This year Easter fell on April 1, against April 16 last year, so month-against-month comparisons mislead. Compare Easter week with Easter week, March and April combined, and weeks matched by day before drawing conclusions.

The month that looks too good

Easter Sunday was yesterday. Last year it fell on April 16.

That shift moves a great deal of activity from one month to another. This year Good Friday fell in March, along with the whole run-up to Easter. Last year both were in April.

So for many businesses, March this year will look stronger than March last year and April weaker, or the reverse for those that close over Easter, without anything about the business having changed.

Anyone reading monthly reports in the next few weeks, including whoever sets the budgets, needs to know that before drawing conclusions.

Who sees the distortion

The second group is the one that usually misses it. A business closed on Good Friday lost a working day in March this year that it lost in April last year.

Working days, counted

Count the days your business was open, and the comparison changes.

For a business open on weekdays and closed on Good Friday, March last year had 23 working days. March this year had 21: one fewer weekday in the calendar, and Good Friday on top.

April reverses it. Last April had 19 such days; this April has 21. Close on Easter Monday as well and each April loses one more, leaving the same gap.

Two days in a month of twenty-odd is a large share for a business whose enquiries arrive on working days. Divide by working days before comparing, or compare the two months together.

Compare Easter with Easter

For businesses whose Easter is busy, the useful comparison is the holiday itself, wherever it fell.

Compare the week leading up to Easter Sunday this year, March 26 to April 1, with the same week last year, April 10 to 16. Then compare the weeks either side in the same way.

In Universal Analytics, that means a custom comparison rather than the previous year option: set both date ranges by hand, matched to the holiday.

Compare by week, matched by day

Even without a holiday, months make awkward comparisons. This March had five Saturdays; last March had four.

Comparing whole weeks, Monday to Sunday, removes that problem, since every week contains one of each day.

When comparing with the previous year in Analytics, the option to match the day of the week lines up Mondays with Mondays, rather than with whichever weekday fell on the same date last year.

Put March and April together

The simplest fix for a monthly report is to stop reading March and April separately this year.

Both Easters fall inside the two months combined, so March and April together compare fairly with March and April last year, at least as far as the holiday goes.

The first quarter is affected too, because Good Friday and part of the Easter run-up fell inside it this year and outside it last year. Read quarterly figures with that in mind.

A worked example

A butcher shop in a suburban shopping plaza takes online pre-orders for holiday roasts and hams, collected in store.

Its March report showed pre-order visits and orders well above last March, and the owner was ready to credit the new email newsletter and spend more on it.

Comparing Easter week with Easter week told a different story. The weeks before each Easter looked much alike, and most of the March gain was the holiday arriving early.

The newsletter may still have helped, and its own tagged links can show what it brought in directly. The decision about spending more on it now rests on those figures, not on a calendar shift.

April will look poor for the same reason, and the owner now expects it.

Ads and email follow the same calendar

Advertising and email reports carry the same distortion. An Easter campaign that ran in April last year and in March this year makes both months look strange in AdWords.

Spend, clicks and enquiries all moved with the holiday, so this March will look expensive against last March, and this April cheap against last April.

Compare campaign with campaign, and the Easter emails with last year's Easter emails, rather than one month's totals with another's.

Notes in the report itself

Universal Analytics lets you add annotations: short notes pinned to a date on the charts. Add one for Good Friday, Easter Sunday and Easter Monday, and one for any closure.

Annotations can be shared with everyone who uses the account, so whoever reads the report next year sees why March jumped without having to remember.

Add school spring breaks as well. They move from year to year too, and a business serving families feels them as much as Easter.

Do the same in whatever spreadsheet holds your sales figures. A column for holidays turns a puzzling number into an obvious one.

Next year it runs the other way

Easter next year falls on April 21, almost three weeks later than this year.

March next year will look weak against this March, and April strong, for exactly the same reason. Anyone setting next year's targets month by month should allow for it now.

Write the dates into the annotations and the budget spreadsheet while the reason is fresh.

The counter-case

Some businesses can ignore all of this.

A business with no holiday pattern, open every day and serving customers whose needs do not follow Easter, will see little difference between the years.

Very small numbers also make the adjustment pointless. When a month brings a dozen enquiries, one or two either way is noise, and a full year's comparison says more.

For everyone else, compare Easter with Easter, weeks with weeks, and March with April together, before deciding anything.

This month's report

  1. Note both Easter dates before reading anything.
  2. Count working days in each month.
  3. Compare Easter week with Easter week.
  4. Match days of the week in comparisons.
  5. Combine March and April for the monthly view.
  6. Add annotations for the holidays.
  7. Warn whoever reads the report.

Step seven is the one that prevents a bad decision. The figures will mislead anyone who does not know about the shift, and a sentence at the top of the report costs nothing.

Another report that rewards careful reading is covered in site search and what visitors type into the box.

Frequently asked questions

Why does March look so strong this year?

For many businesses, because Easter came on April 1 this year rather than April 16, bringing Good Friday and the Easter run-up into March.

How should I compare the Easter periods?

Compare the week before Easter Sunday this year, March 26 to April 1, with the same week last year, April 10 to 16, using a custom date comparison.

How many working days did March lose?

For a business open on weekdays and closed on Good Friday, March had 21 working days this year against 23 last year. April has 21 against 19.

Should I compare March and April together?

Yes, for monthly reports this year. Both Easters fall inside the two months combined, so the pair compares fairly with the same pair last year.

What are annotations in Analytics?

Short notes pinned to dates on the charts in Universal Analytics. Adding them for holidays and closures explains odd figures to whoever reads the report later.

Will next year's comparison be affected?

Yes, the other way round. Easter next year falls on April 21, so March will look weak against this year and April strong.

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.

Monthly report about to reach the owner?

Send it to us first. We will mark what Easter moved and what actually changed.

Start a Conversation