Forecasts

Analyze

Description & Overview

Open

Page "Analyze" is designed for building forecast models and in-depth study of forecast influencing factors. The page includes two elements:

Open

• "Parameters" area is designed to select and configure the basic parameters for the forecast calculation;

• "Results" area is designed to show the forecast results, factors that affecting the forecasts, as well as to study their nature.

How we build forecast

MySales has the most technologically and mathematically advanced system core (engine), tested on hundreds of millions different positions and stores, which is responsible for building a stable quality forecast for each position in each store.

The forecast is built at different levels, starting with the most aggregated levels, ending with the most detailed:

• Product group - all stores

• Product group - region

• Product group - store

• SKU - all stores

• SKU - region

• SKU - store

Forecasting is an automatic, multi-stage, iterative process that has a hierarchy and works with a huge data array. When building a forecast, all available history is analyzed at the high levels, up to the last 3-4 years of sales. This is a huge amount of data. Just imagine. If you have, for example 100 stores and 10,000 SKUs, then this is one million combinations multiplied by 3-4 years of sales history. At each recalculation of all forecasts, MySales analyzes gigabytes of information for a small retail chains and terabytes of information for chains with hundreds of stores.

Describing the MySales forecast calculation algorithm, we would like to warn you against trying to implement such an algorithm yourself. Even if you manage to implement such an algorithm in a reasonable amount of time, it will take years to test it on millions of different positions, catching errors, and also spend a lot of effort on optimizing calculation performance to bring it to an acceptable level before you can get economic benefits from using it.

We can distinguish the following main stages of building the MySales forecast, which system performs for each combination at each level:

• Downloading and preparing data from the DBMS, as well as preparing the MySales file storage, which is used to optimize the forecast calculation speed, as well as to reduce the load on the DBMS when processing huge amounts of data. Not only historical sales, prices, discounts, balances, arrivals are loaded, but also the product hierarchy, product directory, geographic hierarchy, store directory and external data (weather, macroeconomics, competitor prices)

• Inclusion of data for analogues by stores or by position. At the same time, the system always sees the latest sales data for a new store or new position and uses analogues data only for those periods where there is no information about the new product

• Formation of data packages for forecast calculation in a multiprocessor environment using multiple (usually from 3 to 9) threads

• Primary cleaning of sales data from key influencing factors to calculate trend and seasonality. This stage is important in order to separate the impact of prices, discounts and promotions on sales and correctly calculate seasonal uplifts. Considering that in retail the amount of historical data for many SKUs is rarely large enough to exclude periods where the influence of discounts or prices took place due to the constant rotation of the assortment. So, MySales, instead of excluding it, cleaning forecast from the influence of these factors in order not to reduce the already limited history sales, excluding only the most extreme periods. Also, at all stages, when analyzing sales, MySales excludes periods with significant sales losses due to the fact that the product was absent or it was not enough on stock to ensure sales. This approach gives the advantage that, in fact, the trend and seasonality are calculated correctly even if the sales history is limited to only one year. And missing periods may be filled from higher forecasting levels

• Seasonality calculation. MySales considers both multiplicative (seasonal coefficient) and additive seasonality. This is necessary choose the most optimal method of seasonality application after analyzing whether the volatility of the seasonal periods increases along with the trend (general growth or decline in sales). Moreover, for the product groups level, seasonality of the average price of the group is also calculated, which is necessary to predict the average price of the group over a long horizon, since seasonal peaks usually have a higher average price in the group.

