You're staring at a spreadsheet. Worth adding: column B says revenue went from $47,000 to $52,000. Column C shows customer churn dropped from 8% to 5.On top of that, 2%. Your boss asks: "So what's the increase? What's the decrease?
And your mind goes blank.
Not because you don't know the words. Because nobody ever taught you how to use them in the wild — with real numbers, real stakes, and real pressure. Less friction, more output.
What Is Increase and Decrease
At its simplest, increase means something got bigger. Decrease means something got smaller. That said, that's it. No Latin roots required.
But here's where it gets useful: in math, business, science, and daily life, we don't just say "it went up." We measure how much* it went up. And we express that change in two main ways — absolute and relative.
Absolute change
This is the raw difference. New value minus old value.
If your rent jumps from $1,200 to $1,350, the absolute increase is $150. If your commute shrinks from 45 minutes to 32 minutes, the absolute decrease is 13 minutes.
Simple subtraction. Nothing fancy.
Relative change (percentage change)
This is where most people trip up. Relative change puts the difference in context — it tells you the size of the change compared to where you started*.
The formula:
(New Value − Old Value) ÷ Old Value × 100
That rent example? 125. Times 100 = 12.$150 ÷ $1,200 = 0.5% increase.
The commute? Times 100 = 28.13 ÷ 45 = 0.289. 9% decrease.
Same absolute numbers can feel totally different depending on the baseline. A $150 increase on a $1,200 rent stings. 2 million budget? A $150 increase on a $1.Rounding error.
Why It Matters / Why People Care
You see "increase" and "decrease" everywhere. Worth adding: performance reviews. Medical charts. Headlines. Dashboards. Utility bills.
And how they're presented changes what you believe.
The framing trap
A supplement company runs an ad: "Users saw a 50% increase in energy!"
Sounds huge. But if the baseline was 2 out of 10 on a subjective scale, and it went to 3 out of 10 — that's technically a 50% increase. Also barely noticeable.
Flip it: "Crime decreased by 20%.In practice, if it went from 500 to 400? But if it went from 5 incidents to 4 in a town of 50,000, that's statistically noise. That's why " Good news? Maybe. That's real.
Context is the whole story. Without the baseline, a percentage is just a prop.
Decision-making runs on this stuff
- Should you refinance? Compare the percentage decrease* in monthly payment against closing costs.
- Hiring a vendor? Their pitch says "30% faster delivery." Ask: 30% faster than what? Your current average? Their competitor? Last year's best month?
- Tracking fitness? "I increased my bench by 15%." Cool — but over what timeframe? Two weeks or two years?
The numbers don't lie. But they omit*. A lot.
How It Works (and How to Calculate It)
Let's walk through the mechanics. Not because you can't Google the formula — because doing it by hand once or twice builds intuition that sticks.
Step-by-step: percentage increase
Scenario: Your freelance income went from $3,200/month to $4,100/month.
- Find the difference: $4,100 − $3,200 = $900
- Divide by the original* value: $900 ÷ $3,200 = 0.28125
- Multiply by 100: 28.125% increase
Round to 28.1% or 28% depending on context.
Step-by-step: percentage decrease
Scenario: Your site's bounce rate dropped from 68% to 52%.
- Difference: 68 − 52 = 16 percentage points (careful — this isn't percent change* yet)
- Divide by original: 16 ÷ 68 = 0.2353
- Times 100: 23.5% decrease
Notice: the absolute* drop is 16 points. The relative* drop is 23.5%. Both are true. Neither tells the full story alone.
When the baseline is zero (or negative)
This breaks the formula.
If you had 0 customers last month and 12 this month, you can't calculate a percentage increase. Division by zero is undefined. Same with negative baselines — a temperature rising from -10°C to -5°C looks* like a 50% increase if you plug it in blindly. But it's not. It's a 5-degree rise.
Workaround: Use absolute change. Or reframe: "We went from zero to 12 customers." Honest. Clear. No math crimes.
Compound changes (the sneaky one)
Say a stock drops 20% on Monday, then gains 20% on Tuesday. Back to even?
Nope.
Start at $100.
- Monday: 20% drop → $80
- Tuesday: 20% gain on $80 → $16 gain → $96
You're down 4% overall.
For more on this topic, read our article on ethnic religion ap human geography definition or check out 50 examples of balanced chemical equations with answers.
