How to Calculate Success Rate

Updated September 2026 · Written and maintained by the Progression Agency strategy team

Success rate is the number of successes divided by the number of attempts, times one hundred. The arithmetic is trivial and almost nobody gets it wrong. What goes wrong is everything that happens before the arithmetic: what counts as a success, what counts as an attempt, and over what period. Get those three wrong and two people can compute the same figure from the same data and both be right, which is how a metric stops being useful. This page covers the formula, the denominator problem, worked examples across sales, support, email and payments, and how to tell an honest success rate from a flattering one.

What is success rate?

Answer first: success rate is the proportion of attempts that achieved a defined outcome, expressed as a percentage. It measures efficiency — how well attempts convert — as distinct from volume, which measures how many succeeded in total.

Success rate versus success rating

The two phrases are used interchangeably in everyday speech, and where they differ it is usually this: a success rate is a calculated percentage, while a success rating is frequently a subjective score or a band derived from one. If somebody quotes you a success rating, the useful question is whether there is a rate underneath it and what its denominator was.

Why efficiency and volume are different questions

A team that makes ten attempts and succeeds nine times has a ninety per cent success rate and nine successes. A team that makes a thousand attempts and succeeds a hundred times has a ten per cent rate and a hundred successes. Which is performing better depends entirely on what an attempt costs and what a success is worth, which is why both numbers must be reported together.

The formula

Answer first: divide successes by attempts and multiply by one hundred. That is the whole calculation, and it is not where the difficulty is.

Success rate = (S ÷ A) × 100

Where S is the number of attempts that met the defined success criterion, and A is the total number of attempts made in the same period.

Failure rate = 100 − success rate.

A worked example

A sales team made 240 qualified approaches last quarter and closed 54 of them. The success rate is (54 ÷ 240) × 100 = 22.5 per cent. Reported properly, that reads: 22.5 per cent close rate on 240 qualified approaches, Q1 — because the percentage, the sample size and the period together are the claim, and any one of the three removed makes it unverifiable.

The denominator problem

Answer first: almost every misleading success rate comes from a denominator that quietly excludes inconvenient attempts. Deciding what counts as an attempt is a more consequential decision than deciding what counts as a success.

The conclusion: the only defensible position is a written definition of an attempt, applied consistently, with any exclusions stated explicitly and the excluded count reported alongside. If an exclusion cannot survive being written down, it should not be made.

Defining success before you count it

Answer first: write the success criterion down in one sentence, and make it binary. A criterion that requires judgment produces a rate that changes depending on who counts.

1. Make it binary. Did it happen or not. ‘Partially successful’ outcomes need their own category, not a share of the numerator.
2. Make it observable. Somebody other than the person who made the attempt should be able to verify it from a record.
3. Make it time-bound. A sale that closes eleven months later — does it count in the quarter of the attempt or the quarter of the close? Decide once.
4. Write it where everyone can see it, not in one team’s spreadsheet.
5. Record the date you defined it. If the definition changes later, the trend before and after that date is not comparable and somebody needs to know.

Where success rate is used

Answer first: the same formula appears across sales, marketing, support, email, payments and hiring, with entirely different normal ranges. A three per cent conversion rate and a ninety-eight per cent deliverability rate can both be excellent.

The conclusion: when comparing a success rate against any external figure, check whether both used the same definition of an attempt. In most cases they did not, and the comparison is meaningless rather than merely imprecise.

Success rate against the metrics it gets confused with

The most important row is the second. Volume of successes is what people report when the denominator is inconvenient, because it always rises with effort. A team can double its successes while halving its success rate, and only reporting both reveals it.

How to tell an honest success rate from a flattering one

1. Ask for the denominator. A percentage without one is a claim rather than data.
2. Ask for the sample size. Ninety per cent of ten is not the same statement as ninety per cent of ten thousand.
3. Ask what was excluded. If anything was, ask how many and on what rule.
4. Ask when the definition last changed. A rate that improved when the definition changed did not improve.
5. Ask for the trend, not the snapshot. One period is an anecdote.
6. Ask to see it by segment. An average across segments frequently hides the finding.

Small samples

Below roughly thirty attempts, a success rate is unstable enough that a single outcome moves it materially. That does not make it useless, but it means it should be quoted with the raw numbers attached — ‘four of six’ rather than ‘sixty-seven per cent’ — because the second sounds far more authoritative than the evidence supports.

Averages across segments

A single success rate across several segments can conceal the entire finding. If one channel converts at fifteen per cent and another at two, the blended figure of eight tells you nothing actionable and quietly protects the failing channel from scrutiny.

