Over a million developers have joined DZone.

Advanced Analytics in Procurement

DZone 's Guide to

Advanced Analytics in Procurement

Melding advanced analytics with business data is key to maximizing the potential of any business' procurement department. Zone Leader Sibanjan Das offers up a few possible use cases to marry the two.

· Big Data Zone ·
Free Resource

The procurement department is a key business segment for most companies and the backbone of many organizations. They are involved in decisions that impact the cost of carrying out an activity, buying goods or services, and building strategic relationships with stakeholders and suppliers. Their role plays an important part in driving the profitability of an organization by helping to streamline processes, negotiating to reduce material costs, and to identify better supply sources.

In the software industry, the Procure to Pay (P2P) cycle is a common way to track the end-to-end procurement process—from raising a purchase requisition right up until the payment to the suppliers. The P2P cycle has a series of stages as we can see from a standard diagram below. What we have shown below is a standard process. Depending on the business, this process has many variations.

Image title

As procurement is an essential business function, there are many Key Performance Indicators (KPI) used by various industries to track their procurement department's performance. Broadly, they are categorized into three groups according to their primary purposes: Reduce Spendings/Costs, Improve Quality, and Deliver Quickly. One needs to balance the KPIs in these three goals for procurement to achieve results. If you want more details about these indicators, check out Deltabid's Procurement KPI article. 

  1. Costs KPI - Cost Avoidance, Cost Reduction and Procurement ROI 

  2. Quality KPI - Supplier Quality Rating, Rework and Scrap Value in Dollars, Return to Vendor Cost, Ratio of Rejection, % of certified suppliers

  3. Delivery KPI - Lead Time

Apart from these procurement KPIs, there are other KPIs that are shared by different business functions that provide a holistic view of an organization's financial performance such as:

  • Days payable Outstanding(DPO): An important KPI which tells how long it takes a company to pay to its suppliers. It provides information on the outstanding and overdue payables and also forms an essential element of the Cash Conversion Cycle.
  • Cash Conversion Cycle: The Cash Conversion Cycle (CCC) is equal to the time is takes to sell inventory and collect receivables less than the time it takes to pay the company's payables,

  • Total Landed Costs: It is the total price of a product once it has arrived at the destination. It includes the original price of the product, transportation fees, and other handling charges.

  • Payables vs Receivables: An essential part of liquidity analysis to see if enough funds are coming in from receivables to pay for the outstanding payables.

  • Unreconciled Payments: Helps us to understand the payments that are not matched with invoices, receipts, or bills. This encourages us to correct any errors with the payment processing and also detect potential frauds/overpayments to suppliers. 

  • Invoices paid after due date: An important KPI to understand the value of invoices that are paid after a due date. Invoices paid after a due date might attract late payment charges and may lead to a bad reputation in the market. 

Advanced Analytics in Procurement

Advanced analytics are required to improve and optimize the procurement function. It equips professionals working in procurement with substantial information to support them in doing their jobs better. Analytics can help them reduce costs by identifying instances where they have unnecessary spending, monitor suppliers by their deliveries and negotiate for better deals, quickly identify exchange frauds and let intelligent machines manage manual tasks like quotation creation and supplier scoring.

Deloitte predicts Procurement in 2020 will look very different than it does today. This future world will alter existing assumptions and introduce emerging hazards. It will require additional skills, knowledge, and tools to address entirely new challenges while solving current ones more creatively. So, it is extremely important to take advantage of the new technological advancements such as IoT and digitalization. Advanced analytics is important to convert the big data from all these new sources into smart data for smart procurement. 

Below are few advanced analytics exercises which can benefit a procurement team:

1) Spend Analytics: Spend visibility is necessary to understand how much, where, and with whom organizations are spending money. This can help to identify potential opportunities to reduce expenditures. Reducing the spend is a fantastic way to increase ROI. Decreasing spend is complex, but once you do it, it directly affects the bottom line. Big Data has changed the definition of spend analytics. What makes analyzing spending complex is the decentralized amounts of messy data from different systems, where maverick transactions can carry out through different mediums such as purchase cards, wire transfers, purchasing platforms not integrated to payables, and multiple procurement/sourcing systems.  

