The visible problem is rarely the problem

Most bad solutions are good solutions to the wrong problem. Somebody describes a symptom, everyone starts fixing, and three weeks later the symptom returns.

A complaint is not a problem statement. *"The website is slow"* could mean the server, the images, one page, one person's connection, or that a client said so once. Until you know which, anything you do is a guess with a budget attached.

Write the problem as an observable fact. Who is affected, what exactly happens, when, and how often.

> Weak: *customers are unhappy with delivery*

> Usable: *in the last month, 14 of 60 orders to Karachi arrived more than 3 days late, and 9 of those customers complained*

The second version already suggests where to look and tells you when you have fixed it. The first cannot be solved because it cannot be measured.

Separate the problem from the solution somebody has already chosen. *"We need a new courier"* is a solution wearing a problem's clothes. The problem might be the courier, or the packing time, or that orders are placed after the daily pickup. Accepting the framing hands you somebody else's assumption.

Ask who says it is a problem. Sometimes one loud voice is standing in for nobody. Sometimes a quiet, expensive problem is being ignored because nobody complains about it. Frequency and cost matter more than volume.

And ask what happens if nothing changes. Some problems genuinely do not need solving, and saying so is a legitimate answer that saves everyone's time.