Improving a success rate, honestly

Answer first: there are only three levers — improve the attempts, improve the execution, or narrow what you attempt. The third improves the rate without improving the business, which is why it needs watching.

The conclusion: rows four and five raise the number without changing anything real, and they are common precisely because they are cheap. Reporting rate alongside volume, with the definition and its change history, makes both visible.

Success rate over time: trends, not snapshots

Answer first: a single period’s success rate is an anecdote. What tells you anything is the trend across several periods against a definition that did not change, with the volume shown alongside.

How many periods before a trend is real

Answer first: three at minimum, and more if the volumes are small. Two points make a line and a line is not a trend — one unusual month between two ordinary ones produces two apparent trends pointing in opposite directions, both of which will be explained confidently at the next meeting.

What to do when the definition changed mid-trend

Answer first: break the series. Show the periods before the change and the periods after as two separate lines with the change date marked, rather than a single line that appears to move. A continuous line across a definition change is the most common way a chart tells a lie without anybody intending one.

Seasonality in success rates

Answer first: compare like periods. Many success rates are seasonal — close rates fall in holiday periods, support resolution rates fall when volumes spike — and comparing December to November produces a finding about the calendar rather than about performance. Year-on-year comparison of the same month is usually the honest view.

Building a success rate into a report nobody argues with

Answer first: publish the method alongside the number. If anybody can reproduce the figure from the raw data using your written definition, the argument moves from the number to the performance — which is where it should be.

1. Publish the definition of success and of an attempt, dated.
2. Publish the exclusion rules, if any, with the excluded counts.
3. Publish the raw counts next to every percentage.
4. Publish the period and use consistent period boundaries.
5. Publish by segment as well as in aggregate.
6. Publish the change log so a shift in the line can be attributed.

Reporting a success rate properly

Answer first: the percentage, the numerator, the denominator, the period and the definition. Five elements, one line. Anything less and the reader cannot check it.

Good: “Close rate 22.5% (54 of 240 qualified opportunities), Q1 2026. Qualified opportunity defined as a recorded discovery call with a named budget holder.”

Bad: “Close rate 22.5%.”

• The percentage is the headline, and it is the least informative part.
• The raw counts let the reader assess reliability instantly.
• The period prevents comparison across incomparable windows.
• The definition prevents two teams reporting different numbers for identical performance.
• Any exclusions, with counts, stated rather than assumed.

Related reading on the surrounding measurement discipline: what retention means and how to calculate it and how performance is measured in marketing.

Measurement and analytics, from the people who publish the platforms

Publicly available talks from Google Ads, Think with Google, Ad Age, HubSpot, Ahrefs and Neil Patel on measurement, attribution and analytics — the discipline that makes any rate trustworthy. None of these are ours; each is credited to its channel by name and upload date, every identifier was verified live before publication, and each tile loads its player only when you click it.

Calculating TAM

The formula, and the honest way to use it.

The TAM formula in its bottom-up form is the one worth using: number of potential customers multiplied by average annual revenue per customer. The top-down form — taking an industry report’s market size and applying a percentage — produces a larger number and almost no information, because the assumptions are someone else’s.

The value of calculating it bottom-up is that each input is checkable and arguable. How many businesses of this type exist in the geography you serve, how many have the problem, what would they pay. Those three numbers can be researched and defended; a percentage of a global figure cannot.

Rate versus ratio versus percentage

A rate is events over a period, a ratio compares two quantities, a percentage is a ratio expressed per hundred. Reports mixing the three produce numbers that cannot be compared across months.

Choose the denominator before the campaign

Success rate calculated on impressions, sessions, qualified leads or opportunities gives four different numbers from the same data. Deciding afterward is how a figure becomes unfalsifiable.

Sample size decides whether the number means anything

A success rate from twelve attempts moves wildly with one more outcome. Below roughly a hundred events, report the raw counts alongside the rate or the rate misleads.

Segment before averaging

A blended rate across channels hides that one is excellent and another is failing. The average is the least useful view of a mixed portfolio.

Watch for survivorship in the denominator

Excluding the attempts that never completed inflates every rate. If a step drops people silently, they belong in the denominator.

Report the trend, not the snapshot

A single period’s rate is noise; three periods is a signal. Any dashboard showing only the current figure invites the wrong decision.

Define failure as carefully as success

Most disputes about a success rate are actually disputes about what counted as an attempt.

State the period alongside the rate

A success rate without a window is unfalsifiable. Quarterly and monthly figures from the same data routinely disagree, and the one quoted is usually the flattering one.

Frequently asked questions