In recent decades, procurement has been recognized as a strategic business function. Procurement and supply chain management impact every aspect of business operations. They have transformed from clerical tasks to professionalized fields deeply integrated into the fabric of a business’s operations.
The new importance of procurement gave rise to efforts to rationalize, digitize, and centralize procurement and the management of procurement data. The adoption of eProcurement and enterprise resource planning applications led to the automation of many procurement processes. Today, we stand on the cusp of another procurement revolution, as machine learning and other cognitive technologies give rise to intelligent automation.
Cognitive Technologies In Procurement
Artificial intelligence, machine learning, and deep learning excel at recognizing patterns and making predictions. Unlike rules-based categorization systems of the past, modern machine learning technologies don’t have to be painstakingly programmed with decision logic. They process massive amounts of data and learn for themselves how to turn inputs into outputs that provide valuable insights.
Cognitive technology is used today in many industries. Artificial intelligence, natural language processing, robotics, and computer vision automate processes that once required extensive manual labor and human analysis. The result is intelligent automation that uses predictive analytics and data classification to help professionals to make decisions.
KPMG predicts that spending on the application of cognitive technologies to automation will rise from $12 billion today to greater than $230 billion by 2025. Businesses are already transforming processes to take advantage of automation. According to IBM, 60% of businesses are optimizing processes for automation. Almost half are training humans to work with machines, and nearly a third are using machines to make recommendations that impact decisions.
Procurement and supply chain management will be transformed as cognitive technologies mature. Businesses will employ intelligent automation to manage and assess suppliers, categorize procurement data, make predictions about their needs and the ability of suppliers to meet them, monitor and manage compliance, increase procurement efficiency, and streamline the processing and analysis of procurement data.
SaaS eProcurement platform providers are today exploring ways in which artificial intelligence and intelligent automation can be integrated into their platforms to generate value for users.
SAP is working on bringing intelligent automation to its eProcurement platform. According to the company, “These technologies effectively meld large amounts of data across internal demand, spend, and supplier and network data with social, mobile, and cloud technologies to make procurement smarter so you can work better.”
Spend analysis is one area in which SAP Ariba already uses cognitive technologies. B2B businesses may have to analyze and categorize thousands of invoices, a labor-intensive and expensive process. SAP uses convolutional neural networks — a type of deep learning algorithm — to analyze invoices, extract data, and classify the information according to useful criteria. This application alone could reduce the cost of procurement, but SAP is also working on applying similar technologies to contract intelligence, supplier risk management, and sourcing.
SAP is far from the only eProcurement vendor invested in cognitive technologies. Coupa developed the Coupa AI Classification system for “standardizing, normalizing, and enriching spend data.” Coupa takes advantage of the massive amount of spend data it has access to, training machine learning models that can classify spend data and help procurement professionals to make decisions based on accurate and timely information.
Oracle, of which NetSuite ERP is a subsidiary, is a pioneer in the application of cognitive technologies to business problems. The company has added machine learning features to its enterprise software products, including its enterprise resource planning platform, Oracle ERP Cloud, and its supply-chain management platform, Oracle SCM Cloud.
In 2018, NetSuite was billed as the world’s first intelligent cloud suite. It implements AI across many domains, including finance and procurement, which benefit from AI and machine learning for payment analysis, cash flow predictions, and risk analysis. Netsuite also includes AI solutions for Human Resources, manufacturing, marketing, and customer services.
Preparing For The Cognitive Revolution
The leading ERP, eProcurement, and spend management platforms have invested in cognitive technologies to bring intelligent automation to their users — an investment that will increase in the coming years. But to make the most of intelligent automation, B2B buyers have to understand how these new techniques and processes can be leveraged by their business.
I have already mentioned that almost half of businesses are training their employees to work with machines. Training is a critical first step. Intelligent automation is focused on reducing manual work and streamlining procurement processes, but there will be a human in the loop for the foreseeable future. Procurement professionals will be expected to work in partnership with artificially intelligent machines.
B2B buyers also face the challenge of integrating their suppliers so that data is available to eProcurement platforms. Without data, cognitive technologies can’t reach their full potential. Manual processing of procurement documents — requisition orders, purchase orders, invoices, shipping notifications — is slow, expensive, and error-prone. Worse, many suppliers and manufacturers conduct business via email, locking useful data away in silos.
The answer is to integrate buyer eProcurement platforms with supplier eCommerce applications, automating the flow of data between platforms via punchout catalogs, eInvoicing, and other process automations.
To be most useful, eProcurement platforms need accurate, up-to-date data from across the supply chain, including small suppliers. Custom ad-hoc supplier enablement and integration projects don’t scale to the long-tail of the supply chain, which can mean that the data from thousands of transactions is unavailable for intelligent automation and analysis by machine learning algorithms.
Fortunately, there is a solution. A cloud integration gateway can “translate” between incompatible platforms, allowing buyers to quickly integrate a large proportion of their supply chain for automatic data interchange. As cognitive technologies become increasingly crucial to eProcurement and supply chain management, fast and inexpensive integration between seller eCommerce applications and buyer back-ends becomes an essential service.
In a few years, today’s cutting-edge machine learning applications will seem rudimentary. Businesses will reap rewards from the intelligent automation of inefficient manual processes, and those that embrace cognitive technologies will flourish. We are at the beginning of a cognitive revolution.