What it means
These estimates work by breaking one big unknown into a short chain of numbers you can either look up or sensibly guess. Market size becomes customers multiplied by price, hiring cost becomes salary multiplied by a loading factor, and a project payback becomes investment divided by monthly saving.
The point is to answer a decision question quickly rather than accurately. If a rough estimate says a project is worth $50,000 a year and building it costs $2,000,000, no amount of extra modelling will change the answer, and you have saved weeks of work.
Good practice is to keep the chain short and write the assumptions down where others can see them. Three or four steps is usually enough, and each assumption should be a number a reasonable colleague could dispute, because the dispute is where the value actually comes from.
Always test how sensitive the answer is before you act on it. Change the one or two assumptions you are least confident about, see whether the conclusion flips, and if it does, you have just identified exactly where the detailed analysis should go next.
The main risk is that a rough figure gets repeated until it hardens into an apparent fact. Label estimates clearly, keep the working attached, and replace them with proper analysis before they find their way into a board paper or a budget.
In practice
Real-world examples.
Example
A founder asked how much a new warehouse will cost estimates 40,000 square feet at $9 per square foot per year, which is $360,000 of rent, then adds roughly 30% for utilities, rates and fit-out amortisation to reach about $468,000 a year. It is enough to know the idea needs $500,000 of annual gross profit to justify.
Example
A marketing manager evaluating a trade show works out that 900 booth visitors at a 5% lead conversion gives 45 leads, and at a 20% close rate that is 9 customers. Against a $60,000 event cost, each customer must be worth more than about $6,700 for the show to break even.
Example
A finance director sizing an automation project estimates that three staff spend 40% of their time on manual invoice matching. At a fully loaded cost of $70,000 each, that is 3 x 0.4 x $70,000 = $84,000 a year, so a $150,000 system pays back in under two years if it removes most of the work.
Formula
Calculation
Estimate = Starting population x Filter rate x Conversion rate x Value per unit
A software team wants to know roughly how big a new product for mid-sized logistics firms could be before committing engineering time. They start with 12,000 companies in the target size band, then assume that 25% are realistically reachable and actually fit the specification: 12,000 x 0.25 = 3,000 qualified prospects.
Assuming a 4% conversion rate in the first year gives 3,000 x 0.04 = 120 customers. At an average contract value of $9,000 a year, the estimate is 120 x $9,000 = $1,080,000 of annual recurring revenue.
The sensitivity test matters as much as the answer. Halving the conversion assumption to 2% gives 60 customers and $540,000, while holding conversion at 4% and raising the contract value to $12,000 gives 120 x $12,000 = $1,440,000. A range of roughly $540,000 to $1,440,000 is enough to decide whether to fund a proper business case, and it tells the team that pricing and conversion, not market size, are the assumptions worth researching first.Case study
Seen in the real world.
Tallow Street Coffee is a fictional chain of eleven cafes invented for this illustrative example. Its founders were considering a home delivery subscription for roasted beans and had been quoted $180,000 for the platform, warehousing setup and first year of operations.
Before commissioning any research, the operations lead did the arithmetic on a single page. The cafes served about 14,000 distinct regular customers a year; assume 8% would try a subscription, which is 1,120 sign-ups; assume 60% survive the first three months, leaving 672 active subscribers; at $34 a month with a 45% contribution margin that is 672 x $34 x 12 x 0.45 = $123,379 of annual contribution. Against $180,000 of cost, the first year would lose money and the second would only just recover it.
The estimate did not kill the idea, it redirected it. Because the answer hinged almost entirely on the 8% trial rate and the 60% retention figure, the founders spent $4,000 running a small pilot in three cafes to measure both properly, rather than spending $180,000 to find out. The illustrative lesson is that a rough calculation is most useful when it tells you which assumption to go and test.
Watch out
Common mistakes.
- Adding false precision, such as quoting an estimate as $1,083,400 when every input was a rounded guess, which invites people to trust it far more than they should.
- Building a chain of eight or nine assumptions, where the compounding uncertainty makes the final number close to meaningless.
- Letting a rough figure travel into budgets and board papers without the assumptions attached, so nobody remembers it was only ever an estimate.
Questions
People also ask.
How accurate should a back-of-the-envelope estimate be?
Right to within a factor of about two is usually enough, because the purpose is to decide whether to investigate further, not to set a target.
When should I stop estimating and build a proper model?
As soon as the decision is genuinely close, the money at stake is large, or the sensitivity test shows the conclusion flips on a plausible change of assumption.
Is it acceptable to use these in a formal business case?
As supporting context yes, but the case itself should rest on researched inputs, with the rough estimate shown as the screening step that justified the work.
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%Related