How Estimating Works — A Complete Guide

A shopping basket holds three items priced at £4.95, £2.49 and £7.99. Before reaching the till, a shopper wants to know roughly how much the basket will cost — not to the penny, but near enough to check that a £20 note will cover it. Nobody works out £4.95 plus £2.49 plus £7.99 to the exact penny in their head for a question like that. They round the prices to £5, £2.50 and £8, add them to get £15.50, and know at once that the answer is about £15. Estimating is that skill: replacing awkward numbers with friendlier ones, doing the easy arithmetic, and getting an answer that is close enough for the decision at hand.

This guide explains what estimating means, how front-end estimation and rounding to estimate work, what compatible numbers are, and how to estimate percentages and fractions. It works through a shopping total, a multiplication and a division, and it shows where estimates go wrong, so you can judge when an estimate is good enough and when an exact answer is needed instead.

What estimating means

To estimate is to find an answer that is close to the exact answer without doing the full calculation. The estimate is deliberately approximate: it trades a small amount of accuracy for a large amount of speed and simplicity. An estimate of £15.50 for a basket that actually costs £15.43 is a good estimate, because it is only 7 pence out and it answered the real question — will £20 be enough? — in seconds. An estimate is not a guess. A guess is a number chosen without working; an estimate is a number reached by a shortened calculation, and its closeness can be judged.

Estimating is useful whenever the exact figure does not matter, or is not yet known. Checking a bill, judging whether a journey is affordable, or deciding whether a calculation on paper looks sensible are all estimating jobs. It is not useful when the exact figure is the whole point: a bank balance, a medicine dose, or the amount of concrete for a foundation must be exact, and an estimate there would change the meaning. The first question in any estimating problem is therefore the same as in rounding: is an approximate answer good enough here, and if so, how close does it need to be?

Front-end estimation

Front-end estimation is the simplest estimating method, and it works by keeping only the leading digits of each number and ignoring the rest. To estimate 47 plus 38, keep the front ends — 40 and 30 — and add them to get 70. The exact answer is 85, so this estimate is 15 too low, which is a large error for a small sum. Front-end estimation always underestimates when numbers are added, because the ignored digits are all positive, and the error grows as more numbers are added. It is quick, but it is the crudest method in this guide, and it should only be used when a very rough figure will do.

Front-end estimation is more useful for a first, rough check than for a final answer. To estimate 873 plus 486, the front ends are 800 and 400, giving 1,200, while the exact answer is 1,359. The estimate is 159 too low, but it does tell the reader at once that an answer of 259 or of 13,590 would be wrong. A better estimate for the same sum rounds each number first: 900 plus 500 gives 1,400, which is only 41 above the exact answer. Whenever the numbers have more than one digit, rounding before calculating almost always beats front-end estimation, because the rounding errors partly cancel instead of all pointing the same way.

Rounding to estimate

Rounding to estimate means rounding each number in the calculation to a friendly value, then doing the arithmetic with the rounded numbers. The level of rounding is chosen to suit the question: round to the nearest 10 for a quick check, or to the nearest whole number when the numbers are small. To estimate 19 times 6, round 19 to 20 and multiply: 20 times 6 is 120, while the exact answer is 114. The estimate is 6 too high, which is close enough to confirm that the exact answer is in the right neighbourhood and that an answer of 1,140 would be a slip of a decimal place.

The same method works for division. To estimate 312 divided by 8, round 312 to 320, because 320 divides neatly by 8: 320 divided by 8 is 40, while the exact answer is 39. The estimate is one too high, which is excellent for a mental check. To estimate 52 times 7, round 52 to 50: 50 times 7 is 350, while the exact answer is 364. In each case the rounded numbers were chosen because they make the arithmetic easy — multiples of 10, numbers that divide neatly — and the estimate is then only a short step from the exact answer.

Estimating with multiplication needs one extra care: rounding both numbers in the same direction compounds the error. To estimate 96 times 11, rounding both numbers up to 100 times 11 gives 1,100, while rounding 11 down to 10 gives 100 times 10, which is 1,000. The exact answer is 1,056, which sits between the two estimates. Rounding one number up and the other down, where that is possible, keeps the errors pulling in opposite directions and gives a closer estimate. When both numbers must be rounded the same way, expect a larger error and say so.

Compatible numbers

Compatible numbers are numbers that work together easily in the calculation being done. For division, they are numbers that divide without a remainder. To estimate 6,480 divided by 9, the nearest number that divides neatly by 9 is 6,300, because 6,300 divided by 9 is 700, while the exact answer is 720. The estimate is 20 too low, but it was reached in one mental step. For multiplication, compatible numbers are round numbers: to estimate 287 times 4, use 300 times 4, which is 1,200, against the exact answer of 1,148.

Compatible numbers are chosen, not found by a rule, and the choice is part of the skill. For 6,480 divided by 9, another reader might round to 6,480 itself and notice that 6,480 is 9 times 720 — but that is the exact calculation, not an estimate. The point of a compatible number is that it is easier than the original: 6,300 is easier because it is a multiple of 100 that happens to divide by 9. With practice, spotting the nearest friendly number becomes quick, and it is the method that experienced mental calculators use most often for division.

Estimating percentages and fractions

Percentages can be estimated by breaking them into easy parts: 10 percent, 5 percent and 1 percent. To estimate a 15 percent tip on a £68 meal, work out 10 percent, which is £6.80, and 5 percent, which is half of that, £3.40. Adding them gives £10.20, which is the exact tip. A quicker estimate rounds £68 to £70 first: 15 percent of £70 is £10.50, which is 30 pence too high but near enough to decide on the tip. The rounding-first route is the estimating route; the breaking-down route is the exact route, and the two can be compared to check each other.

