P50 or P80: the quoting decision that decides whether you eat
Quote the median and half your fixed-price jobs go underwater. Quote P80 and you stay safe but lose bids. The portfolio math decides.
The median is not a safe number to quote
Your Monte Carlo run returns a distribution of possible costs. P50 is the cost the project stays under half the time. If you quote at P50 plus your margin, you are pricing a coin flip: on the runs where cost lands above the middle, the job loses money or eats the margin.
It feels safe because the median looks like the average everyone uses. It is not. On a project whose cost distribution skews long — which almost every real project does — the median sits noticeably below the expected value, and the expected value itself sits below the level you would want to survive on.
P50 maximises how often you get the job. P80 maximises how often the job pays for itself. Those are different businesses.
What the percentiles actually promise
A P80 quote is built so that the project lands inside budget in roughly eight out of ten simulations. That is not padding — it is buying back the tail of your own uncertainty. The premium you pay for it is the gap between the P50 cost and the P80 cost, which is the price of the risk you are no longer eating yourself.
On a sample project with a P50 cost of $36,000 and a P80 cost of $44,000:
| Quote basis | Your quote | Expected margin | Probability of loss |
|---|---|---|---|
| P50 + 15% | $41,400 | Thin | Roughly 1 in 2 |
| P80 + 15% | $50,600 | Healthy | Roughly 1 in 5 |
| P90 + 15% | $55,200 | Healthy | Roughly 1 in 10 |
Same project, same margin rule, three very different businesses. The P50 quote wins more often and loses money on a huge share of the jobs it wins. The P80 quote loses some bids outright — but the bids it wins mostly make money.
Treat it as a portfolio game
If you bid a handful of jobs a year, one underwater project can erase the profit from three good ones (see the fixed-price death spiral). If you bid dozens, a deliberate P50 strategy with strict loss limits can be rational — but only if you actually have the margin pool and the appetite to survive the losers.
The decision is not which number is correct. It is which loss rate your firm can survive, and that is a portfolio question, not a single-bid question. Most firms under price risk precisely because they have never been shown the loss probability attached to the number they quote.
The free P50 vs P80 explainer walks a full project through both quoting styles. If you want to see your own minimum quote at a chosen confidence, the minimum bid price calculator reverses the math: give it your cost distribution and target probability, and it returns the price.
Choosing your percentile on purpose
Pick your percentile the way you pick a portfolio: consciously, with the downside named. Decide first how many losing bids you can absorb in a year, then let that decide the percentile — not the other way around. If a competitor forces you under your survivable percentile, the correct move is often to walk away, because a bid you cannot survive winning is not a bid, it is a liability.
When you are ready to see the full picture on your own project — cost distribution, loss probability at your chosen quote, and the reverse-priced minimum — the app runs the whole model in your browser.
The takeaway
P50 and P80 are not two flavours of the same quote; they are two different risk positions in the same project. The median quotes you into trouble more often than you think, because half the distribution is on the wrong side of it. Name your loss rate first, then let the percentile follow — and treat any bid you could not survive winning as a bid you should not make.