• Calculation of the trend of overall growth or decline in sales. It is also carried out after cleaning sales from the main influencing factors (price, discount, promotional), as well as after cleaning from seasonality. The system also calculates average sales and median using cleared sales

• Filling in the original matrix of historical values of predictors (influencing factors). A number of predictors are calculated values, for example, the ratio of the price for the current period to the average price for previous periods

• Analysis how sales depends on influencing factors (price, discount, weather, macroeconomic factor, which most often acts as the exchange rate, cannibalization, etc.). When calculating each influencing factor, the system selectively cleans it of other, most significant influencing factors, for example, of promo and seasonality. When analyzing price dependence, the system also analyzes the trend of inflation or deflation in order to clear historical prices from such an impact. A number of factors are analyzed separately for the low, high and medium seasons, so that price elasticity, for example for ice cream, differs for winter and summer

• If it is not possible to determine the dependencies of the influencing factors at detailed levels, then the system takes key ones (for example, the influence of price, seasonality) from higher levels. For example, for the SKU-store level, such dependencies can be taken from the SKU-region or SKU-all stores level, from the group-store, group-region, or group-all stores levels

• Filling the target matrix of future predictor values. Here, a differentiated approach is used, which for some factors can be a simple average or median, for some - predicted values, and for some - information that was entered by the user in the customer’s system and loaded into the MySales data warehouse. It is recommended that you always load into MySales the assortment matrix, future prices as soon as they become known and enter the data about promotional activities

• Calculation of correlation coefficients and formation of an automatic forecasting model based on these coefficients using those factors that affect sales

• Testing automatic, as well as pre-configured forecast models. In this case, the predictive model is understood as a set of influencing factors.

• Choosing the model that gives the best accuracy on past sales. When evaluating each model, periods where the product was not enough to ensure sales, as well as promotional periods that have the greatest impact on sales, are excluded

• After the best model is selected, MySales builds forecast for future periods

• A forecast is also carried out with a number of actions aimed at ensuring its stability and reliability, for example, calculating the minimum sales values, calculating the autocorrelation of the model error, in order to adjust the forecast for the coming weeks and make it more accurate

• The next stage is the calculation of promo uplifts. At this stage, the system applies the promotion coefficients of uplift to the forecast generated by a set of neural networks (Dusya), or data from comparable promos, if such was found. Promo uplifts are adapted to the individual characteristics of each predicted position in each store to take into account the individual characteristics of different combinations and their sensitivity to promotional factors. The minimum and maximum limits of the promotion uplifts are also calculated to ensure a stable result for new positions where the history of the promotion is not enough

• After calculating promo uplifts, the system performs rebalancing of the model taking into account all the influencing factors. This is necessary to balance the influence of the price and the promo effect in the promo forecast, because they often have a high correlation

• At this stage, the forecast is ready, now we calculate the safety stock. SS is calculated as the standard deviation of the forecast from sales in previous periods. SS is also, as in the example above with ice cream, different in summer and winter, so the forecast is divided into 3 ranges: high, low and medium and safety stock is calculated separately for each range

• Further, the forecast is used to calculate price recommendations so that you can determine the most optimal price for sales in value or margin

• The final step is to calculate the possible economic effect in the past, in the form of increased sales, or sales losses for periods in which influencing factors were not known to the system, as well as a stocks reduction

It is also worth noting that the system has a separate algorithm and calculation sequence for new positions that do not have an analogue: for such positions, the system uses the sales of the average position in the product group, adjusting them to the price elasticity of the group using the specific price of the new item, and also applying a number of other restrictions and calculations to make the result more stable and accurate.

There is also a separate forecasting algorithm for expanding distribution. Example: a position was sold in 10 of 30 stores in the region, given the good dynamics and potential, the category manager decided to list it in all 30 stores in the region. In this case, the system uses sales dynamics at the level of SKU-region, group-region and group-store to build forecast for this position in a store where there is no sales history and this position is being listed for the first time.

The set of all factors that the system takes into account is set by the user and can be individually adapted. The default set of factors is described in the "Default Predictors" section.

Daily forecast adjustment

After completing all forecast calculation stages, the system adjusts it for the upcoming weeks (usually two, which is a configurable parameter) based on the most recent days of sales data. To do this, the system loads sales data at the SKU-store-day level for the last 28 days (configurable parameter) and selects the most relevant 7 (or fewer, also configurable) days of sales:

• Separating promotional and non-promotional sales. Past promotional sales adjust only the promotional forecast for the next one to two weeks, while regular sales adjust the regular forecast accordingly.

• Filtering out days with insufficient stock. The stock level must be greater than zero and also exceed 30% of the average daily sales (configurable parameter). In some cases, a minimum absolute stock level can be set as a parameter.

