What is AI in Procurement?

Artificial intelligence is often defined as having computers perform tasks that normally would require human intelligence.

Basic machine learning (ML) technology is already used by some procurement applications in areas such as spend analytics and contract analytics. This is mostly limited to automating processes.

In terms of AI in procurement, think of AI as a supercharged multiplier to procurement data, systems and teams.

The combination of AI software, algorithms and machines can do what humans cannot possibly do: quickly analyze massive amounts of mundane and seemingly unconnected data to reveal patterns, correlations, suspicious or fraudulent spend, and other anomalies.

These insights can then present opportunities for cost reduction, avoiding expensive surprises, more efficient use of human capital, and mitigating risk proactively.

In a sense, it’s both brains and brawn: the brute speed and reach of modern information processing are what enable AI into ever-increasing applications and enterprise areas. When that analysis horsepower is combined with human intelligence and judgment it can uncover new insights that are not just interesting, but actionable.

Specifically, AI in procurement means crunching through thousands or millions of data points from within and outside the enterprise, from disconnected and disparate data sources. These can include:

Contact us to learn more about AI in procurement.

  • Transaction details
  • Inventory records
  • Consumption and usage data
  • Contract terms and rates
  • Inventory turnover
  • Warehouse utilization
  • Product stock-outs
  • Supplier fulfillment
  • Commodity pricing
  • Market information
  • Historical pricing
  • Industry baselines

The possible inputs are virtually endless.

ai-center-supply-chain-icons-850x567

How AI Will Help Procurement Advance Analytics Beyond Basic Spend Analysis

Historically, spend analysis projects focused on determining how much the was spent by vendor and quantity in search of improvement and efficiency opportunities.

These functions will always be important and some senses have become more difficult. Enterprises are more decentralized. Procurement and provisioning is now happening within functional departments, business units, and geographies. And data are often resident in multiple systems, including p-card and T&E. Getting an enterprise-wide view, especially for mid- and long-tail spend, continues to be elusive.

AI provides the benefit of an analytics engine that can accommodate semi-structured and other data beyond just transactions.

Nick Heinzmann does a great job of laying out 3 examples of AI will help Procurement

  1. AI for Strategic Sourcing
  2. AI for Supplier Management
  3. AI for Supplier Risk Management

Below is a quick summary.

AI for Strategic Sourcing

The traditional approach to spend analytics helps procurement organizations reduce, avoid or recover costs with their suppliers. Historical analysis of spend data can help somewhat in the supplier selection process, but processing a typical quarterly batch of data is hardly the most actionable information on which to base a strategic sourcing effort.

Heinzmann suggests that a more intelligent route is to expand the role of traditional spend analysis in strategic sourcing. For example, by running spend analysis reports before, during and after a sourcing event — a far more complete picture than an out-of-date snapshot. This is now possible via AI-driven robust data acquisition, cleansing and classification processes.

Procurement can be better prepared and informed as it enters into a sourcing event with real-time internal and external data.  Knowledge about suppliers and current market conditions, and the ability to run real-time reports during event (such as breakdowns by product, service and carrier; cross-supplier and cross-carrier comparisons; and variance and outlier analyses) will drive a data-driven strategy and award decision.

AI capabilities also can provide real-time classification to analyze spend patterns with the selected supplier post-event. This can include spend-to-date reports, invoice analysis, realized savings tracking and maverick spend analysis. These capabilities to help tackle pesky tail spend ensure purchases are within compliance.

AI for Supplier Management

After a sourcing event, a supplier becomes just one of many, all of which need to be monitored, managed and evaluated to ensure the relationship is delivering value.

Moving beyond how much the business spent, with whom, in what quantity, Heinzmann advocates a more expansive approach to include scenarios beyond the confines of traditional spend analysis such as operations and logistics. This type of capability needs to ingest semi-structured data and other content beyond just transactional data. AI and machine learning-enhanced analytics can do this quickly and accurately.

For example, instead of just looking at purchase orders and invoices, Procurement can also help determine inventory overhead costs and predict stockouts based on things like inventory turnover and warehouse utilization. New capabilities can extend to include analysis of the supply base by geography and provide insights and track fulfillment times, underlying commodity and fuel costs, and other overhead costs.

These examples illustrate how Procurement can better understand the total cost of doing business with a supplier.

AI for Supplier Risk Management

Progressive procurement organizations are expanding their analytics efforts into a critical adjacent territory: risk.

Risk is more tricky to measure but the potential dire consequences of a supplier fulfillment interruption, breach, default, or other lapse are well understood.

AI-based analytics can cleanse and accurately classify supplier data in near real-time. Heinzmann suggests that what makes AI even more of a game-changer is when procurement enriches this data with external content without significant time delay.

Data sources can go far beyond integrated market price and commodity data feeds. Financial risk scores, sustainability and corporate social responsibility (CSR) scores and similar third-party data sources related to risk can also be incorporated and assessed in the supplier risk profile.

AI and machine learning can uncover trends within these and other key data relationships, leading to broader risk reduction and, in some cases, predictive analytics related to price and margin with suppliers.

What’s more, this deeper intelligence can be used to compare supplier performance across various benchmarks, presenting procurement with an opportunity to continuously improve the supply services it provides stakeholders. 

Read the full article, How AI Will Help Procurement Advance Analytics Beyond Basic Spend Analysis.

Examples of AI-Driven Insights for Procurement

To us, AI and machine learning are not mysterious and encompasses more than software, servers and data.

AI blends these elements along with human guidance in order to simplify complex enterprise neural systems in order to achieve specific goals. AI can and will enable companies to achieve higher levels of operational intelligence, efficiency and effectiveness in order to invest in growth, innovation and their people.

insights-for-procurement-dashboardAI-driven insights include but are not limited to opportunities around:

  • Aggregating demand

  • Consolidating suppliers

  • Best pricing comparisons

  • Reducing maverick spend

  • Identify preferred suppliers

  • Fraud and abuse

Contact us to learn more about AI in procurement.