How this instrument works
A compound growth rate takes two readings of the same metric — a starting figure and an ending figure — separated by some number of years, and returns the single steady annual pace that would carry the first to the second. It works on revenue, subscriber counts, headcount, unit sales, or any other quantity a business tracks across a multi-year stretch, not only money sitting in an account. The output is not a record of what any one year actually did; it is a hypothetical constant rate that reproduces the same overall multiple if it applied evenly every year.
The formula takes the ratio of ending to starting value, extracts its years-th root, and converts that root into a percentage. Extracting a root rather than dividing the total percentage gain by the number of years is the entire point: growth compounds, so a business that manages to double over five years needs a rate well under the 20 percent a flat average would suggest — nearer 14.87 percent, because each year's gain builds on a base that already carries every earlier year's gain forward. Skipping the root and dividing gain by years is the single most common misreading of a multi-year growth story, and it always overstates the true steady pace whenever growth is positive.
Two readings and a year count are all the arithmetic sees, so it says nothing about the path between them — a metric that surged early and stalled later reports the same rate as one that grew in an unbroken straight line, and comparing two growth stories still means asking how each one actually arrived at its number. It also breaks down once the starting figure sits at zero or below, since no constant compounding rate connects a zero base to any finite result, and it treats every year identically even though a real business's growth is rarely that tidy.
- Enter Starting value, $ — the reading at the beginning of the stretch you're measuring, whether that's revenue, users, or any other tracked figure.
- Enter Ending value, $ — the same metric's reading at the end of that stretch.
- Set Years elapsed to the span between the two readings; decimals are fine, so eighteen months is 1.5.
- Read Compound annual growth rate (CAGR), % — the one steady annual pace connecting the two figures.
- Run it again with a different Years elapsed to line up two growth stories of unequal length on the same annual footing.
Worked example — revenue that doubled in five years
Set Starting value, $ to 10,000 and Ending value, $ to 20,000, with Years elapsed at 5 — a small business whose annual revenue went from $10,000 to $20,000 over five years. Dividing 20,000 by 10,000 gives a growth factor of 2; raising that to the power 1⁄5 gives roughly 1.148698; subtracting 1 and multiplying by 100 puts Compound annual growth rate (CAGR), % at 14.8698.
That figure sits well below the instinctive shortcut of splitting a 100 percent total gain across five years, which lands on a flat 20 percent a year. Applied honestly — compounding 20 percent for five years running — that guess would carry $10,000 to about $24,883, well past the $20,000 actually reached; the true steady rate has to be smaller than the naive average precisely because compounding lets a lower rate go further than flat multiplication over several years.
The same arithmetic works on any steadily tracked figure, not only revenue — headcount, active users, unit sales, or a chain's store count all convert into a comparable annual pace once two readings and a year span go in. A three-year-old company posting a 40 percent rate here and a ten-year-old company posting 18 percent are, on this one measure, growing at genuinely different speeds regardless of how much history sits behind each of them, which is why the rate belongs on a board slide more than the raw multiple does.
Questions
Why isn't the growth rate here just the total percentage change divided by the years?
Because growth compounds instead of adding up. Splitting a 100 percent total gain over five years into a flat 20 percent a year overstates the true pace — compounding 20 percent for five years would actually carry $10,000 to about $24,883, well past the $20,000 this example reached. The real steady rate has to be lower, which is what the root in this formula solves for: 14.87 percent, not 20.
Does this work for metrics that aren't measured in dollars?
Yes — the arithmetic only needs two readings of the same unit and a year count between them, so subscriber counts, headcount, unit sales, or store locations all convert into a steady annual pace the same way revenue does. Enter the earlier count as Starting value and the later one as Ending value; the percentage on the output describes a rate of change, not currency.
Can I compare two businesses with different amounts of history on this number?
Yes, and that is one of the more useful things a compound growth rate does. A three-year-old company and a ten-year-old company cannot be compared on their raw multiples — tripling in three years and tripling in ten years describe very different speeds — but both convert into one steady annual percentage that sits on the same scale regardless of how long each history actually ran.
Does a high reading here mean growth happened at that pace every single year?
No. Only the starting and ending readings are used, so a metric that jumped early and flattened out afterward, or one that dipped in the middle and recovered, reports the identical rate to one that grew in an unbroken straight line. Treat the output as a summary of the whole stretch, not as a claim about any individual year inside it.
What happens if Starting value is zero or negative?
The instrument blocks it. Dividing by a zero starting value has no answer, and a negative starting figure sends the fractional exponent outside real numbers entirely, so no constant annual rate can connect it to any ending figure. A metric that starts at zero — a brand-new product line, say — needs an absolute or per-unit measure instead of a growth rate for its first stretch.
How is this different from tracking year-over-year percentage change?
Year-over-year change reports what happened between two adjacent periods, one pair at a time, and can swing from a strong quarter to a weak one. This instrument instead smooths an entire multi-year stretch into a single number, which is more useful for setting a multi-year target or comparing trajectories than for explaining why last quarter looked the way it did.
References
Read this first: This instrument shows arithmetic, not advice. Real offers add fees, taxes and terms that vary by lender and place — verify the figures against your actual paperwork before deciding anything.