• Handling cases of insufficient stock. If sales occurred despite low stock, the system may still consider them under certain conditions. For example, if a given day’s sales exceeded 75% of the forecast for that day, the system applies an upward adjustment factor. The assumption is that if stock had been sufficient, sales would have been higher:

• By 50%, if the product likely ran out in the second half of the day.

• By 25%, if the product likely ran out at the end of the day.

• Filtering out days with significant price variations. Days where the price deviation exceeded 15% (configurable parameter) are excluded.

Calculating lost sales

For days when an item was part of the assortment matrix in the last 1–2 weeks (configurable parameter), the system calculates lost sales or missed demand for days when there was no stock or an insufficient stock level.

Similar to the Daily Forecast Adjustment calculation, insufficient stock means that either:

• The stock level was zero at the end of the day, or

• The stock was less than 30% (configurable parameter) of the forecasted sales for that day.

When distributing the weekly forecast into daily values, the system also applies day-of-week coefficients and considers price elasticity if the price changes during the week.

If the system finds valid sales data from the last 28 days (configurable parameter) that meet the required criteria (i.e., sufficient stock was available, promotional and regular sales are considered separately, and price fluctuations were within acceptable limits), it calculates the average daily sales. This calculation adjusts the identified 7 or fewer (configurable parameter) sales days using the same factors applied in the daily forecast adjustment:

• Day-of-week coefficients

• Price elasticity and discounts

• Seasonality, weather, and other influencing factors

During the adjustment of average daily sales, the system takes into account the seasonality and weather conditions of the week in which the stock shortage occurred, as well as the day-of-week coefficient, price, and discount on the specific date for which the calculation is performed.

As a result, the system generates an adjusted forecast for the specific date.

Lost sales are calculated using the formula:

Lost Sales = max(Forecast} - Actual Sales, 0)

This applies only if stock was insufficient according to the criteria described above.

Lost sales calculated during the forecast process are stored in the FCST_SKU_LOST table, linked to the forecast ID for the last 2 weeks (configurable parameter). Based on this table, the SALES_SKU_LOST table is updated, maintaining a long-term history of lost sales at the SKU-store-day level. Additionally, data may also be aggregated and stored in the SALES_SKU_LOST_WEEK table, which contains weekly-level lost sales information.

Next, the system adjusts daily sales based on influencing factors, aligning past data with future forecast weeks:

• Day-of-week coefficients

• Price elasticity and discounts

• Seasonality, weather, and other factors

Finally, for each future week—separately for promotional and regular sales—the system calculates adjusted average daily sales. The forecast is then fine-tuned within acceptable limits, which depend on the number of selected days and can be defined via parameters.

Parameters

"Parameters" area is designed to choose and configure the basic parameters of the forecast model. It consists of the following categories:

Open

• Stores

• Groups

• SKU

Data selection happens in a next way:

• button  opens the drop down list for selecting data. It shows previously used data, and has an option to select new data when you click on "select...";

Open

• button  allows you to select all the data for relevant category;

• кнопка  позволяет не строить прогноз для отдельных SKU, а строить только на уровне Группы.

Open

Button "Settings" opens additional parameters for configure the forecast (to hide additional parameters, you need to click this button again):

Open

• parameter "End weeek" – is responsible for choosing the last week for which the system will see historical data, even if there is historical data after chosen week (last week of the training sample). Used to verify the constructed forecast;

• parameter "Profile" – selection the type of forecast:

  • "52-weeks" – 52 weeks forecast calculation;

  • "52-weeks log" – 52 weeks forecast calculation with a detailed report on the process of building a forecast model (the report can be seen in the tab "Models");

  • "26-weeks" – 26 weeks forecast calculation ;

  • "Brain 1000" – forecast calculation using a neural network in 1000 iterations;

  • "Brain 3000" – forecast calculation using a neural network in 3000 iterations;

• parameter "Supplier" – supplier selector. The supplier selection is the same as selection store / group.