2) Business Process Analytics:
 Let's think of an organization that wants to optimize the procure to pay process. The traditional approach is a standard interview-style discussion with business users, and their answers form a baseline, an 'As-Is' process, for process improvement. However, most of them provide an ideal answer about the standard process followed. This forms the average 70-80% times the process is followed in the organization. The remaining 20% are the exceptions that are not so visible and the areas that need attention for improvement. Here, the process analytics can aid to mine this information for the event logs. You can get more information about this in my previous article Business Process Analytics.

3) Internet of Things (IoT) Analytics: IoT is going to change the way procurement is done. As Spend Matters and Vroozi wrote in their jointly produced research paper Declaration of the New Purchasing: A Buying Manifesto:

Article 12: The Internet of Things (IoT) will surround us, creating unprecedented levels of visibility into the consumption patterns of what we buy and how we use it, creating feedback loops and changing how we manage demand.

All purchases (including money spent on people and labor) will be tracked and monitored continuously. From apps on devices that track movement and access to tagging equipment and supplies, we will create unprecedented sets of data from which to analyze and make better decisions. The feedback loops between different tagged assets, items, people, and customer activities will further create new levels of visibility through meta-data analytics, changing the very basis of how we assign our time and effort to different activities in procurement.

This unprecedented growth of data is going to make the procurement department focus more on analytics than ever before. Advanced data transformation techniques and modeling can improve spend management and catalog content by knowing exactly what is being used and exactly what is needed. Also, IoT is going to help track requests of over-specified items that are not required and can help to control unnecessary spending.

3) Procurement Frauds: Advanced analytics can build models that identify attributes or patterns that can identify fraud. For example, anomaly detection can detect a sudden drift in historical purchase or supplier payment pattern. Clustering can help with comparing similar peer purchase requestor groups and recognize behaviors that are drastically different from what would be required for that group or type of procurement. Profiling can assist in determining potential fraudsters by matching its attributes with known fraudsters attributes. Text Mining on the requested item/supplier's profile can help with identifying good and bad issues/providers for the same specifications. 

5) Robotic Process Automation: RPA can assist procurement staff in handling manual tasks, thereby eliminating human errors and freeing them up to concentrate on better things. If we analyze common pitfalls in a procure to pay process, we can find many functions such as lack of standardized invoices from suppliers, inconsistent data and documents, process inefficiencies, duplicates, and reactive fixes for overpayments. All of these can be automated using automated assistants. Also, intelligent chatbots can be used to communicate with suppliers to get their queries resolved on minor and day-to-day issues.

6) Supplier Rationalization: It is a process to optimize the supply base and find an optimal mix of suppliers that can drive down cost. For this exercise, it is crucial to know the details of the products they supplied (any quality defects or the number of on-time deliveries), suppliers details (Supplier rating, Supplier certifications, Reputation, etc.). This necessitates analyzing a huge amount of data, profiling each supplier, and predicting whether it is beneficial to maintain the relationship with them in the future.

7) Supplies Price Monitoring: Are the suppliers charging us the correct amount? Can we drive cost savings from falling prices? Monitoring prices from several e-commerce and supplier web portals can assist procurement professionals to drive cost savings better. Advanced analytics and RPA comes in handy here where the web portals can be crawled automatically and provide summarized and relevant information to the professionals. 

These are just some of the areas where advanced analytics can benefit a procurement department of any organization. Have I missed any other use cases for advanced analytics in procurement? I am sure there are plenty which I'm not aware of and have not covered here. Readers, please share your views and ideas on other procurement areas that can be benefited by advanced analytics and disrupting technologies.

big data ,data science ,analytics ,advanced analytics ,procurement

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}