A manufacturer's rep spent nearly six hours building a complex quote, pulling specs from three different sources, checking compatibility by memory, and calculating freight manually. The quote went out. It had errors. The customer noticed before the rep did. That is one kind of quoting problem: time lost, credibility dinged, deal delayed. There is a second kind that gets less attention. It shows up in the next conversation with that same customer.
Most post-mortems on a bad quote focus on the immediate damage. The margin was off. The revision took two days. The rep had to call the customer and explain why the number changed. Those costs are real and measurable, but they are also the easier ones to see.
The second cost is behavioral. A customer who received a quote that did not match their invoice, or who watched a price change between proposal and close, starts the next conversation differently. They ask more questions. They push harder on price because they are not sure the first number you give them is the real one. They take longer to sign because they want to wait and see. None of that shows up in a CRM field, but it shows up in cycle time and close rate on the follow-on business.
The first deal costs you margin. The second deal costs you trust, which is harder to rebuild than margin.
This is worth being specific about, because "the quote was wrong" can mean four or five different things.
Manual pricing lookups. A rep who calls the inside sales team, or cross-references a spreadsheet, or asks the one person in the company who knows the bolts pricing is introducing a transfer step where errors live. Every handoff between a rep and a pricing source is a place where the number can come back wrong, stale, or misapplied.
Bundle logic held in memory. In many manufacturing shops, the person who knows which components go together and what they cost as a package is a single individual. If that person is out, on a call, or just hard to reach, the rep makes their best guess. Best guesses on complex bundles produce inaccurate quotes.
No approval gate on custom discounts. When a rep can type any number into a discount field, the quotes that go out at 30% off are not necessarily the ones leadership approved. The deal closes. Finance looks at the margin. Someone asks how that happened. The answer is usually that nothing stopped it.
ERP data not available at quoting time. If your material costs shift and your price book in your quoting tool does not update to match, reps are quoting from stale data. The gap between the quoted price and the actual cost lands on the invoice, or on ops when they discover the margin problem after the order is in.
Each of these is a specific failure mode with a specific fix. They share a common structure: the rep is making a decision that should be handled by a system.
Margin leakage from inaccurate quotes happens in three places.
The first is the direct gap between what was quoted and what was invoiced. If the quote was built on a price book that has not been updated in six months, and material costs moved, the invoice reflects a number that does not hold the margin the deal was supposed to carry.
The second is the revision cycle. A quote that goes back and forth two or three times before the customer signs is a quote that delays close. Sales cycles that stretch because the numbers keep changing are not just annoying. They are pipeline velocity problems.
The third is the discount that gets added to close a deal that should have been clean. When a quote has a visible problem, the path of least resistance is often to discount rather than reopen the negotiation. The deal closes. The margin does not.
The goal is not a system that makes it impossible for anything to go wrong. The goal is a system where the error vectors are known and controlled.
That means pricing logic lives in the tool, not in a person's head or a spreadsheet. When a rep builds a quote, the numbers come from a source that reflects current costs. If a price changes in your ERP, it changes in the quote tool. The rep does not have to call anyone to find out what to charge.
It means configuration rules are enforced before the quote goes out, not discovered when ops reads the order. If two components are incompatible, the system does not allow the combination. If a bundle requires a specific add-on, the system prompts for it.
It means approval workflows exist for the cases that need them. A standard deal at standard pricing goes straight to the customer. A deal with a non-standard discount routes to a manager. The rep does not decide which category applies. The system does.
The aspiration behind all of this is simple: the number on the proposal is the number on the invoice. When that is consistently true, customers stop asking whether they can trust your quotes. That is the second deal coming back.
If you are looking at how automation fits into this, automated quoting walks through how the mechanics work.