Tag Archives: supplier data

Teeing Up For AI in Procurement: It’s All About One Thing…

The benefit of AI for procurement is clear – the question, then, is what will it take to effectively put it to use?

Over the last year, machine learning and artificial intelligence (AI) technologies have graduated from the class of “emerging tech” – they’re here now, they’re increasingly sophisticated, and their adoption will only continue to accelerate.

We’ve seen machine learning and AI go mainstream in consumer tech environments, and they are rapidly shifting from hype to reality in enterprise environments as well; however, enterprise executives are still working to understand how AI applications can move beyond specific product features to influence broader business functions and strategies.

Let’s take a look at the procurement department, for instance. Procurement and purchasing professionals have a lot to gain from leveraging AI. In fact, AI has the potential to completely transform how organisations manage their spend, from automating invoice coding based on learned criteria, to predicting potentially fraudulent transactions, and preventing rogue spending before it happens.

The benefit of AI for procurement is clear – the question, then, is what will it take to effectively put it to use?

Gartner’s report, “Start Preparing Now for the Impact of AI on Procurement,” states that “technologies’ need for data will force application leaders in procurement to ensure access to the necessary internal and external data sources.”

Essentially, the first step to getting predictions out of AI is to capture all data – internal data, external data and third-party, public data. Furthermore, procurement professionals should be asking themselves if they have the volume, the quality and the completeness of data needed to leverage AI within their department.

Ticking each of these boxes can feel like an arduous process, but a good starting point is to hone in on three particular sources of data that provide the greatest visibility into spend:

1. Supplier Data: This means capturing data from 100 per cent of suppliers in the procurement system. Not just the largest multi-national suppliers who use sophisticated EDI or XML formats, but the whole tail. This should include mid-tier suppliers that may be using online portals or emailing PDF invoices, all the way down to the smallest “mom and pop” businesses, who continue sending paper invoices. Using an open commerce network that accepts and supports all invoice formats and requires no changes on the supplier’s end enables 100 per cent supplier onboarding and captures all transactional data. To gain true visibility and power future platforms, procurement and finance leaders must aggregate as much financial data as possible beginning with supplier data.

2. User-Driven Data: The ability to capture user-driven data–specifically, buying insights that track 100 per cent of all purchasing requests that run through the system, is vital. Visibility into employee spend ultimately depends on how user-centric procurement tools, technologies, and processes are designed. The bottom line is: procurement systems shouldn’t be designed for the procurement department. They should be catered to potentially thousands of employees around the world that are buying things in their organisation.

Searching for orders, dynamic routing and approvals, and guided buying, for instance, should be easy to navigate and fit seamlessly into the way employees already work. The key is to create a system that users adhere to not because they have to, but because it’s the easiest way to get what they want from preferred vendors at the negotiated price, providing another layer of spend visibility.

3. Invoice Data. By nature, the accounts payable function is primed for intelligent automation. There is a huge opportunity to use AI for things like improving processing efficiencies and reducing costs, increasing discounts and eliminating late payment fees, for instance.

But, these enhancements can only be achieved if the invoice data feeding into AI is complete. That means procurement needs to capture 100 per cent of invoices, irrespective of format (paper, PDF, electronic) and irrespective of invoice type (PO-based, non-PO based, invoices for direct spend, for indirect spend, for facilities and utilities, etc) –  truly, any and all. Whatever the invoice, it should be captured.

These three particular sources of data can truly position a company to take advantage of all the benefits AI promises just over the horizon. Elements of machine learning, AI and predictive analytics already exist within procurement today. Forecasting budgets for approvers, alternative cost-effective suggestions during a user’s shopping experience and intelligently aggregating POs based on purchase trends are just a few commonplace applications. But to take advantage of any of these applications, and future opportunities to gain a competitive advantage, data is an absolute prerequisite. Only when armed with data – especially from suppliers, users and invoices – can procurement make the most of their investment in AI technology, enhance spend visibility and optimisation, and ultimately, boost the organisation’s bottom line.

Continue reading Teeing Up For AI in Procurement: It’s All About One Thing…