU, Channel, Location level demand predictions which take into account Historical sales trends, Seasonality, Cyclicity and Outlier Corrections. In the next stage, the Factor adjusted demand predictions are made by including factors like Changes in pricing, Marketing/Promotional Campaigns, Holiday & Weather. The system analyzes the impact of these factors on demand and adjusts the baseline predictions accordingly. These prediction numbers can still be edited/changed by users based on their judgment or business knowledge. The updated prediction numbers are then finally pushed to the Inventory Planning module.

How do you predict demand for newly launched products?

Predicting demand accurately and planning inventory for newly launched products or stores can get tricky as there’s no historical data available. Our solution simplifies this by intelligently associating new products or new stores with existing ones that have similar attributes. Eventually, the system learns continuously and trains itself to predict demand and plan inventory for new products/stores independently.

How do you help measure the impact of promotions on demand uplift?

The Promotions Planning feature of Kronoscope allows you to systematically record any promotions or marketing events that you are planning to run. You can create the event thereby specifying the event type, duration and event segment. This helps in capturing the real time impact of promotions on demand levels. The system then learns from how demand looks during these events. This is then factored in to predict the promotional uplift in demand when similar promotions/marketing events happen in the future. This way, demand forecasting and inventory planning for promotions becomes much easier and more systematic.

Solving Promotional Uplift Challenges With Kronoscope →

We don’t change prices very often, how will you be able to establish price sensitivity?  

There could be few SKUs for which you may not have experimented with prices before and you would want to do it now. There could also be new SKUs that don’t have a history of price points. In this it might be confusing to set a new price. In this case we establish price sensitivity by using dimension hierarchy. We take into account the price elasticity of a cluster of products that have similar attributes. The system intelligently learns from similar SKUs and recommends optimum price points. This way, it becomes easier to run price discounts to meet revenue targets even for SKUs whose prices haven’t been changed in the past.

How do we establish price sensitivity using Dimension Hierarchy?

For example our dimension hierarchy is [SKU_ID, Sub_Category] if we have a SKU Active White Toothpaste in Branch - Los Angeles and we have never changed the price of it. The system checks price sensitivity at SKU_ID level across all the branches. If price changes are not available at SKU_ID level as well then it will check at Sub category level (at all toothpaste level) and apply that sensitivity on the current SKU - Branch combination.
‍

Measuring The Impact Of Strategic Pricing with the help of Price Elasticity and Price Optimization →

‍