Year-over-year vs. month-over-month: choosing the comparison

Month-over-month numbers are easy to read and easy to misread. The choice between year over year vs month over month comes down to what you're trying to see. Year-over-year compares a period with the same period a year earlier, which cancels out most seasonal patterns. Month-over-month compares consecutive months, which shows recent direction but carries seasonality with it.

Neither comparison is better in general. Each one has a job, and each one fails in predictable ways. Knowing both failure modes is most of the skill.

What does each comparison actually measure?

A year-over-year change answers: compared with the same month last year, are we up or down? Because July is compared with July, the summer effect is present on both sides and mostly cancels. What remains is closer to the underlying trend, plus whatever was unusual in either year.

A month-over-month change answers: compared with last month, which way are things moving? It responds quickly. If something shifted in the past few weeks, this comparison will show it first. But it also includes every seasonal swing between one month and the next.

Suppose a brand of hot cocoa sells 900 units in November and 1,400 in December. The month-over-month jump looks dramatic. Now suppose the same brand sold 950 in November and 1,500 in December a year earlier. The year-over-year view says December was slightly down. Both statements are true. Only one describes the business.

When does month-over-month mislead?

Most often in seasonal categories. Cold drinks, ice cream, holiday candy, school supplies: any category with a strong annual shape will show big month-over-month moves that are just the calendar at work. Reading those as trends is the most common mistake in retail reporting.

It also misleads when months differ in length or in their mix of weekdays. February has fewer days than March. A month with five weekends can outsell a month with four for reasons that have nothing to do with demand. Adjusting to a per-day basis helps, and seasonal adjustment goes further by estimating and removing the usual pattern.

When is year-over-year the wrong choice?

When last year was strange. If a store group had a supply disruption, a major promotion, or a holiday that fell in a different month a year ago, the year-over-year comparison inherits that oddity. The result can look like a big swing that's really an echo of the base period.

It's also slow. A change that began three weeks ago barely moves a year-over-year number, because the comparison is anchored to last year's level. By the time it shows up clearly, you may have missed the moment to respond.

And it struggles with new items. A product that launched eight months ago has no year-ago base. Any year-over-year figure for it is either undefined or meaningless.

How should you report both?

Lead with the comparison that matches the question. For "is the business healthy?" year-over-year usually fits. For "did something change recently?" month-over-month, ideally adjusted for days and season, fits better.

Then label it plainly on every chart and table. Readers switch between the two without noticing, and a headline that mixes them causes more confusion than either one alone. A short note on what was unusual in the base period saves a lot of follow-up questions.

The July 2026 same-store sales report is one example of how a monthly release frames its comparisons, and earlier releases are on the NRS Insights news page.