Google Analytics: How to Select Your Date Range
You’ve got a problem - you need to figure out which days are best for conversions on your landing pages.
The answer is easy, you log into google analytics and pull up a week’s worth of data and you find that Monday’s are the best day for conversions. This answer may or may not be right. It often depends on a number of circumstances.
The more coherent way to seek out the right answer is to normalize the numbers. A week is a small sample size in time compared to a month, 6 months or a year. Even a linear choice such as a year may not be enough for insights.
It’s when you start comparing data week over week, month over month or year over year that the real insights begin to appear.
In a linear example, your date choice is simple - you choose based on your need:
- A Week - 1st week, 2nd week, 3rd week etc
- A Month - 1st month, 2nd month, 3rd month etc
- A Year - 365 days
In a week over week / month over month / year over year example - you choose on your need & the day of week.
- Monday through Sunday for a week over week, then compare to either previous period or to previous year, same week
Best Practices for selecting date ranges:
- Start the range at the beginning of the week (Sunday)
- End the range at the last day of the week (Saturday)
- Never use the current day in the range - the number will change
- Go for 7 day chunks at minimum - day on day may work however there can be extenuating circumstances that might cause a massive drop or lift
Look at the below examples:
I clicked on the 18th of March in the above example - though the dates are aligned, the day of the week is not. This is remedied by making the choice of day of the week as the priority over choosing the date of the month.
In the above example - the dates are March 22nd, 2016 & March 24th, 2015 - however they are both a Tuesday.
Even though the above example shows a mismatch in the date, the days are aligned - making it possible to arrive at an answer that isn’t skewed by the “seasonality” or the mismatched days.
Choosing your the right day to start and end for the time categories is critical.
Step #1 - Choose the time period you want to compare against
- Day 1 can be any day, though beginning of week is best
- Last day should be the day of the week that precedes the 1st day
Step #2 - Click on the “Compare to” box - default is previous period
- The previous period does not compare based on day, it starts with date
- This example is a Month over Month, Year over Year example
Step #3 The Fix
- Move the Start day to align with Wednesday
- Move the End day to stop on Tuesday
Now you’re set.
Here’s a Month over Month example which coincides with the previous period:
Note that the chosen range is symmetrical - you’re golden, get to analyzing that data!