Stop defending a number. Present a range.

"Sure, your product saves time. But how much time? For us?"

If you sell into a buying group, you have heard a version of this. It is the moment the conversation leaves the product and starts on the number. Defend the number and you sound like a pitch. Hedge it with no model behind the hedge and you sound like you do not know.

There is a better object to put on the table: a range, and the assumptions that produced it.

Buyers are walking in with a short expected payback and a higher bar for the case. That is their expectation. It is not a cycle time Serre claims to hit.

What the URL is named after

Monte Carlo is one way to explore combinations of assumptions: draw each input from a range, many times, and look at the spread. The name of this page is that technique, because that is the search and the old argument.

It is not required, and a run count is not evidence. Bounds that hug the expected value read as a wish. Bounds that span an order of magnitude read as a shrug. Either problem is visible without simulating it. What the reviewer needs is the low, the expected and the high, each tied to an assumption they can change.

Serre does not sell a simulation count. The arithmetic is deterministic. The same inputs produce the same result, and narrative text does not touch the sum.

Why a range is easier to review than a point

When you present one ROI number, a finance reviewer's instinct is to find the optimistic assumption. When you present a range, the conservative figure is already the answer to "what if this does less?"

The conversation can move to the inputs. "We think the affected headcount is closer to 35 than 60." That is a discussion about assumptions, and the formula stays visible so any output can be walked back to them.

The goal is not a larger claim. It is a claim that shows its bounds. A range that admits uncertainty is more reviewable than a point that pretends uncertainty is gone.

What the buyer should be able to see

They should not have to read a thousand rows. They should see three readings of one model:

• Conservative. The downside set of assumptions. This is the reading a skeptical reviewer will start from. It still has to be labeled as a projection.
• Expected. The middle set. It is not more true because it sits in the middle. It is the set you are willing to call most likely, and they can disagree.
• Upside. The favorable set. It is there so the hopeful reading is not smuggled into the expected case.

Each reading traces to bounds someone chose. If the expected case is questioned, walk the inputs. "This assumes 3 to 7 hours a week, with 5 as the most likely. Does that match the work?" Hours valued at a labor rate are a valuation of capacity, not a cash saving, until a named capture action and an owner exist. Do not add the operational, budget and financial readings of one benefit together.

Stop performing certainty

The question "how much value can you actually deliver for us?" does not have a single right answer. Pretending it does is what makes the conversation adversarial. The buyer knows the number is a projection. You should say so in the case.

A range says: given the bounds we can both see, here is the spread. That is not a weaker claim. It is the claim that can be checked. Value engineering for every seller means the rep can put that model in front of the buyer without waiting on a specialist. A person still approves what enters the library. Turn every deal into a defensible case for change. The change is still a projection, not a result.