Forecasting the future of procurement

One of the main challenges for procurement professionals is smart business buying: accurately forecasting demand and identifying what stock will be needed and when. Placing orders too early could lead to excess, obsolete or unsold stock; placing orders too late could lead to overpaying or even lacking vital components or business essentials when needed.

Predictive procurement uses the power of machine learning and data analytics, drawing on historical usage to help organizations identify what they have spent in previous years, months or weeks, and anticipate future demand.

In time, tools could factor in wider changes in demand, shifts in buying habits such as the use of more sustainable items, or even anticipated disruption to supply chains caused by weather or other events.

The adoption of such technology has been growing steadily in recent years, and 60 percent of procurement professionals currently use procurement analytics or reporting tools, according to Amazon Business’ 2024 State of Procurement Data Report.

Amazon itself is already making use of machine learning, using neural network models to anticipate demand for products, helping ensure it can meet orders and deliver within tight timeframes.

What are the main benefits that come with predictive procurement? 

Armed with such information, procurement professionals can plan exactly when to buy items and devise a strategy. This could involve running a tender exercise, where suppliers are encouraged to put forward their best prices in exchange for a volume commitment, or seeking to spread orders over different suppliers to reduce the risk of single-sourcing.

With greater knowledge of their requirements, buying organizations have the potential to negotiate discounts and build stronger relationships with suppliers. This could help position them as customers of choice should supply be limited.

Having a clear strategy of what needs to be bought and when and with which supplier can also help internal customers better manage their own requirements, as long as this fits within agreed parameters. Three in 10 procurement professionals want to decentralize purchasing so others can more easily buy items for their teams, the State of Procurement report suggests.

Making more use of machine learning can also yield savings, with the potential to reduce costs in the supply chain function by between 35 and 65 percent.

How can predictive procurement help organizations manage risk?

At a time when 68 percent of procurement professionals are focused on developing supply chain resilience, according to Deloitte’s 2023 Global Chief Procurement Officer Survey, having additional insight into the future could prove invaluable.

From a supply risk perspective, such information can help ensure there is less risk that a business will run out of items, particularly if it builds in a buffer, so slightly more stock is purchased than anticipated.

This will also reduce financial risk by helping organizations avoid unexpected price increases, which can occur if inflation rises or orders are placed with tight turnaround times. Buying early and in bulk can give buyers more negotiation power, lowering prices.

Conversely, buyers may wait until later to place an order, knowing they have sufficient stock to use first. This can bring about cash flow benefits, from lower prices should demand or costs fall. Armed with information about their needs and when businesses can make better calls regarding potential risk and reward.

What are the potential concerns about predictive procurement?

Like any technology, the information that comes out will only be of value if the data going in is correct. Hence, businesses understand what they have bought historically to make accurate predictions. Where this still needs to exist, it may be a better strategy to build up this information over the next year or two before attempting to forecast expected future demand.

Tools such as Amazon Business Analytics and Spend Visibility can help Amazon Business customers identify what they are spending and with which suppliers. This information can then be used to make decisions that create more value for money and provide the system for building predictive procurement capabilities.

Being aware of macroeconomic trends and pricing forecasts, however, can help procurement teams with their forward-planning process, and assist them in determining when a good time to buy would be.

It’s also essential to ensure human involvement in any process. Machine learning can identify patterns and trends, but only humans can make strategic decisions that will benefit the business.

How can organizations take advantage of predictive procurement?

According to the Amazon Business’ State of Procurement Report, 98 per cent of decision-makers are already planning investments or upgrades in analytics and insights tools, automation and AI-driven optimization of purchasing decisions in the next few years, so this is very much on procurement’s radar.

The challenge for procurement is to pick the right time to invest, at a point when the technology is established enough to be proven but not so late that it will lose its competitive advantage.

Amazon Business’s smart business buying tools can help organizations get started with predictive procurement. It can anticipate reorders and predict what they’re likely to need, helping to reduce inventory costs by as much as 20 per cent. It can also provide the wider management information that’s required for making more long-term, strategic decisions.

Originally published in Business Reporter.