Decision Inc. has assisted a South African-based multinational retailer migrate its disparate data environment to the Qlik Sense analytics platform to facilitate self-service business intelligence (BI) and enable the group to optimise insights from its data.
Its South African retail business operates more than 200 full line stores (comprising fashion, home and beauty, and food), and more than 400 stand-alone stores in the country. The group has close on 45,000 employees and serves more than 17 million customers across the southern hemisphere.
Moving to self-service
To align with its long-term strategy of enabling governed self-service, the retailer needed to migrate its QlikView analytical environment to the new generation Qlik Sense data analytics platform.
One of the most significant challenges the organisation wanted to resolve was to have self-service business intelligence (BI) capabilities enabled that would see the in-house development of models and visualisations to analyse data more effectively. However, because of the complexity and size of the organisational infrastructure, its data was spread out across multiple environments and locations.
“Decision Inc. was called in to demonstrate how to build a common data layer that would not only optimise how data is managed inside the organisation but also provide business users with access to data from different environments. Through the self-service approach that Qlik Sense would facilitate, the retailer would be capable of creating its own BI models, dashboards, and reports within a regulated environment managed by the internal IT BI team,” says Rudie Vermaak, Senior Client Partner at Decision Inc.
Furthermore, the shift towards self-service would realise better insights from their data faster than was possible before. The business would have the ability to adapt faster to changing market conditions. The current Decision Inc. pilot project runs in the Amazon Web Services (AWS) cloud with consumer-specific data.
Complex data management
“Many organisations are looking at ways to enable self-service BI for their end users and staying up to date with the latest technologies. But the complexity of the environment that would impact more than 1 200 users meant that Decision Inc. had to work closely with their internal team to formulate and design a strategy that would ensure the migration happen with minimum disruption,” adds Vermaak.
Part of this entailed identifying which data and data sets the different business units at the retailer use. Once done, it would be easier to repurpose the data and performance enhancements to improve the speed of data modelling.
For instance, the time it took to load processes were reduced from an hour to just a few minutes with this enhancement. More than 400 models had to be converted from QlikView to the more advanced Qlik Sense environment. Decision Inc. consultants worked with the various business units to put the systems and processes in place to improve what was used before.
“The goal was to empower the retailer to have better (and more) insights into their data. The company managed the change management process internally even though Decision Inc. did the upskilling of their internal teams. There was extensive work involved to make the user experience like what employees were used to. To this end, Decision Inc. built a positioning on how best to view the data for the retailer,” adds Vermaak.
A large part of the migration had to be done during lockdown conditions in South Africa.
“Despite what was a difficult operating environment, the Decision Inc. team managed to work together and finish the project ahead of schedule. The project began in June 2019 and was initially set to be completed by July 2020. When lockdown hit, Decision Inc. assigned six consultants to work remotely on the migration resulting in it being completed ahead of time.”
This was mostly due to a formalised project management plan in place to keep track of all the initiatives. Decision Inc. had frequent ‘show and tell’ sessions with the retailer to ensure every aspect of the project was done according to the client’s expectations.
“The migration is testament to the ability of Decision Inc. to deliver quickly on client needs with no delays in working remotely. We are showing that we can work with any client irrespective of their geographic location and deliver on time and budget,” adds Vermaak.
Leveraging multi-environment data
Through Qlik Sense, Decision Inc. has enabled the retailer to gain insights into billions of records of data.
“The organisation can now effectively leverage on-premise and cloud data through the AWS hybrid environment we enabled. One of the smaller data tables encompass 100 million records while the biggest accesses 1.6 trillion records. This can now happen in real-time using the Qlik Sense environment where the retailer can go into their history and access all relevant information,” says Vermaak.
Given the sheer scale of the project, Decision Inc. put in place a best practice approach that could be used across the data sets. This result in 340 models being built, 130 front-end models (the actual visualisations) being developed, and six out of the nine business units embracing Qlik Sense.
“While there is still more that can be migrated, this falls outside the scope of this particular project. The Decision Inc. team was able to remote many duplicate extractors that took up space and resources. The retailer can now be more efficient in its data management, embrace self-service BI using less resources, and continue to rely on the best practices Decision Inc. has put in place for the organisation,” concludes Vermaak.
Location: South Africa
Function: South African-based multinational retail company comprising fashion, home and beauty, and premium food lines.
Challenges: Required self-service business intelligence capabilities to build their own models and visualisations to analyse data.
Solution: Qlik Sense data analytics platform