I watched a two-minute clip from The Diary of a CEO this week and it has stayed with me. Two stories, one idea.
The $12,000 floor
Adam Neumann, the WeWork co-founder, tells it. They were fitting out their second building and needed to lay the floor. They had $90,000 left. Every quote for the job came back at around $200,000, and the floor alone was priced at $30,000 to $45,000 by every contractor they asked.
So he asked why. Not "can you do it cheaper", but why. Break the quote down: how many pieces of flooring, how many nails, how much glue, how much rubber, how many hours of work per person, and what each person costs.
Nobody could answer. The number was not built from anything. It was the number everyone quoted because it was the number everyone quoted.
He then called the top three contractors and offered them a deal: you earn 15% above cost, which is a fair margin, and I will price every piece myself. They agreed. The first floor came in at $12,000.
One of them told him nobody had ever worked with him that way.
The nine-day video
Steven Bartlett's reply is the part I keep coming back to. His company has a Slack channel called "paper walls". A paper wall is something that looks solid, a cost, a deadline, a "that's just how long it takes", and turns out to be paper the moment someone pushes on it.
He tells the story of an animator who had worked with him for four years. Each video took nine days. Bartlett wanted a weekly channel, so he needed it in three or four. He asked why it took nine days.
The answer: the laptop was old. A new laptop.
Four years of nine-day videos, and the fix was a purchase order. Nobody had asked, and the animator had never mentioned it. He did not know he was standing behind a paper wall himself.
Why this matters more than it sounds
Both stories are the same failure from opposite sides. In the first, the buyer accepts a number because every seller agrees on it. In the second, the manager accepts a number because the person doing the work has always delivered it that way. In both cases the number feels like a fact. It is a habit.
I see this every week, in software estimates more than anywhere else. "That feature is three sprints." "The integration takes a month." "The report has to be built by hand every Friday." Sometimes that is true. Often nobody has broken it down since the first time it was said out loud.
AI has made this sharper. A lot of work that "takes two days" took two days in 2022. Ask the person doing it what the two days are made of, and you often find a step a tool now does in minutes, or a bottleneck that is nothing to do with the work at all, like Bartlett's laptop.
How to push without being a jerk
Neumann's method is worth copying, and so is the tone of it.
- Ask for the parts, not a discount. "Walk me through what goes into this" is a question people can answer. "Can you do it cheaper?" is a negotiation, and everyone defends the number.
- Agree on what fair looks like first. He did not squeeze the contractors' margin. He agreed it with them, then made cost visible. That is why they said yes.
- Ask "why" about the obstacle, not the person. The animator was not slow. He was stuck. If the question sounds like blame, you get a defence instead of the laptop.
- Make it a shared habit. The point of a "paper walls" channel is that pushing on a number is normal, so nobody has to be brave to do it, and people start pushing on their own walls before you have to.
Try this week
Pick one number your business treats as fixed: a supplier price, a lead time, a recurring task that "always takes" a day. Ask the person closest to it to break it into its parts. Then ask why for each part.
Most walls are real. Some are paper. You only find out by pushing.