Arkestro

Arkestro is a predictive procurement tool that suggests a fair price before a sourcing event starts, using machine learning on past pricing, market, and category data, rather than waiting for supplier quotes first. Buyers use that price to run automated, multi-round negotiations in which suppliers get feedback on how their bid compares and can revise it, without a buyer managing each round by hand. It integrates with existing source-to-pay suites and ERP systems rather than replacing them. It is made by Arkestro, which reports customers including Chevron, Koch Industries, and GAF and cites average savings of 18.8% across spend categories. Pricing is based on the volume of spend routed through the platform, not a per-user fee, and is not published.

This fits a procurement team that already runs a source-to-pay system and wants to speed up negotiation on recurring or moderately complex categories without adding headcount. A sensible pilot is one spend category with several qualified suppliers and a history of price data, so the predictive pricing model has something to learn from. Because pricing is tied to spend volume rather than seats, ask Arkestro directly how the fee scales as more categories are added, since public numbers are not available. The reported 18.8% average savings and 60% shorter negotiation cycles come from Arkestro's own case studies and are not independently audited, so validate them against your own category data during a pilot. It works best as an add-on to an existing sourcing process, not a replacement.