Most errors in business reporting are not exotic statistics mistakes. They are percentages computed against the wrong base, averaged when they should not be, or described with the wrong word. This piece walks through the handful of calculations that cover most day-to-day reporting, with the trap attached to each.

Percentage change: always divide by the old value

Percentage change is (new − old) ÷ old. If quarterly revenue rises from ₹120 crore to ₹138 crore, the change is ₹18 crore ÷ ₹120 crore, which is 15%.

The base is what makes increases and decreases asymmetric. A fall of 20% followed by a rise of 20% does not return you to where you started: ₹100 becomes ₹80, and 20% of ₹80 is only ₹16, so you end at ₹96. To recover a 20% fall you need a 25% rise. This matters when someone says a metric has “recovered” after a bad month.

Percentage change also breaks when the base is zero, very small or negative. Going from 2 enterprise customers to 10 is a 400% increase, which sounds dramatic and tells you little. Going from a loss of ₹5 crore to a profit of ₹3 crore has no meaningful percentage change at all; dividing by a negative number produces a sign that misleads. In these cases, report the absolute change and say why the percentage is omitted.

Percent versus percentage points

When the thing being compared is itself a percentage, there are two different “changes”. Suppose an online store’s conversion rate falls from 3.0% to 2.4%.

  • The absolute change is 0.6 percentage points (3.0 − 2.4).
  • The relative change is a 20% decline (0.6 ÷ 3.0).

“Conversion fell 0.6%” is ambiguous at best and wrong at worst. Interest rates, margins, market shares, attrition rates and conversion rates should be compared in points when you mean the arithmetic difference, and in percent when you mean the proportional one. Saying both is often clearest: “down 0.6 points, a 20% relative drop”. The same applies to basis points in finance, where one basis point is one hundredth of a percentage point.

Growth over several years: CAGR

Suppose a company’s annual revenue went ₹50 crore, ₹80 crore, ₹60 crore, ₹90 crore over four consecutive years. The year-on-year growth rates are +60%, −25% and +50%. Their simple average is about 28.3%. But if you grow ₹50 crore at 28.3% for three years, you land well above ₹90 crore. The simple average overstates growth because percentage changes compound.

Compound annual growth rate answers: what constant yearly rate would take the starting value to the ending value over the same number of years? The formula is (end ÷ start) raised to the power 1/n, minus 1, where n is the number of years between them (here 3, not 4). For this company: (90 ÷ 50) to the power 1/3, minus 1, is about 21.6%. Check: ₹50 crore grown at 21.6% for three years gives ₹90 crore.

Two cautions. CAGR ignores the path entirely; it would be the same if the middle years had been a smooth climb or a near-collapse, so show the yearly values alongside it. And CAGR depends heavily on the start and end points chosen. Starting from a pandemic-hit year or ending at an unusually strong one can make the same business look very different. Ask why a particular window was picked.

Share of total, and why you should not average percentages

Share of total is each part divided by the sum of all parts. If South contributes ₹42 crore, West ₹35 crore and North ₹23 crore of a ₹100 crore total, the shares are 42%, 35% and 23%. Shares must add to 100%; if your table shows 101% because of rounding, say so in a footnote rather than quietly fudging one number.

A share can fall even when the part grows, if the rest grows faster. “South’s share dropped from 45% to 42%” may hide strong absolute growth in South. Report both absolute values and shares when the story is about mix.

The other classic mistake is averaging rates that have different bases. Two stores: Store A converts 10 of 100 visitors (10%), Store B converts 300 of 10,000 (3%). The simple average of the two rates is 6.5%. The actual combined conversion is 310 out of 10,100, about 3.1%. The simple average treats a tiny store as equal to a large one. To combine rates, add the numerators and add the denominators, then divide. This is also why “average of daily conversion rates” and “monthly conversion rate” in two dashboards rarely agree.

A short checklist before publishing a percentage

  • Is the base the old value, and is it large and positive enough for a percentage to mean anything?
  • If comparing two rates, have you said points or percent, and meant it?
  • For multi-year growth, is it CAGR with the correct n, and are the endpoints reasonable?
  • Are combined rates weighted by their denominators?
  • Does the reader also see the absolute numbers?

None of this is advanced, which is precisely why errors here are so damaging. A wrong percentage in front of a CFO undermines trust in every other number on the page.

Written by the RP Strativa Team. Figures in the examples are illustrative.

Leave a Reply

Trending