When you press "run" the forecast calculation begins. Detailed information about the forecast calculation progress is displayed under the progress bar. At the end of the forecast calculation, audible warning will occur.

Open

Results

"Results" area is designed to display the forecast results and the factors that affects them, as well as for studying their nature.

Open

"Results" area consists of two components:

• analysis area – designed for in-depth study of the forecast model, predictors and their influence on the forecast;

Open

• results display area – designed to display the results of the forecast and the values of the parameters.

Open

Button "More" opens additional displaying parameters of the forecast results (to hide additional parameters, you need to press the button "Less"):

Open

You can setup results displaying in next ways:

• change the displaying of forecast date to weekly/monthly;

Open

• select forecast range to display (from the start date to the end date);

Open

• сhange the display of the forecast sales by sales volume or by sales value.

Open

To apply the changes, click "Show".

Additional parameters tab has the pane that allows you to work with files:

Open

• To load previously calculated model to analyze it you have to select file in field "Select a file…" field, and then press load button  ;

• Для сохранения построенной модель прогноза в формате json (для возможности ее загрузки в будущем) нужно нажать кнопку  ;

• save and download the calculated forecast and the configuration matrix (forecast parameters) in csv format press  .

Analysis area

Analysis area is one of the "Results" area components, it is designed for in-depth study of forecast model, its components and their influence on forecast. Area is used to manually analyze the forecast by SKU, to understand forecast numbers, to analyze the factors that affect it. This is a very important part of the software, since it allows the user to get complete information about the influence of various factors on the forecast, to understand how sales change with price changes, and also to see in numerical terms how different factors will change sales.

Analysis area is represented by header and elements panel.

Header displays:

Open

• Group Name / SKU;

• Stores count / Store name;

• – button for downloading results into csv file.

Elements panel allows you to get detailed information about all positions and consists of next elements:

– drop-down list, which elements allow analyzing sales and forecasting model by SKU. It consists of:

Open

• SKUs chart – visualizes "Components" element in the form of a chart for all SKUs (the user himself selects which component to display on the chart). Details on all components are described in the Forecast analysis components guide;

Open

• SKUs table – visualize "Components' element values as a table for all SKU (the user himself selects which component to display on the chart);

Open

• SKUs list – allow selecting for analysis forecast model for a particular SKU. The forecast model for the selected SKU is displayed as a separate area under the parent area (in list, SKU element disappears);

Open

Stores – drop-down list, which elements allow analyzing sales and forecasting model by Store. It consists of:

Open

• Stores chart – visualizes "Components" element in the form of a chart for all stores (the user himself selects which component to display on the chart). Details on all components are described in the Forecast analysis components guide;

Open

• Stores table – visualize "Components' element values as a table for all Stores (the user himself selects which component to display on the chart);

Open

• Stores list – allows selecting for analysis forecast model for a particular Store. The forecast model for the selected Store is displayed as a separate area under the parent area (in list, Store element disappears);

Open

Components – Allows you to add a new component, parameter, result or characteristic of the calculated forecast model to the results display area. Different components display either their influence on the forecast in sales volume or value, and also display historical values of factors that affect the forecast. The use of components allows a more detailed study of the influence of factors on the forecast, as well as understanding why the forecast has one or another value. Anomaly sales and excluded weeks are also added using components (described in detail in the Anomaly & Excluded section). After being added, the component disappears from the list; The component is returned to the list after it is removed from the display area. For more detailed information about the "Components" elements, use the Forecast analysis components guide; section;

Open

Forecast – designed for visual analysis of the forecast results on the chart. The chart allows you to visually estimate built forecast, as well as using components to understand the forecast behavior (for example, using the stock component, you can see changes in the stocks balance and understand that, for example, sales fell precisely because of shortages);

Open

By default chart displays historical sales, forecast, and historical sales shifted by 52 weeks to the right (for visual forecast estimation) Also chart displays selected components. The user can turn off the display of any element by clicking on its name under the chart.

