To do that, it used Celonis Process Mining, a data analytics technology, to uncover hidden inefficiencies, automate and streamline supply chain processes. The results were impressive: Vodafone cut process costs by 11 percent (reducing the cost per purchase order from $3.22 to $2.85), reduced time to market by 20 percent and achieved 100 percent process transparency in less than six months.
Here’s what the global communications company does, and what it faced: with 470 million mobile customers and 14 million broadband customers, mobile operations in 26 countries and fixed broadband operations in 17 countries, Vodafone provides a range of voice, messaging, data and fixed communications. It needed to transform its massive purchasing operations and adapt back-end processes, such as purchase-to-pay, to match changing business demands. Vodafone wanted to leverage the opportunities presented by the digital age, such as process automation and big data analytics, to get its operations to the next level. The gap between performance losses and determining the appropriate course of action to improve the situation was too wide to ensure the efficient operations desired.
Needless to say, driving this transformation is very challenging, with hundreds of thousands of transactions, which accumulate an enormous amount of data. Vodafone annually manages more than 800,000 purchase orders, 5 million invoices and 40 million assets. The company lacked the necessary insights into how processes were executed and performing, especially compared to the desired “to-be” state.
Massive Amounts of Data
With the goal of creating the “perfect purchase order” and achieving business excellence, Vodafone selected Celonis for its big data analytics technology, Celonis Process Mining. Process mining is a technology that sifts massive amounts of data to reconstruct what actually happens in companies’ business processes, spots the inefficiencies and suggests improvements automatically.
The idea behind process mining is simple: use the tremendous amount of data accumulated in a large organization to understand human behavior, automatically reconstruct the way the organization works and find ways to improve. But how to do it? Process mining achieves this by automatically reengineering a model for the organization out of the data.
In Vodafone’s case, it leverages the key systems used for supply chain execution, such as SAP ERP, Ariba and SAP SRM. The reengineered model then shows which cases are handled efficiently, and where the organization needs to improve. Leveraging machine learning and artificial intelligence, the system can recommend improvement measures, which basically creates an X-ray and advisory “machine” for the day-to-day business.
This machine allowed Vodafone to easily build a bridge across the information gap between performance losses and determining the appropriate course of action, as the technology automatically learned how Vodafone’s processes worked and was able to detect hidden vulnerabilities. Altogether, this helped Vodafone’s big data analytics team to transform the organization and create the “perfect purchase order process”.
The “perfect purchase order” aimed to improve the procurement process and reduce the numbers of errors in the ordering process, to ensure “right-first-time ordering” for all customers. This can be challenging, given the size of the operation, variety of different suppliers and quickly changing business needs in the adapting telecommunications market. Vodafone needed a technology to shine a light on its processes – a tall order given the company has more than 10 terabytes of data housed in various storage systems and applications, and also has data on SAP’s HANA relational database management system.
The Celonis technology discovered multiple variations and inefficiencies in the organization. For example, many of the orders with the longest throughput times were showing multiple deviations from the standard process, before the purchase order was even released to the supplier.
Origin of Deviations
Vodafone was able to see the deviations could be attributed to three main causes:
1. Deviations due to low automation levels, when users need to manually insert values, which could be automatically filled by the system if it had a solid set of rules on the back end
2. Deviations due to incorrect data at the source (i.e., tax codes and payment terms)
3. Deviations which could be corrected by enhancing best practices within the buyers’ community (i.e., streamlining material categories, bundling demand with desired suppliers, etc.).
When Celonis alerted Vodafone to low automated transactions, the technology highlighted where the company could improve by using robotic process automation. Since a high level of human interaction was still needed in these transactions, Celonis was able to determine how robotic process automation could easily and automatically replicate the humans’ role without error and at a faster speed, saving the company valuable time and money.
The initiative proved to be highly beneficial to Vodafone’s purchasing operation. Specifically, Vodafone was able to create 100 percent transparency, achieve a significant cost reduction and improve time to market.
Prior to Celonis, running a root cause analysis every time something went wrong took around two days to complete. Analysis now takes seconds, saving 48 hours per issue. That saves more than a week’s worth of valuable time employees were spending trying to figure out the problem with every four issues discovered. This helped Vodafone eliminate errors in procurement, as the company could see deviations from the standard process and intervene. Further, Vodafone was able to maximize catalogue buying and speed up the release of a purchase order to a supplier. By achieving 100 percent transparency into their operations, Vodafone is now able to make all business decisions based on facts rather than assumptions.
Through the “perfect purchase order” initiative, 85 percent of purchase orders are now completed correctly the first time, surpassing the organization’s 80 percent goal in less than six months. With 85 percent of purchase orders completed correctly the first time around, Vodafone reduced the cost of each process order from $3.22 /PO to $2.85 per PO, surpassing The Hackett Group’s world-class standard. This provided Vodafone with an 11 percent cost-savings improvement.
In a dynamic and competitive environment, time to market is critical in order to stay ahead of the competition. Through de-bottlenecking the purchasing operation, Vodafone was able to improve time to market by 20 percent, or two full weeks.
Click here to view the video
Timely, incisive articles delivered directly to your inbox.