Recovering 15% without losing the relationship

Context

On a lump-sum job, change is where the contract quietly stops meaning what it says. Requests for information arrive in volume, a fraction of them carry a cost impact, and each one is a small negotiation conducted under time pressure by people who mostly want the work to keep moving.

Over the course of the project the change channel ran to more than a thousand RFIs. The question was never whether change would arrive. It was whether it would be priced or absorbed.

The mandate I gave myself

Not to review the contractor’s number. To produce my own.

A review argues about somebody else’s basis — their quantities, their productivity assumptions, their markup. You end up negotiating a discount off a figure you didn’t author, which means the starting point was set by the party with the most to gain from it. An independent estimate replaces that figure rather than discounting it, and it changes what the conversation is about.

So I ran a check estimate myself on every RFI that carried a cost impact.

The decisions

Estimating independently, then negotiating to a number both sides could accept

The check estimates were the position, not the opening bid. Where mine and theirs diverged, the gap was a question about a specific assumption — a quantity, a rate, a duration — rather than a disagreement about the total. That is a solvable conversation.

We settled at numbers both sides found acceptable. That was the objective. A settlement the contractor resents is a settlement you pay for later, in cooperation you no longer get.

Letting the record do the arguing

We ran Procore on site, and it was used properly. That meant the project had a contemporaneous record of what was asked, when it was asked, who responded, and how long they took.

Entitlement arguments are usually arguments about memory. This one wasn’t. When a delay position depended on who had held a decision up, the answer was in the log — including the occasions when the answer was us. A record that can embarrass both parties is the only kind either party will accept.

Knowing the contract better than the negotiation required

Delay damages and performance incentives were built into the agreement before any of this started. That matters more than it sounds: if the contractor did not meet the dates, the incentives were simply not earned. No argument was necessary, because the mechanism already existed and both parties had signed it.

Where delay events did arise — including the disruption everyone in the industry absorbed in 2020 — the position came from forensic analysis against the contemporaneous record and the contract terms, not from a negotiated view of what felt fair.

Reading the entire drawing set

The largest single exposure was the gap between the issued-for-bid drawings and the issued-for-construction set — the design development that happens between the price and the build. Holding a position on that requires knowing exactly what changed, which requires reading every sheet.

There is no shortcut. Somebody has to do the comparison, and the party that has done it sets the terms of the discussion.

Outcomes

Total project cost reduced by 15%, through renegotiated change orders and a rebuilt change lifecycle that closed the approval gaps allowing scope to enter unpriced.

The contractor kept working with us on the same terms afterwards.

What I carry forward

Facts preserve relationships. Positions destroy them.

Everything above was grounded in evidence — an estimate I could show my basis for, a record neither party could dispute, and contract terms both had already agreed. None of it required anyone to lose an argument in front of their own management, which is usually what makes commercial disputes expensive.

You are not taking money back. You are pricing the work correctly and showing your work.