Nearly a century ago, in the mid-1920s, a group of grape farmers in South Africa began making wine together, eventually forming the first winery in South Africa, Stellenbosch Farmers’ Winery. In the autumn of 2000, Stellenbosch Farmers’ Winery merged with Distillers Corporation, becoming Distell. Distell is the largest producer of wine, spirits, cider, and ready-to-drinks in South Africa, and is headquartered in Stellenbosch, South Africa.
Distell has branch offices and distilleries in five continents and provides alcoholic beverages to folks across the world. Their primary markets include South Africa, Taiwan, and the United States. As the business expands into more markets and countries, so does Distell’s datasets and workloads. Distell’s on-premises infrastructure included siloed storage and other disparate systems, so data was scattered across the network. It would be costly to expand its capacity and scalability while centralizing data, but Distell’s analytics capabilities couldn’t grow and improve without those qualities.
Distell had an established, trusted relationship with Microsoft, so it looked toward a data warehouse solution based on Azure Synapse Analytics. Together, and with Distell’s other vendors, Distell’s technology team built and deployed a cloud-based platform that could support the company’s market expansion and analytics goals.
Data-driven decisions require a dependable platform
Many of Distell’s distribution, supply chain, and manufacturing decisions rely on big data analytics. Distell collects data from its distilleries, customer and country demographics, local wine and spirits markets, and anywhere its products are served or sold. These analytics inform how Distell understands and caters to its customers—like the fact certain regions prefer peaty tasting whiskey versus something sweeter, or how many cases of cider to provide to a favorite neighborhood pub based on customer demand.
“We wanted a single source of truth to provide a holistic view to the business, one where we could build multiple models using multiple data sources—all from a single repository,” Nielesh Maharaj, Cloud Consultant at Distell, says.
Distell had other key objectives, too. It wanted to use the platform to set its brand apart from its competitors and enable it to modernize other aspects of the business—like smart manufacturing, predictive maintenance, and making operations more environmentally friendly.
“Our cloud vendor needed to align with our culture and values, too,” adds Krushen Pillay, Head of Information and Communications Technology (ICT) Operation at Distell. “Azure sits on fully fledged, green data centers, which supports Distell’s commitment to reducing its carbon footprint.”
“At the end of the day, we chose Azure because we trust Microsoft technology.”
Nielesh Maharaj, Cloud Consultant, Distell
Saving time and labor with a trusted, easy-to-integrate technology
Pillay and Maharaj had experience working with Azure and found its flexibility quite appealing. Maharaj says, “The interoperability of Azure was also a big benefit for us. Transitioning on-premises source data or external third-party data into a single data lake with the integration mechanism that Azure provides is one of the easiest I’ve worked with.”
Distell had previously deployed Microsoft Office 365, which showed the executive team the power and possibility of the cloud. Between leadership’s positive experience with Office 365, the technology team’s trust in Microsoft, and the close relationship with Microsoft technical and early adoption teams, it was easy for the executive team to approve Azure for Distell’s data warehouse project. Distell looped its third-party database vendors, SAP and Orbis, into the conversation, and communication remained transparent and clear between Distell and all three technology vendors.
Distell’s data analytics team worked with SQL, so onboarding to Azure was easy as it uses the same query language. Maharaj says, “Bypassing the learning curve for a new query language led to a direct cost saving.”
Pillay says, “Microsoft’s FastTrack and customer support teams were generous with assistance and support, helping us learn as we simultaneously deployed the platform.” Because of this hands-on assistance and collaboration, training and deployment took about three weeks. By the fourth week, the proof of concept was fully deployed.
From initiating the cloud-based data warehouse project conversations with the executive team to deployment of the proof of concept, the process took four months. After deployment, the excitement about the platform, support from Microsoft teams, and the Enterprise Skills Initiative app encouraged other Distell technology team members to pursue Azure training and certification.
“We forecast that moving our analytics to the cloud could be a three-year process, but because of our teamwork with Microsoft and the simplicity of Azure, we reduced that to a month.”
Nielesh Maharaj, Cloud Consultant, Distell
Following a proven data warehouse model
In designing its analytics solution, Distell followed Microsoft’s modern data warehouse blueprint. Azure Data Factory ingests all of Distell’s raw data—from vendors, customers, distilleries, bottling plants, marketing, internal ERP and CRM systems, and SAP and Orbis data sources—piping it into Azure Data Lake for storage.
Then, using Azure Synapse Analytics, the platform transforms and links the data to Power BI. With Power BI, users can generate a report with a few clicks, thereby reducing manual reporting time and labor costs.
The first workload stream that Distell focused on for its proof of concept was its data analytics department. The technology and data analytics teams saw the value of Azure-backed analytics as soon as the proof of concept began pulling in data from Distell’s ERP and CRM systems.
“Because of analytics we’re getting out of our data lake—looking at consumption data from the previous weeks—we know which products and which sizes of those products to manufacture each day,” says Pillay. One of its plants can produce more than 90,000 bottles of cider an hour, so forecasting can help Distell save on production and distribution costs and make sure product isn’t sitting un-purchased on store or pub shelves.
The Distell analytics team then built models using machine learning, AI, and predictive analytics to use historical data to forecast one, three, five, and even 10 years into the future.
Endless possibilities with Azure and Microsoft
The proof-of-concept project helped the technology and leadership teams instantly see how an Azure-based data warehouse can help increase the accuracy of their forecasting. Maharaj and Pillay already have other projects in mind.
To improve the customer experience, Distell plans to use Azure to capture reseller data and compare and analyze it against consumer data. This will help it pinpoint which ciders, wines, spirits, or ready-to-drinks its customers are enjoying at the bar or buying from the corner store. This information can then be analyzed against other company data to help inform manufacturing and distribution decisions.
Another future goal is smart manufacturing, using Azure and IOT devices within a distillery or bottling plant to collect and analyze production data. This could provide insights into machinery performance and allow Distell to predict when a machine will need maintenance.
There will be other SQL migrations, too, and there is excitement to see what augmented reality and AI could do for Distell. Maharaj and Pillay see potential to use Azure in Distell’s call centers and e-commerce platform. Eventually, Distell wants to reduce its footprint even further by going serverless and tracking and retrieving its glass bottles with QR codes.
Azure has proved itself in the eyes of Distell leadership, so when approached about these future project ideas, they only ask Maharaj and Pillay what resources they needed to make it happen. Similarly, when Maharaj and Pillay go to Microsoft technical teams with their Azure plans, the teams ask Distell what it needs from Microsoft to turn ideas into reality.
Pillay says, “The Microsoft teams are always willing to assist us in our cloud and analytics journey.”
“And that’s why we’ll stick with Microsoft as we move forward through our cloud journey,” Maharaj adds.
“Having worked with other public cloud offerings, Azure is definitely leaps and bounds ahead in terms of how dedicated they are to helping the customer achieve their organizational vision and how simple Azure is to use.”
Nielesh Maharaj, Cloud Consultant, Distell
Follow Microsoft