Open

Predictors – designed for visualization and assessment of the impact of predictors on the predictive model (those predictors that were included in the model are displayed). Allows you to analyze, because of what parameters the forecast has one or another value. For more information on predictors, see Predictors. This element is very important for understanding the forecast itself. It displays the numerical values ​​of the influence of various factors on the forecast and understand on what exactly the sales depend in different periods. User can turn off the display of any element by clicking on its name under the chart.

Open

Price – designed to visualize the chart of price elasticity of demand. This chart allows the user to see how demand changes when the price of a product changes using price elasticity. The system calculates average sales in volume, value and margin relative to different prices. The system also calculates theoretically optimal price values ​​to maximize sales in value and margin. The area around the chart line indicates the possible range of sales for a particular price value. It is built based on the forecast error, it is logical that the larger the error, the larger the area and the greater the range of predicted sales values.

Open

There are 3 interpretations of price elasticity (depending on what it is considered). To select, click on the required field; the current price value is indicated by a dot on the graph):

• Volume – influence on sales volume, purchase price – estimated purchase price;

Open

• Value – influence on sales value, purchase price – estimated purchase price, best sales price – price, that, theoretically gives the highest sales;

Open

• Margin – influence on sales margin, purchase price – estimated purchase price, best sales price – price, that, theoretically gives the highest margin.

Open

Trends – designed to visualize the influence of prices, discounts and weather predictors on sales on the chart with a range of error. It allows the user to see how changes in these factors will change sales.

Open

The necessary parameter is selected from the elements of the "Predictors" drop-down list; if the necessary parameter is not in the list, the influence of this factor is not observed.

Open

There are 2 interpretations of influence (current value is marked with a dot):

• Volume – influence on sales volume;

Open

• Value – influence on sales value.

Open

Models – designed to study and analyze forecast model, to study the influence of various predictors, to analyze reports on model calculation, to select another forecast model in case of disagreement with the model selected earlier.

Open

The element has 3 display variants:

• general – for the aggregated forecast model of stores and SKU, is displayed when the model tab is selected at the group or SKU level until individual stores are selected. Displays models for all selected stores, as well as for regions.

Open

• normal - displayed after selecting a specific store. It can be displayed for all SKUs as well as for selected ones (to select this parameter, you have to choose one particular store), it displays:

  • Forecast estimate - shows values for different forecast estimates;

  • Tested trial models - built models, their assessment, the impact of the predictor options previously specified in the file for each model. The model that the system has selected as the best based on the standard error is highlighted in green. The user can manually change the model by clicking next to the selected model in the "Fix"

  • Correlation Coefficients - coefficients of dependencies of different components on each other and on sales;

  • Trends - formulas for recovering the values of forecast elements;

Open

Open

• detailed – for detailed studying of the forecast calculation process model, it is displayed as a log of calculation process and works only in the "52-weeks log" mode (log automatically appears in Model area)

Open

Results display area

Results display area – is one of the "Results" area components, that is used to display forecast results, historical data values, recovered data, and other resulting components.

Open

By default, the following components are displayed:

• week in ISO format;

• historical sales volumes (or values) for the relevant week;

• sales volume (or values) forecast for the relevant week;

• historical sales volumes (or values), shifted 52 weeks ahead in time, for the relevant week.

Also, it is possible to add other components by using the analysis area component "Components"; for more detailed information about the components, use the "Forecast analysis components guide" section.

Forecast analysis components guide

Components of the forecast analysis (element "Components") are designed for detailed analysis and study of the calculated forecast model.

Components of the forecast analysis are divided into the following categories:

• volume – components that display the results of the forecast in the form of sales volumes;

• value – components that display the results of the forecast in the form of sales values;

• original – components that display historical data and intermediate forecast results (predictors that were used at any of the forecast stages) based on historical data;

• competitors - components that display competitors price for the selected SKU from competitors;

• regressed – сomponents that represents the influence of price factors, weather factors and the width of the assortment of the group;

• economy – components that represents the economic effect when correctly used by MySales;

• models – components that represents forecast when using alternative forecast models (not shown for aggregated models).

Category 'Volume'

Category volume consists of the following components:

• order – represents the sales forecast with safety stock (recommended results for ordering);

• seasonality – represents the influence of the seasonality on sales volume (taking into account the coefficient of influence on the forecast);

• trend – represents the influence of the trend on sales volume (taking into account the coefficient of influence on the forecast);

• region – represents the region sales influence on sales volume (taking into account the coefficient of influence on the forecast);

• group – represents the group sales influence on sales volume (taking into account the coefficient of influence on the forecast);

• auto error – represents the influence of unknown factors on sales volume (taking into account the coefficient of influence on the forecast). It is based on the avg error component, and the duration of the component effect can be 2, 4, or 6 weeks, depending on which duration gives the best result (min error);

• min – represents the influence of the sales conversion on sales volume, in the case of a forecast below the historically limiting volume (taking into account the coefficient of influence on the forecast). The limiting sales volume is the lower quartile of sales by default;

• price % – represents the price changing influence on sales volume (taking into account the coefficient of influence on the forecast);

• discount – represents the discount influence on sales volume (taking into account the coefficient of influence on the forecast);

• discount % – represents the discount (in percent) influence on sales volume (taking into account the coefficient of influence on the forecast);

• rate – represents the dollar rate influence on sales volume (taking into account the coefficient of influence on the forecast);

• a. price – represents the actual price (price with discount) influence on sales volume (taking into account the coefficient of influence on the forecast);

• price – represents the price influence on sales volume (taking into account the coefficient of influence on the forecast);

• item count. – represents the influence of the width of the available assortment (the width of the assortment on the stock) on the sales volumes (taking into account the coefficient of influence on the forecast);

• group disc – represents the influence of the average discount of all SKUs in the product group that the SKU belongs to (taking into account the coefficient of influence on the forecast);

• group price – represents the influence of the average price of all SKUs in the product group that the SKU belongs to (taking into account the coefficient of influence on the forecast);

• base – represents the base sales level (taking into account the coefficient of influence on the forecast);

• temp. – represents the average air temperature during the week influence on sales volume (taking into account the coefficient of influence on the forecast);

• rain – represents the proportion of raining days of the week influence on sales volume (taking into account the coefficient of influence on the forecast);

• snow – represents the proportion of snowing days of the week influence on sales volume (taking into account the coefficient of influence on the forecast);

• anomaly - displays the volume of anomaly sales, and also allows you to enter the value of anomaly sales in the opened field. Anomalies are described in detail in the section 'Anomaly & Excluded';

• исключения - displays excluded weeks, and also allows you to exclude weeks in the field that opens. Excludes are described in detail in the section: 'Anomaly & Excluded'.

If any components from the list are not displayed, this means that they were not included in the predictive model as predictors.

The sign ® or the prefix "reg" before the component means that this component is present in the recovered value (regressors).

Category 'Value'

Category value consists of the following components:

• sales – historical sales value;

• forecast – forecast of sales value;

• last year saels – historical sales value, shifted 52 weeks ahead in time;

• order – represents the sales forecast with safety stock (recommended results for ordering);

• seasonality – represents the influence of the seasonality on sales value (taking into account the coefficient of influence on the forecast);

• trend – represents the influence of the trend on sales value (taking into account the coefficient of influence on the forecast);

• region – represents the region sales influence on sales value (taking into account the coefficient of influence on the forecast);

• group – represents the group sales influence on sales value (taking into account the coefficient of influence on the forecast);

• auto error – represents the influence of unknown factors on sales value (taking into account the coefficient of influence on the forecast). It is based on the avg error component, and the duration of the component effect can be 2, 4, or 6 weeks, depending on which duration gives the best result (min error);

• min – represents the influence of the sales conversion on sales value, in the case of a forecast below the historically limiting volume (taking into account the coefficient of influence on the forecast). The limiting sales volume is the lower quartile of sales by default;

• price % – represents the price changing influence on sales value (taking into account the coefficient of influence on the forecast);

• discount – represents the discount influence on sales value (taking into account the coefficient of influence on the forecast);

• discount % – represents the discount (in percent) influence on sales value (taking into account the coefficient of influence on the forecast);

