New Forecasting system for Klarna Operations!

Klarna is a fast growing company with a mission to become the world's favourite way to buy. To make that happen it is key to create a user experience that is superior in all aspects. That ambition combined with Klarna’s rapid growth brings certain challenges to business operations. On the one hand operations needs to make sure that there is enough personnel to serve both existing and new clients in order to maintain a high level of customer satisfaction. On the other hand operations need to keep track on the growing costs relating to customer service operations.

Varjus Consulting has been helping Klarna's Operations team to build a new advanced forecasting system to predict the amount of consumer initiated contacts to customer service points over various channels. These forecasts are needed on daily levels for operational purposes as well as on monthly levels for financial planning purposes. The longest forecast horizon period spans over one year.

The forecasting solution for Klarna's Operations is a "hybrid" solution. The known weekly and monthly regularities are modelled using statistical time series forecasts. However, Klarna's expansion to new markets and new countries means that everything can’t be forecasted solely based on the past data. Thus, the time series forecasts are enhanced with a set of specific business rules. On top of that, there may be a need for manual overrides on day-to-day basis which needs to be taken into consideration too.

Klarna requires high accuracy levels to offer their users a great experience in all aspects of their user journey. The delivered forecast solution has been able to reach the required accuracy, and thereby guarantees that there is enough personnel in customer service operations to provide a superior service, but without over staffing with negative budget effects.

The good forecast accuracy is a corner stone for the success of Klarna's Operations team. At the same time the solution is highly automated and easy to use for skilled end-users. This frees Operations team from time consuming data preparation and allows them to focus on more business critical activities when serving their users.

40% Sales Increase for Helsingin Sanomat!

Varjus Consulting has been helping Sanoma Data and Analytics- team with building a new analytical environment to support cross Business Unit customer analysis. As part of this project Varjus Consulting was assigned to build an analytical CRM model for passive customers for Helsingin Sanomat. The objective of the analysis was to identify the most probable buyers to be included in the direct marketing campaign for these customers. The analytical model was to be build from the data contents of the newly created analytical environment in order to prove the usability and functionality of this environment.

The accuracy of this predictive model was tested and analysed in a real life situation on two separate direct marketing campaigns during the summer 2014. Comparison of the analytical model was performed agaist the normally used customer lists. The end result of the campaign analysis showed that within the tested campaigns the predictive model built by Varjus Consulting was able to generate 40% more sales on average than the normally used campaign lists.

The building of the model took only 10 days of work from beginning to customer scoring using the new analytical environment. And, the built model can now be automatically refreshed and used in all future direct marketing campaigns for passive customers, generating even more revenue for Helsingin Sanomat in the future.

SanomaNews has a new Data Mining environment

SanomaNews Media sales team decided to build their own, in-house Data Mining environment on world's leading SAS Data Mining technology. The in-house Data Mining skills and technology provides SanomaNews Media sales team wider flexibility and faster reactivity within the area of customer analytics. SanomaNews Media sales team is now able to combine the best of both worlds - i.e. out-sourced Data Mining expertise with the in- house flexibility and reactivity - as needed.

Varjus Consulting was selected to help with creating a flexible and efficient Data Mining environment to support the every day analytical needs of the SanomaNews Media sales team. Together with the SanomaData's IT specialists and SanomaNews CRM- analysts Varjus Consulting was able to design and build a Data Mining environment that allows building completely new data mining models fast and easy. For example, during the hand-over project the team was able to build a completely new predictive Data Mining model from the scratch - within a one single day! This included the creation of a new target variable, refreshing the existing ABT (Analytical Base Table), conducting the Data Mining analysis and scoring the existing customer base with this new Data Mining model.

TeliaSonera designed a new common data mining methodology

Varjus Consulting is proud to inform you that it has been helping TeliaSonera to design a new, common, data mining methodology for TeliaSonera's all Nordic CRM operations including Baltics. The new project methodology allows TeliaSonera to divide the needed data mining phases into well separated steps (phases), which each can be outsourced and insourced as needed to maximize the total efficiency.

The designed methodology is common for all Nordic and Baltic operations and thus it allows greater co-operation between local operations. Now it is possible for TeliaSonera to start building common processes on top of the different data mining models. It also becomes possible to build a model in one local operation and then share these between other local operations.

Varjus Consulting was selected to help TeliaSonera in this task due to vast experience both within the CRM analytics and telecommunications industry.

Elisa's B2B-division is the first reference for Varjus Consulting

Elisa's B2B-division wanted to speed up their CRM analysis and activities with the help of SAS software. Prior to use of SAS tools the customer analytics and generation of contact lists required the use of several different software tools. This process was quite labour intensive and prone to errors as data needed to be moved from one application to another. Varjus Consulting assisted Elisa's Corporate Customer CRM department to build a customer data mart to help the automation of the labour intensive steps of the CRM analysis process.

CRM data mart is a pre-processed intermediate data storage of the customer related information. Data marts are typically used in situations when there exists several similar kinds of activites requiring the very same data to be extracted from the DW (or other data sources) over and over again. With the help of the CRM data mart Elisa's Corporate Customer CRM department is now able to run more customer analysis and generate more customer intelligence with the same amount of resources.

One of the Elisa's B2B-division's CRM activities used to take 3-4 days of manual work on monthly basis. With the help of SAS software the required data was stored into the CRM data mart for an easy access to the end-users. Now, this CRM action can be conducted with just a half-day's exercise releasing more time for additional CRM activities and customer analysis.