For business

Holidays move. Your growth numbers move with them.

Eid, Diwali, Lunar New Year, Carnival, Easter and Thanksgiving land on a different date every year. When one slips into another month, month-over-last-year growth jumps in one month and sinks in the next, even though nothing about the business changed. Below: check any year, see how big the effect is and when it evens out, plan the year ahead by market, and take it out of your numbers. Planning dates instead? See Eid dates 2027.

Was it in the same week last year?For Eid, almost never. The Islamic year is about 11 days shorter, so Eid comes about 11 days earlier every year. Does it change monthly growth?Yes, whenever the holiday, or the busy days before it, crosses into another month. Is the effect permanent?No. It's a shift in timing, so the two months cancel out. The full year only changes when Eid crosses New Year (around 2033). What about next year?Working days per month, market by market, and a heads-up for every holiday that changes month. Is there a method?Yes: measure the holiday's effect, take it out of both years, then compare. It works for any moving holiday.

Was it in the same week last year?

Where it lands, year by year

What it does to monthly growth

A made-up business that grows exactly 8% a year, every day, with one holiday pattern on top (). Growth is sales per day this year against the same month last year, so leap days don't count.

Reported growthHoliday-adjusted growthUnderlying growth, 8%Holiday this yearlast year
Show the numbers

Does it even out?

Reported growth, year to dateUnderlying growth, 8%

Mostly, yes. A moving holiday doesn't add or remove sales; it moves them. What one month gains, a neighbouring month gives back, so growth year to date is off only between the two months and back on track once both are in.

Quarters are affected when the move crosses a quarter end: Eid al-Fitr moved from April 2024 (Q2) to March 2025 (Q1), so in 2025 Q1 looked better and Q2 worse than they really were.

The full year is affected only when the holiday, or the days around it, cross New Year. Because the Islamic year is shorter, that happens for Eid al-Fitr around 2032 to 2034: Eid falls twice in 2033 (2 January and 23 December), and the busy days before the January Eid land in December 2032. For the retailer in this example, 2032 grows 9.2% and 2033 only 6.7%, on the same underlying 8%. Pick those years above to see it.

Long periods never quite line up. Ramadan lasts about 30 days, so it almost always spans two months, and the split changes every year. A business that's slower or busier through the whole month sees some distortion every year, not only when Eid itself changes month.

The year ahead, by market

Every month and market picked: working days both years, the difference, the holidays and the windows, for a budget sheet.

Working days are weekdays (each country's own weekend) minus the public holidays most people have off, from TimeBento's holiday table; official holidays most offices work through are listed but not counted. Weekends worked to make up for a long break (China, Vietnam) count as working days where they've been announced; China announces next year's in November or December, so its numbers for next year can still change. Italic lines are the business windows from the checker above, with the days they cover in that month: examples of where money moves, not measurements.

Try it with your own numbers

Paste daily figures (a date and a number per line, from a spreadsheet or a CSV export) covering at least the year you pick and the one before. The page takes the holiday window out of both years with the method below, using the holiday, year and business chosen above, and shows reported against adjusted growth. It's worked out in your browser: nothing is sent anywhere.

A method for any moving holiday

This is the idea statistics offices use to publish calendar-adjusted figures, cut down to something a finance team can run in a spreadsheet. It's what the blue bars above do.

Define the windowThe days the holiday really changes for your business: the ten shopping days before Eid, the Eid days themselves, the whole of Ramadan, the week after Lunar New Year while factories restart.
Find the window every yearFrom a holiday table for the countries you sell in. The page's own table has the dates; moon-sighted dates can differ by a day between countries.
Measure the effectCompare the window's days with ordinary days around it (four weeks either side). Better, fit a regression over several years with the window's share of each month as a variable.
Take it out of both yearsReplace the window's days with the ordinary level, in this year and last year, then compute growth. Report it next to the unadjusted figure.
// for month m of year y
effect(m, y)   = Σ over window days d in m:  sales(d) − baseline(d)
adjusted(m, y) = sales(m, y) − effect(m, y)
growth(m)      = adjusted(m, y) ÷ adjusted(m, y−1) − 1

// baseline(d): average day in the 4 weeks before and
// after the window, outside any holiday window

Simpler options, when there isn't time to adjust:

  • Combine the months the holiday moves between. China's statistics bureau publishes many January and February figures as one combined number, because Lunar New Year moves between them.
  • Compare holiday to holiday: the 30 days before Eid this year against the 30 days before Eid last year.
  • Lead with year to date or a rolling three months in the months the holiday moves.

Statistics offices do the full version with regression: the US Census Bureau's X-13ARIMA-SEATS estimates Easter effects this way, and holiday regressors built the same way are used for Lunar New Year, Ramadan and Eid.

Where this could go

This page is a proof of concept. The numbers above are simulated, and the holiday patterns are examples, not measurements.

Want this for your markets and your own numbers?We're building a team plan: the year ahead for every market you sell in, your own holiday windows, monthly reports adjusted for moving holidays, and a heads-up before each one. Tell us what you'd use it for, and we'll let you know when it's ready.

Join the waitlist