Read a sample: 1. Why Estimates Go Wrong
On Monday morning, your manager asks a question that sounds simple: “How long will this take?”
You picture the main task, make a quick mental calculation, and offer an answer. Your manager thanks you and moves on. By Wednesday, you are waiting for information you did not know you needed. By Thursday, a reviewer has requested changes. The number that seemed reasonable on Monday now feels like a promise you cannot keep.
The problem is not necessarily that you are careless or bad at managing your time. You may have estimated the work you could see while leaving out work that had not yet come into focus. You may also have answered a different question from the one your manager thought they were asking.
A useful estimate is not a test of whether you can predict the future. It is a forecast based on what you know, what you assume, and what remains uncertain. Its purpose is to help people make decisions: when to begin, whom to involve, what to prioritize, and when to revisit the plan. An estimate can be valuable even when the eventual result differs, provided it makes the reasoning visible and improves as the work becomes clearer.
This chapter begins with the distinction that supports the rest of the book: an estimate, a deadline, and a promise are related, but they are not the same thing.
An estimate is your current judgment about how much work an assignment may require or when it may be completed. It is conditional on the information available when you make it.
A deadline is a date or time by which something is wanted or required. It may come from a customer commitment, a scheduled event, a business decision, or another constraint. A deadline tells you when an outcome is needed; it does not, by itself, establish how much work is involved.
A promise is a commitment to deliver an agreed result. A responsible commitment considers the estimate, the deadline, the scope, and the resources available. It also makes clear what happens if important assumptions do not hold.
Consider these three statements:
- “Based on similar work, I estimate this will take about two working days of effort.”
- “The team needs the report before Friday’s meeting.”
- “I will deliver the agreed report by Thursday afternoon.”
They may sound compatible, but they describe different things. Two working days of effort might fit before Thursday afternoon if the person doing the work has enough available time, receives the necessary information, and can obtain review promptly. It might not fit if those two days are already filled with other commitments or if an essential approval takes several days.
The distinction matters because workplace conversations often slide from one statement to another without anyone noticing. A manager asks for an estimate, hears a number, and puts a date on a schedule. The person doing the work hears the date repeated and assumes they have committed to it. Neither intended to create confusion, but the uncertainty has disappeared from the conversation.
You can prevent some of that confusion with a short clarification:
> “My current estimate is about two working days of effort. I still need to check when I can get the source information and how quickly the review can happen before I can forecast a delivery date.”
That answer is not a refusal to commit. It separates what you can judge now from what you still need to establish. If the work must be finished by a particular date, the next conversation is about whether the scope, available time, and dependencies make that date workable.
Sometimes a commitment must be made before every detail is known. That is normal. The goal is not to postpone decisions indefinitely; it is to make the uncertainty visible enough that the people making the decision understand the risk.
Throughout this book, we will follow a hypothetical workplace assignment. The example is deliberately ordinary. Estimation problems do not require a large, complex project. They appear whenever a request contains more decisions and dependencies than its first sentence reveals.
Maya is an operations specialist at a growing company. Her manager, Daniel, asks her to prepare a monthly customer-support report for a leadership meeting. He wants a concise summary of recent support activity, notable issues, and suggested areas for attention.
At first, Maya imagines opening the support system, exporting a few numbers, writing a page of observations, and sending the report to Daniel. She has prepared internal summaries before, so the request feels familiar. When Daniel asks how long it will take, she is tempted to say, “One day.”
That answer may eventually turn out to be reasonable. But before treating it as a plan, Maya needs to know what the day includes.
Does Daniel want a written document, a slide, or both? Which support activities should be included? Is there an existing definition of a notable issue? Can Maya retrieve the information herself, or does she need help from someone who manages the support system? Who will review the report, and how many rounds of changes are expected?
Maya does not need to solve every question immediately. She does need to recognize that each unanswered question can affect the forecast. A one-page report may be quick to write but slow to prepare if the source information is inconsistent or the intended audience has not agreed on what it…