The Math Nobody Questions Until It Matters
You take two numbers, figure out how much they changed, then express that change relative to where you started. That is basically the entire concept. The formula is (new minus old) divided by old, times 100. You get a percentage. People usually get it right on paper and then second-guess themselves when it comes to budgets, sales reports, or comparing prices across suppliers. I have seen it happen repeatedly. Here is what actually matters beyond the basic calculation. The denominator has to be the original value, not the new value. Swap those and you get a completely different number that describes a different thing entirely. This mistake shows up in quarterly earnings reports more often than you would expect from grown professionals with spreadsheets.
How To Find Percentage Increase
First, subtract the starting number from the ending number. Call that your absolute change. Then divide that result by the starting number. That gives you a decimal representing the proportional change. Multiply by 100 to convert it into a percentage. Done. Nothing fancy about it. For example, a product cost you 80 dollars last quarter and now costs 110 dollars. The difference is 30 dollars. Divide 30 by 80 and you get 0.375. Multiply by 100 and you are looking at a 37.5 percent increase. You can double-check by reversing it: 80 plus 37.5 percent of 80 equals 110. The math holds up. The tricky part is when the base value is small or fluctuates wildly between periods. I was working on a supply chain analysis last year comparing monthly order volumes for a niche component. The previous month had only 12 units and the next month jumped to 47 units. The percentage increase was 291.7 percent. On paper that looked like a massive spike. In reality the absolute change was 35 units and the business context was a single large customer order that should not have been treated as a trend signal. That is the kind of edge-case where percentage increase without absolute context will mislead anyone who reads the report without thinking.
Another scenario I deal with regularly involves percentages near zero. When your old value is very close to zero, even a tiny absolute change produces an enormous percentage that is essentially meaningless for decision-making. If revenue went from 0.03 dollars to 0.05 dollars, that is technically a 66.7 percent increase. It does not mean anything practical. You should always pair percentage changes with the underlying absolute figures so the reader gets the full picture.
Get the Full Details
Where This Gets Messy in Practice
Compound changes over multiple periods throw off people who assume percentages add linearly. If something goes up 20 percent and then up another 20 percent, that is not a 40 percent total increase. It is 44 percent. Each step compounds on the new base. This is basic but it is the single most common error I see in financial models. Someone projects three years of growth at 15 percent per year and simply adds 15 plus 15 plus 15. The actual cumulative increase is closer to 52 percent, not 45 percent. The compounding effect matters enough that it shifts budget assumptions materially. Directionality is another detail worth noting. Percentage decrease is calculated the same way but you subtract the later value from the earlier value. A drop from 200 to 150 is a 25 percent decrease. The math is symmetric. The interpretation is not, which is why people get tripped up when discussing losses versus gains in the same conversation. There is no free spreadsheet tool that handles all of these cases automatically and correctly without you understanding the underlying logic. Excel and Google Sheets have the PERCENTCHANGE formula in some versions and you can build a simple formula in any other version, but the tool will never tell you whether a reported percentage is meaningful in context. You have to bring that judgment yourself.
If you are working with large historical datasets or need to compute percentage changes across dozens of columns automatically, a short Python script using pandas is far more reliable than manual entry. You can set up a script that computes percentage changes for an entire column in seconds and flags values where the base is near zero. The setup takes about ten minutes and saves hours of spreadsheet work over time. I use this approach for monthly reconciliations instead of building formulas by hand. The main limitation of relying purely on percentage increase is that it strips away scale. A 50 percent increase sounds dramatic until you see it came from two units sold to three units sold. Always show the raw numbers alongside the percentage. Anyone who makes decisions based only on percentages without seeing the base is working blind. There is also the issue of negative bases. If your starting value is negative, percentage change becomes mathematically valid but logically confusing. A change from negative ten to negative five is technically a 50 percent increase, but most people read that as a positive improvement when the arithmetic is pointing at something else entirely. In those cases, it is better to describe the absolute movement and skip the percentage altogether.
When to Use It and When Not to
Percentage increase works well for comparing growth across similar categories, tracking price changes over time, and communicating relative shifts in reports. It breaks down when the base is unstable, near zero, or negative. It also does not help you understand the underlying cause of the change. That requires looking at the data behind the numbers, not the percentage itself. I keep the calculation simple in my own work because simple is less likely to produce the wrong answer. Write down the old value, write down the new value, subtract, divide by the old, multiply by 100. Verify the base is positive and reasonably sized. State the absolute change next to the percentage. That is all it takes to avoid most of the errors I see in practice.
