How Pick n Pay’s migration to the cloud is paying off

Pick n Pay’s bold plan to modernize infrastructure and drive efficiency is beginning to pay dividends as its migration to AWS was successfully completed last year, signaling the digital ambitions of this retail giant.

This move to the cloud lays the foundation for further expansion into other cloud-based applications to deliver deeper insights and better digital services, as well as drive cost savings for the company and the consumer.

The company’s recent trading update confirmed it would accelerate growth opportunities, such as its online offerings, and develop serverless computing with AWS so the retailer is able to rapidly scale when needed, like over Christmas or Black Friday.

The data warehousing team was the first to take its platform to the cloud and is already leveraging AWS’ suite of services and cloud-first applications, like Snowflake, for self-service analytics. And over the past 10 months, sales growth across all online platforms (scheduled, on-demand, and click-n-collect) came in at a robust 69.6% year-to-date, with on-demand year-to-date growth in excess of 100%.

Mark Tudor, Chief Information & Technology Officer, Pick n Pay

Pick n Pay

Although the new foundation promises great things in omnichannel retail for the future, the decision was in discussion for years, according to Pick n Pay CIO Mark Tudor. “Around 2019, we understood we needed to make a significant investment in our infrastructure,” he says. “We had to decide whether to invest in another five-year depreciation cycle with our partner in our existing data, or move to a hyperscaler. At the time, we decided it was better to go with a hyperscaler because we wanted the cost-effective agility this approach would provide in the coming years. We didn’t want to invest in a data center that would commit us to another five-year depreciation cycle. That was really the catalyst.”

Pick n Pay, a multiformat and multichannel retailer with 90,000 employees and an annual turnover of R97.9 billion ($5.3 billion), operates across mostly Southern Africa with stores in Namibia, Botswana, Zambia, Nigeria, Eswatini and Lesotho. It also owns a 49% share of the Zimbabwean supermarket chain, TM Supermarkets. So with such size and reach, the choice of a hyperscaler like AWS made sense. In the beginning, they were firmly focused on Infrastructure-as-a-Service.

“Our view on hyperscalers is that they provide near infinite capacity on demand, which is what we were looking for,” Tudor says.

Managing the migration

The actual process took 10 months and Tudor says they mostly looked at an IaaS-led migration as opposed to a platform rebuild. “Moving from an on-premise infrastructure service provider to a hyperscaler-based one was more of a partner change for our staff as opposed to anything fundamental,” he says. “Now that we’re in the cloud, there are a lot of complementary services and platforms around our infrastructure to better leverage opportunities.”

Critical to the success of the migration was AWS premier consulting partner Lemongrass, which enabled Pick n Pay to migrate its mission-critical SAP (including SAP S/4HANA) and non-SAP workloads to AWS, and provide the ability to manage these workloads with the highest levels of automation and agility.

Eamonn O’Neill, Lemongrass Consulting CTO, explained the rationale: “We typically see potential cost savings of between 45% and 75% resulting from this strategy. These are savings that can then fund initiatives, such as new offerings and tools that deliver superior retail experiences, and drive additional value for the business, such as business intelligence and automation.”

Superior cost management

One of the largest benefits of the migration has been around cost, as Pick n Pay has reduced its infrastructure spend by approximately 40% while driving up its competitiveness.

The trading report also states that the group is on-track to deliver its targeted efficiency savings for FY23 across store operations, support offices, and supply chain, and is also fully focused on delivering its target R3 billion in savings over three years.

“We needed a strategy we felt was right for us going into the future and was also cost-effective,” says Tudor. “That’s worked well for us. The complexity of managing environments in the cloud from a financial perspective is something we already had experience with from taking our data warehouse into the cloud a few years ahead of the main migration.”

In terms of environments on the ground, however, migration to AWS offers another benefit considering how South Africa grapples with the knock-on effects of an unstable power supply. While Pick n Pay spends R60 million a month on diesel for its generators to keep its stores operating through rolling blackouts, its IT infrastructure hasn’t added to this since there aren’t any data centers to manage. “With any data center, there’s the challenge of cooling and managing your utility power effectively,” says Tudor. “We haven’t had to worry about this given it now sits in the cloud. That’s been a massive added win for us.”

Lessons learned

When it comes to smaller retailers on the continent looking to implement something similar, Tudor offered some advice. ‘For smaller retailers, you’d probably do better buying a cloud solution directly as opposed to carrying infrastructure into the cloud,” he says. “Our theory is to look to SaaS first, PaaS second, and IaaS third. The economics of a service you procure from the cloud is more cost-effective with the SaaS service. For example, use Workday in the HR space rather than build your own infrastructure in your own virtualized data center in a hyperscaler. That’s how I’d approach it.” So far, Pick n Pay’s forward-thinking initiatives are having the desired effect of driving efficiencies and creating cost savings for customers. But the power crisis and other retailers innovating and posing greater competition means they can’t afford to be complacent. So leveraging the power of AWS to create new services and offerings will be crucial.

CIO, Cloud Management, IT Leadership