Percentage Change Calculator: Increase, Decrease, and Why the Math Is Asymmetric

Percentage change is one of those calculations everyone thinks they know until a real number proves them wrong. You compare two figures, you reach for the percent button, and somehow you end up with two different answers depending on which value you divide by. Then a colleague says revenue "rose 3 percent" when they meant 3 percentage points, and a quarterly report quietly drifts off by an order of magnitude.

This post walks through the formula, the single worked example that exposes the trap most people fall into, and the three places percentage change actually shows up in daily work: price moves, metrics, and salary.

The one formula you need

Percentage change between an old value and a new value is:

change% = (new − old) / old × 100

The old value is the denominator. That is the whole game. Revenue, stock prices, conversion rates, headcount — anything where you have a "before" and an "after" — divides by the starting number, in absolute terms, so the proper definition is ((new − old) / |old|) × 100.

A quick sanity check: revenue went from 150 to 180.

(180 − 150) / 150 × 100 = 30 / 150 × 100 = +20%

If you had divided by 180 instead, you would get +16.7%, which is wrong. It is wrong silently — the number still looks plausible — which is exactly why the mistake survives in spreadsheets for years. Pick the old value as the denominator every time.

The asymmetry trap: why +25% and −25% are not opposites

Here is the example worth memorizing, because it breaks the intuition almost everyone carries.

Start at 80 and go to 100:

(100 − 80) / 80 × 100 = 20 / 80 × 100 = +25%

Now go back the other way, from 100 to 80:

(80 − 100) / 100 × 100 = −20 / 100 × 100 = −20%

Same two numbers, same gap of 20 units, but one direction is +25% and the other is −20%. That is not a rounding quirk. It happens because each percentage is taken on a different base. Going up, you divide the 20-unit gap by 80. Coming down, you divide it by the larger 100. A bigger denominator produces a smaller percentage, so the trip back is always a smaller percentage than the trip out.

The practical consequence: a +50% gain is not undone by a −50% cut. Add 50% to 100 and you reach 150; subtract 50% of that (75) and you land at 75, not 100. To erase a +50% gain you actually need a −33.3% drop. Anyone quoting "we recovered the loss" after a symmetric-looking swing is usually wrong, and the only way to be sure is to run the actual numbers. The percentage change calculator locks the denominator to the old value so you cannot accidentally pick the wrong base mid-calculation.

Percent versus percentage points

This is the second confusion, and it costs more credibility than the first because it usually happens in public.

If an interest rate moves from 5% to 8%, two true statements describe it:

• It rose 3 percentage points (8 minus 5).
• It rose +60% in relative terms (3 is 60% of the starting 5).

Both are correct, but they answer different questions. "Three percentage points" describes the gap on the percentage scale itself. "Plus sixty percent" describes how much bigger the rate got relative to where it started. Say "the rate rose 3 percent" and a careful reader hears a tiny +0.15-point nudge — not the real move at all.

Polls do this constantly. Support climbing from 40% to 44% rose 4 percentage points, or roughly 10% in relative terms. Conversion lifting from 2.4% to 3.1% is +0.7 percentage points but a much punchier +29.2% relative change. Both framings are honest; what is dishonest is leaving the unit off and letting the audience guess. Lead a write-up with the relative figure if you want impact, footnote the percentage-point figure to stay accurate, and always name which one you mean.

Where this actually matters

Price changes. A $59 item marked down to $39 is a −33.9% discount off the original. But the markup needed to climb back to $59 later is +51.3%, not 33.9%. Same two prices, two different percentages, because of the base-shifting we just covered. Retailers and anyone reselling inventory need both numbers, and they are never mirror images.

Metrics. Quarter-over-quarter, week-over-week, year-over-year — every growth rate on a dashboard is a percentage change with the earlier period as the denominator. The discipline of locking the denominator to the old period is what keeps the finance team from re-deriving the same QoQ number three different ways. When a metric is itself a percentage (a conversion rate, an SLA-attainment rate), that is where the percent-versus-points distinction becomes mandatory.

Salary. A raise from $72,000 to $80,000 is (80,000 − 72,000) / 72,000 × 100 = +11.1%. Read it in reverse — if a pay cut took you from $80,000 back to $72,000 — and that is only −10%. The same dollar gap, framed as a raise, sounds larger than the same gap framed as a cut, and now you know precisely why. The percentage that gets you up is always larger than the percentage that brings you back down.

How I use it

I report on a small product surface, and the metric I watch most is a conversion rate. Early on I wrote "conversion up 0.7 percent" in a weekly note, and a teammate read it as a rounding-error non-event — because 0.7 percent of a 2.4% rate is almost nothing. What I meant was +0.7 percentage points, which on a 2.4% base is a +29% relative jump. After that I started running both figures every single time: the percentage-point gap for the honest footnote, and the relative change for the headline. It takes ten seconds and it has saved me from underselling good results and from overselling mediocre ones. If you are reporting any rate-of-a-rate, build the same habit.

A second-opinion check

Percentage change is the engine behind a lot of business math you might not label as "percentage change" — return on investment, for instance, is a percentage change between money in and money out, with the starting amount as the denominator. If you are working through profit or campaign returns, the ROI calculator applies the same (new − old) / old logic to dollars invested and dollars returned, and seeing it framed that way makes both calculations click.

Whenever two numbers feel like they should be symmetric and the math says otherwise, trust the math. The old value is the denominator, percent and percentage points are different units, and the trip up is never the same size as the trip back down.

---

Made by Toolora · Updated 2026-06-13