Negotiated Away: Revenue Teams Are Leaking Margin Their Pricing Models Already Caught

Most revenue leaders can tell you their average discount. Far fewer can tell you how much margin left the building on deals that closed exactly as forecast. That second number is the one that matters heading into fourth-quarter close, when deal desks are busiest and the pressure to sign is highest. A growing body of evidence says the money is not slipping away because companies lack pricing intelligence. It is slipping away because the intelligence never makes it into the negotiation, and nobody is measuring what happens after the contract is signed.

The Leak Happens on Both Sides of the Signature

In an August analysis, L.E.K. Consulting partners Robert Haslehurst, Stephen Matthews and Vassilis Economides argue that B2B companies lose margin on nearly every deal through prices set below what customers would pay, usage that never gets billed and concessions that outlive their purpose. Before signature, the leak shows up as discounts used to close the quarter. After signature, it shows up as unbilled overages and renewals that quietly carry forward last year's concessions. Citing MGI Research, the firm estimates that post-signature leakage alone costs at least 3% to 5% of revenue.

Technology companies are especially exposed, the authors note, because a single customer relationship can now span subscriptions, usage, services, credits, channel partners and renewals. Every one of those is a place where a price can be set once and eroded many times. And because most of the erosion is small and spread across hundreds of transactions, it rarely surfaces as a line item anyone owns.

The problem is not new, and it is not limited to software. A study of industrial companies by Bain & Company and Pricefx found that businesses lose over 6% of revenue, on average, through hidden discounts and off-invoice leakage such as rebates and unrecovered freight. For a company with $1 billion in revenue, that is a $60 million opportunity. The authors stressed that many of these concessions are not controlled by salespeople at all and are visible to very few people inside the business, which is precisely why they persist.

The pricing model was right. The margin was lost in the hand-off between the team that set the price and the team that negotiated it.

Two Systems That Never Talk to Each Other

L.E.K.'s central diagnosis is structural. Most companies run what it calls a "pricing brain," owned by pricing or finance, which produces segments, discount limits and recommended prices. Separately, they run a commercial execution layer, owned by sales and customer success, which covers enablement, deal desks, partners and renewals. The two share neither data nor reasoning. Sellers often negotiate without pricing input, and deal-desk approvals and renewals happen without the logic behind the recommended price ever reaching the person signing off on the exception.

That gap is why many AI pricing investments disappoint. L.E.K. points to reporting that only about a fifth of companies using AI see clear, measurable value from it, and argues that pricing tools often fail because they are bought separately from the execution layer, so sound recommendations get routed around. As the authors put it, "A modest, well-connected model will outperform a sophisticated model that sellers route around."

When the recommendation does reach the seller, the effect is meaningful. L.E.K. cites a peer-reviewed field experiment in which sellers shown an AI-recommended price during live quoting closed deals at roughly 11% higher profit than colleagues without it. The recommendation did not replace the seller's judgment. It anchored it. Of the four moments where the firm sees AI helping pricing, from designing how to charge through protecting value after signature, target pricing and live deal guidance currently have the strongest evidence.

Price Increases That Never Fully Land

The same execution gap shows up when companies raise prices. Simon-Kucher's State of Pricing 2025, drawn from more than 2,200 business leaders across 28 countries and 39 industries, found that companies realize less than half of their average price increases. A list price change is announced, and then discounts, exceptions and renewal negotiations give a large share of it back. For CROs, that means the revenue plan built on a price increase is often overstated before the first quote goes out.

The common thread across all three sources is visibility. Leakage persists where nobody sees the full path from recommended price to invoiced price, and where the people approving exceptions cannot see why the original number was set. L.E.K. frames governance, meaning who sees the reasoning, who approves overrides and where exceptions are logged, as mattering as much as the algorithm itself. "A recommendation a seller can explain is one they will defend," the authors write.

The CRO Playbook for Closing the Margin Gap

• Measure the gap, not just the discount. Track the difference between recommended price, quoted price and invoiced price by segment and by rep. Average discount hides where margin actually leaks.
• Put the reasoning inside the quote. Surface target prices and the logic behind them in CRM and CPQ at the moment a seller builds a quote, not in a pricing deck reps never open.
• Make every exception leave a trail. Require a logged reason for each discount beyond guardrails and review the log monthly with pricing and finance so patterns become visible.
• Audit renewals for inherited concessions. Before each renewal, flag discounts and credits that were granted for reasons that no longer apply, and reset them deliberately rather than by default.
• Qualify harder before you discount. Deals that need deep concessions late in the quarter are often deals that were never well qualified. Fix the pipeline upstream instead of buying the close downstream.
• Find out whether price really lost the deal. Use buyer interviews on wins and losses to test whether discounts changed outcomes, or simply gave away margin on deals you would have won anyway.

Revenue teams have spent years investing in better pricing models, and the evidence suggests the models are largely doing their job. What has not kept pace is the plumbing that carries a good price from the model to the contract and keeps it intact through renewal. The CROs who recover margin in 2027 will not be the ones with the most sophisticated algorithm. They will be the ones whose sellers can see the right price, explain it and defend it all the way to signature.