Using AI in procurement to reduce millions of dollars and supply chain risk

Tariffs are placing tremendous pressure on procurement professionals. Given they are present, AI-assisted procurement can help offset the increased costs. In addition to reducing spend, AI also enables organizations to maximize their use of competitive bidding.

Recently, I worked with an organization that wanted to increase competitive bidding while getting a better handle on costs. This company had several billion dollars of spend, annually. World-class manufacturers typically competitively bid 80-90% of their annual purchases, but this manufacturer only competitively bid about half of that. The company was almost certainly paying higher costs, and the suppliers may not have been offering the best service or quality.

An organizational change was identified: increase the level of competitive spending. The goals were to drive cost savings while reducing supply chain risk and upgrading the maturity. To do so, we piloted an AI platform, which makes procurement faster and more predictive, using Arkestro. The initiative aimed to increase competitive bidding, reduce costs and supply chain risk, and drive a culture shift.

Arkestro was chosen because the app provides a suggested price (based on historical data, supplier behavior, and market conditions) during an RFP, pushing suppliers toward more competitive pricing and reducing the time. The app automates a lot of the manual work in sourcing and bidding, like inviting suppliers, collecting bids, and comparing them, so companies can receive bids in days, not weeks. Other Arkestro benefits include scalability, customer service, and ease of use.

Approach: Pilot using a Standard Process

The two-month pilot involved a few buyers and category managers with open mindsets and a bias for action. Select categories were identified based on the impact, annual spend, and whether the category had historically been competitively bid. These categories – which totaled more than $1 billion in annual spend but historically had limited competitive bidding – included raw materials, heavy equipment, transportation, and MRO.

We launched the pilot with a two-day training. In less than 2 hours, we had set up a Request for Proposal in the tool, which is the first step in the three-step process. This included identifying the items, suppliers, quantity, historical baseline price, and any specifications.

Next, the RFP was launched. Suppliers were invited to participate via email. With just a few clicks in the app, suppliers could see the RFP details, including the suggested price. Suppliers could accept the suggested price or enter a new price along with other information. The final step includes analyzing the bids, selecting the supplier, and follow-up communications thanking all suppliers .

The actual work for a procurement professional is about 45 minutes. Assuming the RFP is two rounds, each supplier can see their rank compared to other suppliers; the transparency of suggested prices and rating increases faster agreements. Suppliers have a defined time to review the RFP and accept the price or suggest alternate pricing. Overall, it automates and expedites the process, using transparency to build trust in the system.

We held weekly 1:1 meetings with the pilot team members to measure progress, resolve issues with Arkestro’s help, increase engagement, and allow for the right mindset and behavior changes. One of the buyers who had never competitively bid ended up doing an RFP every week and captured at least 20% cost savings every time..

Outcomes: A Competitive, Data-Driven Procurement Function

We captured more significant savings than expected, though the specifics varied by category.

• About 20% for MRO
• About 20% for IT peripherals
• About 8% for raw materials
• About 4% for transportation

We estimated that we would save $40 million-plus annually for these categories alone.

Even in categories that were heavily impacted by tariffs, we were able to realize significant cost savings. For example, one raw material was sourced from China, and with a 32.6% reciprocal tariff, the company’s $3 million annual spend was in danger of ballooning to nearly $4 million. When competitively bid, the company reduced its annual spend to about $2.3 million.

Three excellent and unexpected things happened:

1. Supplier Participation: About 75% of the suppliers participated, which was much higher than expected because many of the suppliers have lower maturity in technology.
2. Time Savings: For most categories, it took 10-20% less time. Transportation saw the most astonishing improvements. The incredibly complex category – involving dozens of plants, hundreds of carriers, and thousands of lanes - went from 0% competitively bid to 100% in a 1/10 of the time it would have normally taken.
3. Impact: The suggested price works. For transportation, about 60% of the bids were below the suggested price, and 100% were below historical pricing. The tool is transparent; suppliers can see their position relative to other suppliers after the first round.

With these results, the company decided to integrate the tool across all of its plants, distribution centers, and procurement professionals.

Results: A Competitive, Data-Driven Procurement Function

The impact was both immediate and far-reaching:

• Tens of millions in cost savings captured in the first year
• 60% of bids were lower than historical pricing
• 60% of incumbent suppliers were retained, which ensured supply chain stability
• The company was able to mitigate the impact of tariffs
• A more aligned and engaged procurement team with a continuous improvement mindset

Want to continue the conversation?

Contact NorthLawn to explore how NorthLawn's AI Practice can support your organization's goals.