You Can’t Dilute Out Poor Quality

I did a small experiment the other day. I put 100 mL of hot water into a 1-litre bottle, then filled the rest with cold. I wanted it back at cold-tap temperature.

But once the hot water was in, I couldn’t get there. Even after filling to the brim with cold, the bottle was still a few degrees too warm. And to get it anywhere near cold again, I had to keep running cold water through until it overflowed. By the time it was close, I’d flushed nearly four litres of extra water.

A small mistake at the start had created a huge amount of rework.

The maths

• Cold tap: 15 °C

• Hot water: 60 °C

• Start: 100 mL hot + 900 mL cold

The mix ends up at:

T = (0.1×60)+(0.9×15) = 19.5∘

That’s already 4.5 °C too warm.

Now flush the bottle with cold water while it overflows. The excess temperature decays exponentially:

ΔT(V)=ΔT0 e(−V/Vbottle)

To get within 0.1 °C of cold water (15 °C):

V=1×ln⁡ ⁣(4.5/0.1) ≈ 3.8 litres

So one small hot splash means almost four times the bottle volume of extra work to put it right.

The consulting lesson

Consulting quality works the same way.

You can have 95% of a report or a solution solid, but if one area is weak — an area of poor logic, a half-tested model, a poorly written section — it pulls everything down. Clients don’t judge the strong sections in isolation. They notice the part that’s off, and that drags the whole product to average.

And here’s the kicker: you can’t just “pour in” more high quality elsewhere to bring the standard back.

Over-compensating in other areas doesn’t restore confidence.

To get back to the target quality, you end up doing multiple rounds of rework — the consulting equivalent of flushing litres of water through a 1-litre bottle.

My rule

Do everything that can be done to avoid low quality in the first place - no warm water.

“Because once the hot water is in, you can never simply top it up with cold and expect it to go away.”