
Recently, I worked with a large sales team that showed me something they’d built: an AI tool that watches their entire sales pipeline. One feature keeps tabs on their renewal book. It pulls from their CRM, reads the conversations logged against each account, scores which renewals are at risk, and tells the team where to push for changes. It’s genuinely impressive work, and they were right to be proud of it.
Interestingly, I’ve had similar experiences with procurement organizations. It’s becoming more apparent that this renewal season will be unlike others, as both sides of the table now have a machine watching the clock and proactively providing guidance.
The Demand Arrives Before the Opinion
On the buying side, procurement teams have spent this year deploying AI agents that do work nobody had the capacity to do. These systems sweep contracts out of inboxes, drives, and legacy folders. They extract terms, pricing, and notice windows. They build a renewal calendar, cross-reference expiring contracts against category spend and supplier performance, and propose negotiation levers automatically. Industry analysts estimate these agents now handle 60% to 80% of routine procurement work.
On the sales side, revenue intelligence platforms are doing the mirror image. They correlate conversation data from calls and QBRs with account health trends and usage signals to predict renewal risk months before anyone schedules the renewal conversation. Gartner projects that by 2027, 95% of seller research workflows will start with AI, up from less than 20% in 2024.
Set aside whether any of this is good or bad. Here’s what it means at the table.
For most of my career, when the other side opened a renewal with a demand, a person had decided to make it. They had a reason, a target, and internal pressure behind them. Their opening move carried information, and a good negotiator could work backward from it to find out what they actually needed.
That’s no longer a safe assumption. On both sides, the opening move is increasingly generated rather than decided. It shows up because a system flagged a variance, not because a human weighed the relationship and reached a conclusion. The demand arrives before anyone has formed an opinion, which creates different risks, depending on where you’re sitting.
If you’re on the receiving end, you can find yourself conceding to a position the person across from you doesn’t actually hold. If you’re on the sending end, the risk is subtler and, I’d argue, worse. You can walk into a negotiation defending a number your own system produced, which you never stress-tested and can’t explain when someone asks you why.
Here are 3 ways to handle it, whichever chair you’re seated in.
1. Find the person behind the position.
A flag isn’t a position — it’s an output.
In the deals we work on, no system emails you a demand. A person does. But, increasingly, that person is carrying a number they didn’t author. It came out of an analysis and landed on their desk with a target attached. They may not have stress-tested it, defended it internally, or thought about what they’d give up to get it.
The trained reflex is to counter. Resist it. Your first job is to find out how much of that position is actually theirs. Ask who else is affected if the scope changes. Ask which stakeholder signed off on the target. Ask what happens on their side if this doesn’t close before the notice deadline. The answers tell you whether you’re facing a considered stance or a printout with a person attached.
That distinction changes how you respond to everything that follows. Treat an output like a hardened position and you’ll concede against a demand nobody was prepared to fight for. You’ll also teach the other side that flagging your contract produces movement. Do that twice and you’ve trained both the system and the team behind it.
The same discipline applies when the output is yours. Before you carry a number into the room, know what you’d trade to get it and what you’ll say when they ask you to justify it. If you can’t answer either one, you’re not ready to sit at the table. You’re just delivering someone else’s conclusion.
A system can generate a demand. Only a person can trade. Find the one who can.
2. Make sure your value survives extraction.
This is where most people are exposed — and it’s new. These systems build their picture from what’s in the data, not from what actually happened. If your team absorbed two escalations at no charge and never invoiced them, they don’t exist in the repository. If your service level consistently outperformed the contract, but the contract records only the minimum, then the minimum is your record.
It cuts both ways. If you’re the buyer, and the supplier missed delivery windows all year but nobody logged it or claimed the credits, those failures aren’t in your system either. Your leverage evaporated the same way theirs did.
So, before renewal season starts, do an audit. Ask what evidence exists in a form that the other side’s systems can read: documented, dated, quantified, and priced (ideally) even if it was credited to zero.
Anything you delivered but never recorded isn’t generosity — it’s just gone. And anything they cost you that nobody logged isn’t a grievance — it’s a story.
3. Trade against the analysis instead of arguing with it.
You’ll not win an argument with a benchmark. The data is the data, and telling someone their comparables are wrong just starts a debate neither of you can settle.
Trade instead. A benchmark reflects a standard offering under standard terms, so if they want the benchmark price, they can have the benchmark deal. Something like, “If you need to be at market pricing, then I’d need to move you to standard terms on response times and volume commitments. Would that work?”
The same move works in reverse. If a supplier’s model says you’re a low-risk, high-dependency account, and their proposal reflects it, don’t argue with their read. Ask what they’d need in order to price you differently, then decide whether you’re willing to give it. Either way, you’re negotiating instead of defending. And you’ve moved the trade-off in front of a human, which is exactly where you want it.
The bottom line.
None of this means the fundamentals changed. Preparation still wins. Trading still beats arguing. Knowing what you want and what you’ll give up still separates the professionals from everyone else.
What changed is how we’re analyzing data and developing positions. A machine can flag a renewal or highlight a recommended change. It can’t trade. That part is still yours.
How is AI changing contract renewal negotiations?
AI is changing contract renewal negotiations by identifying risks, analyzing contract and account data, and recommending positions before negotiations begin. That makes it increasingly important for negotiators to understand who owns a position, document value, and know what they are prepared to trade.
Should I respond to a data-driven price challenge with my own data?
Data alone rarely resolves a price challenge because both sides can produce numbers supporting their position. Use their analysis as the starting point for a trade instead. If they want a different price, offer different terms and let them decide which they value more.
How early should I start preparing for a contract renewal?
Earlier than you used to. If the other side’s systems flag expiring contracts a quarter out, your preparation needs to begin before that flag fires. Otherwise, you’re reacting to their timeline instead of setting your own.
Does the other side using AI mean I should use AI in contract renewals too?
Not necessarily. AI can help organize contract data, identify patterns, and prepare for renewal negotiations. But it can’t replace the human judgment required to test positions, make trades, and manage relationships.
As AI works its way further into our day-to-day, the ability to discern earned knowledge from AI-borrowed knowledge is essential. Rely on Scotwork’s expertise to help you push for the truth.
By Brian Buck | 28.08.26
This past week, we lost Dolly Parton. I have strong memories of riding around in…
By Kirbi Isaac | 07.08.26
A friend and I recently went out to dinner and had a great meal. We…
By Trevor Murphy | 31.07.26
“I have a great idea.” In a marketing agency, this statement comes up in many…