Predicting Quotation Conversion Rates with Machine Learning

Introduction

SAP S/4HANA Suite has built-in analytical features to enable enterprises to monitor and analyze their regular business operations and issues. The embedded real-time analytics offer extensive monitoring and analytical capabilities across several core areas and business processes; both for cloud and on-premise environments.

As specified in my previous blog (Predicting Sales Performances with Machine Learning); embedded predictive analytics offer 3 business scenarios in context with cloud sales i.e. –

1. Sales Volume Prediction

2. Quotation conversion rates

3. Predicted delivery delay

With this blog, I would like to cover the business scenario “Quotation conversion rates” empowered with predictive modelling (using machine learning) and do a detailed walkthrough of its capabilities and how enterprises can benefit from this functionality available in S/4HANA Cloud.

Context

This scenario addresses the net conversion rate of customer requests from a quotation stage to an ordering point (in percentage).

The app “Quotation Conversion Rates” under business catalog “Sales Quotations”, allows to track the status of quotation conversion rate into sales orders before they are expired. It measures the percentage of the net value of order items that have been converted from a quotation item, based on the total net value of quotation items.

Additionally, with embedded machine learning capabilities, this app can provide predictive insights into quotation conversion by comparing actual and predicted results. It helps to predict to what extent the sales quotations can be converted into sales orders for the future periods.

Pre-requisites

To evaluate the status of conversion rate, the quotation items have to meet the following criteria:

· Net value of the quotation should not be ‘0’.

· Quotation validity (expired or future dated quotations are not considered)

· Not fully referenced (header of the quotation)

· Overall quotation status is “Not completed” (header of the quotation)

Predictive Model Training

To understand how the training of a machine learning model works, it is necessary to know the basic concepts of predictive scenario and predictive model.

Predictive scenarios are standard business scenarios that are available as an out-of-box solution within S/4HANA Cloud. These scenarios depict the predictive use cases of an enterprise with specific business goals, type of prediction required for business operations and the historical data required for the prediction.

Predictive Model on the other hand is a model used for predicting outcomes based on a historical data set and are linked to specific predictive scenarios. For a model to predict an accurate outcome, it must be trained at regular intervals with most recent data. Re-training an active model multiple times with most recent data will ensure the predictions are precise and meaningful.

Creation of multiple versions to a model is possible with a different status. Further, the system allows to review the quality and the training status of a model and delete any inactive model version(s) if necessary that have never been used or set to active.

An active model version can be put to a quality check and decide if it represents the business data by analyzing the responses in relation to a business use case. The predictive model performance can be improved by continuously exposing it to new data sets as they are available. There are two indicators available in the predictive model app to evaluate the predictive power and prediction confidence. These indicators can be used to evaluate the performance of a model version.

o The quality indicator “KI” evaluates the explanatory power of the training model when exposed to a data set. A perfect model would possess a KI equal to 1 and a completely random model would possess a KI equal to 0.

o Whereas the robustness indicator “KR” defines the prediction confidence and strength of the model. The degree of robustness corresponds to the predictive power of the model applied to an application dataset.

Solution Model

Quotation Conversion Rates – Valid / Not Completed with predictive capabilities.

For business roles internal sales representative or a sales manager, the analytical app “Quotation Conversion Rates – Valid/Not Completed” conveys the converted percentage of net value of quotations into sales orders. Further, the in-built machine learning capability of this app, enables to predict and compare the quotation conversion rate for the coming periods as well. Let’s look at the functional capabilities of this app in detail below –

View Quotation Conversion Rates – The “quotation conversion rates” tile on the Fiori landing page displays the overall conversion rate of the quotations.

The header area provides a classical filter section to enter exchange rate type and display currency. Add filter option as shown in below snapshot, allows to add additional filter criteria if required.

The content area reflects the data based on the selected parameters from the header area. It can be displayed in 2 different views i.e. –

· Chart view

· Tabular view

The graphical representation (chart view) of quotation conversion rates can be displayed in various dimensions such as sales organization, customer, material, and employee responsible.

Also, the dimension to view Top 10 Quotations by Net Value shows the details of 10 quotations with highest net values that are converted into sales orders within the validity period and the dimension view bottom 10 Quotations by Conversion Rate presents the 10 least converted quotations, especially those that are close to expiring.

Below is the snapshot of tabular view

The toolbar in the chart area allows to customize the content area with various features such as excel download, choose a different chart type, drill-up the values and chart or table view in a different dimension.

Predict Quotation Conversion Rates – the built-in machine learning capability of this app allows to compare the actual and predicted conversion rates using different dimensions.

To view the predicted conversion rate, click on the ‘Show Mini charts’ link on top right-hand side corner in the header area. A mini tile (hidden by default in the overview screen) would appear in the header area with title “Quotation Conversion Rates – Predicted”.

Similar to conversion rate chart type as shown above, the comparison graph between the converted value with the prediction value can also be displayed in different dimensions such as customer, material, sales employee etc.

To enable predictive order value comparison, the default model in the predefined predictive scenario for quotation conversion rate (SLSQTANPREDICTION) must be activated. It is required to create versions by training the model at regular intervals with latest sets of business data using the Predictive Models app.

Re-training an active model multiple times with most accurate data will ensure the predictions are precise and meaningful. If this predictive functionality is not enabled, then this app would display the predicted result as 0%.

Below is the snapshot of tabular view of converted value and predicted value comparison.

Further, it is possible to list quotations with details such as customer, net value, actual and predicted conversion rates, percentage of elapsed time over validity period, and remaining days before expiry. Kindly note that for the remaining days before expiry, the app displays maximum of 1000 days only. Net values can be viewed in preferred currency based on the display currency entered in the header.

Conclusion

The quotation conversion rate app allows to access the current conversion rate as well as predictive value comparison within the same app. This embedded analytical app is easy to navigate, compatible with desktop / tablet devices and offers various functions that can provide true value to business.