Percentage changes don't cancel out. The baseline shifts. This trips up investors, marketers, and anyone tracking metrics month over month.
Year-over-year vs. month-over-month vs. quarter-over-quarter
Same math. Different timeframes. Different noise levels.
- Month-over-month (MoM): Noisy. Seasonal. Good for spotting sudden shifts.
- Quarter-over-quarter (QoQ): Smoother. Standard for business reporting.
- Year-over-year (YoY): Removes seasonality. Best for "real" trend lines.
A retail store seeing a 40% MoM drop in January isn't crashing — it's post-holiday. But a 40% YoY drop in January? That's a fire.
Pick the timeframe that matches the question you're actually asking.
Common Mistakes / What Most People Get Wrong
I've seen smart people make these errors in boardrooms, pitch decks, and published reports. You will too — unless you know what to watch for.
Confusing percentage points with percent change
This is the big one.
Unemployment goes from 5
% to 6%. That’s a 1 percentage point increase. But the percent change is 6 − 5 = 1, divided by 5, times 100 = 20% increase.
Headlines love to blur this. “Taxes up 1 point” sounds mild. “Taxes up 20%” sounds alarming. This leads to both describe the same move. Know which one you’re reading — and which one you’re writing.
Using the wrong baseline
Always divide by the starting* value, never the ending one. And if you’re measuring growth from last year to this year, last year is the denominator. Flip it and your numbers lie in the opposite direction.
Averaging percentages
You can’t average “20% increase” and “30% increase” and call it “25% average increase” unless the bases are identical. In real terms, if one change acted on $100 and the other on $10,000, the weighted reality is closer to 30%. Simple averages of percentages are how bad dashboards get built.
Ignoring sample size
A 50% conversion lift sounds great — until you learn it’s 1 sale out of 2 visitors vs. Small bases produce violent, meaningless percentages. 2 out of 4. Always pair the percent with the raw numbers behind it.
Conclusion
Percentages are a language, not just arithmetic. The formula is easy; the discipline is harder. In real terms, do it by hand once, and the intuition stays. Which means know the difference between points and percent, respect your baseline, watch for compounding, and match your timeframe to your question. After that, the calculators can take over — but you’ll already know when they’re wrong.
down 4% overall.
Percentage changes don't cancel out. The baseline shifts. This trips up investors, marketers, and anyone tracking metrics month over month.
Year-over-year vs. month-over-month vs. quarter-over-quarter
Same math. Different timeframes. Different noise levels.
- Month-over-month (MoM): Noisy. Seasonal. Good for spotting sudden shifts.
- Quarter-over-quarter (QoQ): Smoother. Standard for business reporting.
- Year-over-year (YoY): Removes seasonality. Best for "real" trend lines.
A retail store seeing a 40% MoM drop in January isn't crashing — it's post-holiday. But a 40% YoY drop in January? That's a fire.
Pick the timeframe that matches the question you're actually asking.
Common Mistakes / What Most People Get Wrong
I've seen smart people make these errors in boardrooms, pitch decks, and published reports. You will too — unless you know what to watch for.
Confusing percentage points with percent change
It's the big one.
Unemployment goes from 5% to 6%. That's a 1 percentage point increase. But the percent change is 6 − 5 = 1, divided by 5, times 100 = 20% increase.
Headlines love to blur this. "Taxes up 1 point" sounds mild. "Taxes up 20%" sounds alarming. But both describe the same move. Know which one you're reading — and which one you're writing.
Using the wrong baseline
Always divide by the starting* value, never the ending one. If you're measuring growth from last year to this year, last year is the denominator. Flip it and your numbers lie in the opposite direction.
Averaging percentages
You can't average "20% increase" and "30% increase" and call it "25% average increase" unless the bases are identical. But if one change acted on $100 and the other on $10,000, the weighted reality is closer to 30%. Simple averages of percentages are how bad dashboards get built.
Ignoring sample size
A 50% conversion lift sounds great — until you learn it's 1 sale out of 2 visitors vs. Consider this: 2 out of 4. Consider this: small bases produce violent, meaningless percentages. Always pair the percent with the raw numbers behind it.
Conclusion
Percentages are a language, not just arithmetic. The formula is easy; the discipline is harder. Day to day, do it by hand once, and the intuition stays. Know the difference between points and percent, respect your baseline, watch for compounding, and match your timeframe to your question. After that, the calculators can take over — but you'll already know when they're wrong.
