TL;DR
- Procurement scorecards track savings but rarely count what it cost to get them; a 4% win on a $10M deal can still mean hundreds of thousands in value lost to weeks of manual effort.
- The real culprit is "coordination burden" or the time spent hunting for contract versions, PO history, and chasing other departments instead of doing actual procurement work.
- AI's biggest near-term value isn't negotiation or judgment calls; it's eliminating the painful, repetitive coordination tasks humans dread.
- Target the most painful coordination gap between procurement and another function, not the flashiest use case.
- Waiting for perfect data before deploying AI is often a mistake; match your data-quality bar to the stakes of the decision, and start where imperfect data is tolerable.
Procurement teams are good at measuring what they save, but less good at measuring what those savings cost to achieve. That gap between reported performance and total operational reality is where a significant amount of value quietly disappears, often without anyone noticing until the damage is done and the opportunity is in procurement’s rearview mirror.
Mike Fedorov, Chief Operating Officer and Co-CEO of Applied AI Labs, has more than 25 years of hands-on operations and supply chain experience that he shared in a recent episode of The Sourcing Hero podcast. His perspective on where procurement loses value is sharp and practical, filled with some hard yet necessary truths for every procurement team to consider.
The $10 Million Negotiation That Wasn't as Good as It Looked
A company negotiated a $10 million contract and saved 4%, which looks like nearly $1.5 million on the scorecard. The team celebrated, leadership was satisfied, and the metric moved in the right direction.
But what did it cost to get there? How many hours did the procurement team log chasing stakeholders, digging through systems, or reconstructing the history of the last negotiation? If the answer is eight weeks of intensive work, and the savings were being realized at a rate of $100,000 per week, then $800,000 in potential savings evaporated while the negotiation was still in progress.
But, not surprisingly, that sobering number does not appear anywhere on the scorecard.
There is another question worth asking: was 4% the right outcome? “Did we take too much money from the table?” as Mike puts it. A supplier who accepts a deal feeling like they lost may not walk away, but they will remember it at renewal time. They may also quietly deprioritize the account in ways that create operational problems.
Then there is the opportunity cost. Every hour procurement spends on a transaction is an hour not spent on changing the terms of how the business operates. That rarely shows up in any performance report; instead, it accumulates over time and becomes a significant drain on what procurement can accomplish.
The Coordination Burden
The reason these hidden costs exist, Mike argues, is that procurement is an unusually coordination-intensive function. They touch nearly every part of the business, which means a disproportionate share of a procurement professional's time goes into strengthening the connective tissue between teams rather than into the work itself.
There’s a difference between coordination that is genuinely valuable (the human contact, the relationship-building, the context-gathering that makes good decisions possible) and what he calls “coordination burden.” That is the time spent hunting for the latest version of a contract, digging through an ERP for purchase order history, or chasing someone in finance for information that should have been easy to find. This is not what procurement was hired to do.
“A great procurement person is very good at negotiating and understanding the market and understanding the context and figuring out where the value is."
— Mike Fedorov, Chief Operating Officer & Co-CEO, Applied AI Labs
The coordination burden is what stands between them and that work, and it is also what fills the urgent pile to the point where important work never gets done.
Can AI Lighten This Burden?
The coordination burden is precisely the problem set that AI is now capable of addressing. Not because AI can negotiate, or make judgment calls, or read a room, but because it can do the things procurement professionals least want to do and do them well.
Finding the latest contract version, collating purchase order history, pulling together the context for a renegotiation: these are tasks that AI handles capably and that humans find genuinely painful. As Mike puts it, the relationship between human and digital capacity should be complementary, not adversarial. It is, in fact, the whole point.
For teams figuring out where to start, Mike's advice is to resist the temptation to begin with the most ambitious use case. Start with the most painful coordination problem that creates gaps between procurement and another function, somewhere that value is visibly leaking and where a fix would be both meaningful and achievable within a reasonable timeframe.
How Important Is Data Readiness?
Data quality is an important part of this process that stops a lot of teams before they even get started. The instinct to clean everything up before deploying AI is understandable, but Mike pushes back on that. Waiting for perfect data may mean waiting forever, and it misunderstands what data quality costs in different contexts.
For example, in high-stakes automated decisions, clean data is non-negotiable, but for a use case like preparing a negotiator with context before a supplier conversation, the stakes of imperfect data are much lower. If the AI pulls together 95% of what is needed and gets one thing slightly wrong, the negotiator will catch it.
The result is still far better than doing all of that preparation manually, and the cost of the error is manageable. Starting with use cases that can tolerate current data conditions, and building toward the ones that require more rigor, is the most practical way to start.
For the full conversation, listen to Mike's episode here:
