What it means
A binomial distribution describes a run of identical, independent trials where each trial has only two possible outcomes, usually labelled success and failure. To use it you need three things: the number of trials, the probability of success on any single trial, and the number of successes you want to ask about.
Most business questions are really counting questions: how many customers will renew, how many invoices will be paid late, how many machines will fail this quarter. The distribution turns a single-event probability into a full picture of the likely range of totals, which is what budgets and risk limits actually need.
The calculation combines two ideas: the chance of one specific sequence of wins and losses, and the number of different sequences that produce the same total. Multiplying those two together gives the probability of exactly a given number of successes.
Two shortcuts do most of the practical work. The average number of successes is the number of trials multiplied by the probability of success, and the spread around that average is the square root of the number of trials multiplied by p and by one minus p.
A team with a 40% close rate making 50 calls should expect 20 wins, with a standard deviation of about 3.5 calls. The model assumes every trial is independent and that the probability never changes, which is only roughly true in real markets.
Customers talk to each other, and a downturn moves every default probability at once. Treat the answer as a sensible base case rather than a precise forecast.
When the number of trials is large the shape starts to look like the familiar bell curve, so analysts often switch to the normal distribution for speed. When the probability of success is very small and the number of trials is large, the Poisson distribution is the usual approximation instead.
In practice
Real-world examples.
Example
A subscription software company has 12 enterprise contracts coming up for renewal, each with an independent 85% chance of renewing. The finance team calculates that the probability of every single one renewing is 0.85 raised to the power of 12, which is about 14%. That single number stops the board from budgeting as though a clean sweep were the normal outcome.
Example
A components manufacturer runs a line with a 3% defect rate and samples 20 units from each batch. The expected number of defects in a sample is 20 x 0.03 = 0.6, and the chance of finding none at all is 0.97 to the power of 20, or about 54%. Quality managers use this to see that a clean sample is not proof that the line is fault free.
Example
A small insurer writes 200 identical household policies, each with a 4% chance of a claim in the year. The expected number of claims is 8, with a standard deviation of about 2.8 claims. Reserves are set to cover a bad year of roughly 14 claims rather than the average of 8.
Formula
Calculation
P(exactly k successes) = C(n, k) x p^k x (1 - p)^(n - k), where n is the number of trials, p is the probability of success on one trial, and C(n, k) is the number of ways of choosing k items from n, worked out as n! / (k! x (n - k)!).
Worked example: a sales team makes 5 pitches this month, and each pitch has an independent 40% chance of winning a $20,000 contract. What is the chance of winning exactly 2 of them?
C(5, 2) = 10
p^2 = 0.4 x 0.4 = 0.16
(1 - p)^3 = 0.6 x 0.6 x 0.6 = 0.216
P = 10 x 0.16 x 0.216 = 0.3456, which is 34.56%
The expected number of wins is 5 x 0.4 = 2 contracts, which at $20,000 each is $40,000 of expected new revenue for the month.Case study
Seen in the real world.
Northvale Components is an illustrative, entirely fictional engineering supplier that bids for large maintenance contracts. It submits about 40 tenders a year and wins roughly 25% of them, with each win worth around $150,000 of revenue. The sales director had always budgeted for 10 wins and $1.5 million, and treated any shortfall as a performance failure.
A new finance manager modelled the year as a binomial distribution with 40 trials and a 25% success rate. The average was indeed 10 wins, but the standard deviation came out at about 2.7, meaning that anything from roughly 5 to 15 wins was a perfectly ordinary result. In revenue terms that is a spread from $750,000 to $2.25 million.
The illustrative lesson was not that the sales team was unpredictable, but that the budget had been built on a single point estimate. Northvale switched to funding fixed costs from the pessimistic end of the range and treating anything above 10 wins as upside, which removed a great deal of mid-year panic.
Watch out
Common mistakes.
- Assuming the expected value is the most likely single outcome and then planning as if it will happen, when a wide range of results around it is entirely normal.
- Applying the distribution to trials that are not independent, such as customer churn during a service outage, where one failure makes the next far more likely.
- Confusing the probability of exactly k successes with the probability of at least k, which requires adding up several separate probabilities.
Questions
People also ask.
What does the "binomial" part actually mean?
It simply refers to two outcomes, so every trial must be reducible to a clean success or failure with nothing in between.
Can I use it when the success probability changes over time?
Not directly, because the model assumes a fixed probability, so either split the period into segments with stable rates or move to a simulation.
Do I need to calculate this by hand?
No, spreadsheets have a built-in binomial function, and the value of the concept is in framing the question rather than in the arithmetic.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%