• rate – represents the dollar rate influence on sales value (taking into account the coefficient of influence on the forecast);

• a. price – represents the actual price (price with discount) influence on sales value (taking into account the coefficient of influence on the forecast);

• price – represents the price influence on sales value (taking into account the coefficient of influence on the forecast);

• item count. – represents the influence of the width of the available assortment (the width of the assortment on the stock) on the sales values (taking into account the coefficient of influence on the forecast);

• group disc. – represents the influence of the average discount of all SKUs in the product group that the SKU belongs to (taking into account the coefficient of influence on the forecast);

• group price – represents the influence of the average price of all SKUs in the product group that the SKU belongs to (taking into account the coefficient of influence on the forecast);

• base – represents the base sales level value (taking into account the coefficient of influence on the forecast);

• temp. – represents the average air temperature during the week influence on sales value (taking into account the coefficient of influence on the forecast);

• rain – represents the proportion of raining days of the week influence on sales value (taking into account the coefficient of influence on the forecast);

• snow – represents the proportion of snowing days of the week influence on sales value (taking into account the coefficient of influence on the forecast);

• anomaly - displays the value of anomaly sales, and also allows you to enter the value of anomaly sales in the opened field. Anomalies are described in detail in the section 'Anomaly & Excluded'.

If any components from the list are not displayed, this means that they were not included in the predictive model as predictors.

The sign ® or the prefix "reg" before the component means that this component is present in the recovered value (regressors).

Category 'Original'

Category original consists of the following components:

• seasonality – represents the seasonality of sales;

• trend – represents the trend influence on sales;

• a. price – represents the actual price (price with discount);

• price – represents the price;

• price % – represents how many times the price has changed;

• discount – represents the discount;

• discount % – represents the discount (in percent);

• stock – represents the goods balance on stock;

• group price – represents the average price of all SKUs in the product group that the SKU belongs to;

• group disc. – represents the average discount of all SKUs in the product group that the SKU belongs to;

• region – represents the region sales redestribution influence on sales;

• group – represents the group sales redestribution influence on sales;

• rate – represents the dollar rate;

• rate % – represents the changing in the dollar rate (how many times has it changed);

• seas. mult – represents the seasonal increase in sales in the form of coefficient (how many times will sales increase);

• avail – represents the availability of goods (enough was the goods in the stock for its maximum sale). In store level displaying a "0" or empty space – the product was unavailable, "1" – the goods was enough. In store aggregated level the component represents the number of stores where the goods were avail;

• temperature – represents the average air temperature during the week;

• temperature, Δ – represents the difference between the average temperature for a particular week and the average temperature for the last 13 week

• rain – represents the proportion of raining days of the week;

• snow – represents the proportion of snowing days of the week;

• seas. type – represents a season type (value of 0 for the normal season, -1 for the low season and 1 for the high);

• promo ratio – represents number of days per week in percentage for which promo was held;

• avg. error – represents the average error of forecast;

• auto error – represents the influence of unknown factors on sales volume. It is based on the avg error component and the duration of the component effect can be 2, 4, or 6 weeks, depending on which duration gives the best result (min error);

• median – represents the median of sales;

• base – represents the base sales level;

• min – represents the influence of the sales conversion on sales volume, in the case of a forecast below the historically limiting volume. The limiting sales volume is the lower quartile of sales by default;

• item count – epresents the width of the available assortment (the width of the assortment on the stock);

• receivings – represents the sum of all goods receiving in stores (from the supplier or from the warehouse), minus the write-offs of shortages;

• # of Stores – represents the count of stores in which group or SKU is represented (available at the region level).

Category 'Regressed'

Category regressed consists of the following components:

• reg. a. price – represents the actual price (price with discount) influence on sales;

• reg. price – represents the price influence on sales;

• reg. price % – represents the price changing influence on sales;

• reg. disc. –represents the discount influence on sales;

• reg. disc. % – represents the discount (in percent) influence on sales;

• reg. rate – represents the dollar rate influence on sales;

• reg. rate % – represents the dollar rate changing influence on sales;