Fractions are estimated the same way, by replacing them with friendlier fractions. A quarter of 240 is 60, because 240 divided by 4 is 60, so 24 percent of 240 — close to a quarter — is a little under 60. In fact 24 percent of 240 is 57.60, so the estimate is close. A third of 90 is 30, so 33 percent of 90 is a little under 30. The pattern is the same in every case: find the nearby easy fraction, work it out, and then adjust the answer slightly in the direction the rounding pushed it.

Worked example 1: the shopping basket

Three items cost £4.95, £2.49 and £7.99. Estimate the total, and decide whether £20 is enough. Rounding each price to a friendly value gives £5, £2.50 and £8, which add to £15.50. The exact total is £15.43, so the estimate is 7 pence too high. The estimate answers the real question immediately: £15.50 is well under £20, so £20 is enough, with about £4.50 in change. Front-end estimation would have given £4 plus £2 plus £7, which is £13 — a worse estimate, because it throws away the 95 pence, 49 pence and 99 pence entirely.

This example also shows why the level of rounding matters. Rounding to the nearest pound instead — £5, £2 and £8 — gives £15, which is also close, but it hides the 49 pence in the second item. Rounding to the nearest 50 pence kept that information and gave the closer answer. In a shop, the nearest 50 pence is usually the right level: prices are often set just under a round number, and rounding to the nearest pound systematically underestimates the total.

Worked example 2: estimating a product and a quotient

A builder needs 19 packs of tiles at £6 each, and 312 tiles must be shared equally between 8 rooms. Estimate both answers. For the packs, round 19 to 20: 20 times 6 is £120, while the exact cost is £114. The estimate is £6 too high, which is the cost of one pack, and it is a safe overestimate for budgeting — the builder will not run short. For the rooms, round 312 to 320: 320 divided by 8 is 40 tiles per room, while the exact answer is 39. The estimate is one tile too high per room, which is fine for planning, but the exact count of 39 would be needed before cutting tiles.

Notice that the two estimates err in different directions for different reasons. The multiplication estimate rounded 19 up, so it overestimated. The division estimate rounded 312 up, so it also overestimated the quotient. Had 312 been rounded down to 304, the estimate would have been 38, underestimating by one. Either direction is acceptable for an estimate, provided the reader knows which way the rounding pushed the answer and by roughly how much.

Worked example 3: the estimate that exposes a slip

A student multiplies 52 by 7 and writes 2,644. An estimate exposes the slip at once: 50 times 7 is 350, so the exact answer must be near 350, and the exact answer is 364. The written answer of 2,644 is nearly eight times too large, which is the signature of a misplaced digit — the student has multiplied by something near 50 instead of 7, or has added an extra digit. The estimate did not give the exact answer, but it gave something more valuable in the moment: the certainty that the written answer was wrong.

This is the most important use of estimating, and the reason it is taught before calculator work is trusted. A calculator will faithfully repeat a wrong key press, and a spreadsheet will faithfully repeat a wrong formula, but an estimate made before the calculation gives an independent check. The habit to build is simple: estimate first, calculate second, and compare the two before trusting either.

Common mistakes to avoid

Treating an estimate as exact is the most common error: an estimate of £15.50 is not the price, and it should be reported with the word “about”. The second is estimating with numbers that are too rough for the decision, such as front-end estimation for a budget where £15 of error matters. The third is rounding too early and then rounding again: each rounding step adds its own error, so round once, from the original numbers, and calculate with the rounded figures only once.

Two more traps are worth watching for. One is rounding both numbers in a multiplication in the same direction without noticing, which can double the error, as the 96 times 11 example showed. The other is forgetting the units: an estimate of 40 tiles per room is not 40 tiles in total, and an estimate of £120 for 19 packs is not £120 per pack. State the units in the estimate, just as in the exact answer, or the check the estimate was meant to provide will mislead instead.

Frequently asked questions

How close does an estimate need to be?

Close enough for the decision it serves. For checking that £20 covers a £15.43 basket, being within a pound is plenty. For ordering tiles, being within a few tiles per room is enough to plan, but the exact count is needed before cutting. A good habit is to state the estimate with its likely error: “about £15.50, within 50 pence” tells the reader far more than “about £15” with no sense of how far out it could be.

Is front-end estimation ever the best method?

Rarely, and only when a very rough figure is wanted at once. It is the fastest method, because it uses only the leading digits, but it always underestimates sums and its error grows with the number of terms. Rounding to estimate is better in almost every practical case, because the rounding errors partly cancel. Front-end estimation is best kept for a first, rough check, followed by a rounding estimate when the figure matters.

Should I estimate before using a calculator?

Yes, always. The estimate takes seconds and it is the only check that is independent of the calculator: if the calculator answer and the estimate disagree wildly, one of them is wrong, and the estimate is usually the clue that a key was pressed twice or a decimal point slipped. The Scientific Calculator on this site is useful for the exact calculation that follows the estimate, and the Percentage Calculator is worth checking when a percentage estimate needs confirming, since percentages are where mental slips are commonest.

Can estimating replace exact calculation?

No. Estimating answers the question “roughly how much”, and exact calculation answers “exactly how much”, and the two questions are different. Money owed, quantities to order, and answers in an examination all need exact figures, and an estimate there would be marked wrong or would cost money. Estimating is a companion to exact calculation — a check before it and a sense-check after it — not a substitute for it. When the exact figure is needed, calculate it exactly, and use the estimate only to confirm that the exact answer is sensible.

If you want to check a result, try the matching tools on this site: the Scientific Calculator handles the exact arithmetic after an estimate, and the Percentage Calculator is the usual place where a percentage estimate is confirmed. The Average Calculator is worth checking when an average has been estimated, since averages often come out with far more digits than the data justifies.