Overview

Google BigQuery cloud-based data warehouse can provide global retail companies the data storage resources they require for data warehousing. From marketing data, such as KPIs, target marketing data, click through rate information, to customer recommendations provided from advanced analytic tools; Google BigQuery can store all of that information in an easy-to-use, serverless data warehouse.

Global Retail Customer requirement

When our global retail customer approached us, they had large Point of Sales (POS) data sets. At the time they were replicating their POS data sets from SAP applications to an SAP HANA database. Their HANA database size was rapidly growing, which created a need to move historical data from HANA to a data lake.

They also wanted to gain more insights from their large historical POS data, while also combining it with non-SAP data.

KochaSoft Advice: Google BigQuery for Retail

Through our investigation, we recommended that Google BigQuery would be the best solution for their data lake in terms of cost and flexibility. Not only would they not have to invest in additional infrastructure or increase the capacity of their HANA system by going to BigQuery but also it would be able to scale and keep up with their continued growth.

High-level architecture of the solution:

Above, you can see that the SAP POS is the main transactional system for our customers. The SAP POS data was loaded into SAP HANA using SLT. The analytics on the POS data from SAP HANA is performed in SAP Business Objects and SAP Analytics Cloud.

Then, the historical POS data from SAP HANA is transferred to Google BigQuery using SAP Datahub. To note, the size of the historical POS data was 12TB. Non-SAP data was then loaded into BigQuery using File Transfer and GCP Cloud Dataflow. From this, the SAP data is integrated with the non-SAP data in BigQuery.

Summarized historical POS data from BigQuery is accessed in SAP HANA using HANA Smart Data Access (SDA). This was combined with current data to give a complete overview of their business.

Data Analytics

For our Global Retail customer, we used Google Data Studio for visualization and Cloud DataLab for data exploration. These tools also use applied machine learning to predict future moves in the market and to strengthen target marketing.

Above is a preview of three dashboards that were created in Google Data Studio to display revenue generated by geographic location and online vs. in-store sales. These dashboards also use machine learning to create targeted marketing campaigns to recommend products to customers and the top product category to target.

Another benefit of these dashboards is that our Global Retail customers can view social sentiments from their Twitter feed. In the graphic above, you can see that our customer is experiencing a spike in positive sentiments around last June from a new social media campaign they launched.

Summary

Our solution for our Global Retail customer was a success all around because it helped them tap into customer insights while also giving them a comprehensive overview of retail analytics to provide sales and marketing teams the tools they needed.