Stop using confusing date formats in your charts!

 
 

One of my pet peeves is dates in charts that are ambiguous:

 
 

This can be easily fixed by simply choosing different date formats:

 
 

What, exactly, makes these date formats clear and unambiguous? Two things:

  1. The month is alphabetic (“Feb”, “September”, etc.), not numeric (“02”, “9”, etc.).

  2. The year is four numbers (“2024”, “2015”, etc.), not two (“24”, “15”, etc.)

Here are a few more date formats that meet these criteria and so are clear and unambiguous:

 
 

[Update: Aug. 27, 2024, 6pm EST: This post generated some great discussion on LinkedIn, which I wanted to add to this post.

Some people commented that, if everyone agrees to use a specific date format (most notably, the ISO standard “2024-08-27” format or the “07-27-2024” format that’s commonly used in the U.S.), then we wouldn’t have to worry about readers misreading dates.

The thing is, in order for a standard like that to actually work, everyone would need to follow it without exception. If even a small fraction of report creators don't follow it, the whole thing falls apart because report readers can no longer assume that every report always follows the standard and they'll be forced to double-check what date format is used, every time, by looking for a date like "2024-01-13" or “2024-13-01”, and then inferring the date format used in that particular report.

Ensuring that every report creator on the planet uses the same date format 100% of the time is, for all intents and purposes, impossible, so it's more of a wish than a strategy, IMHO. I think it's easier to just encourage report creators to use alphabetic months and 4-digit years, but then leave the specific date format up to them.

Another reason to not try to enforce a single, standard date format is that different formats might make more sense in different situations. For example, there can be good reasons to list the month first, or the day of the month first, or include the day of the week, the time zone, etc., depending on the context and purpose of the report.

There's also the consideration that most people recognize month names more readily than month numbers (no one I know says that they're "going on vacation in 07"...), so date formats with alphabetic months might be a bit quicker to read.

Some people also commented that purely numerical date formats are better for storing dates in databases and doing certain types of analysis. This is absolutely true, but there’s no rule that says that the format in which dates are stored in a database must be the same as the format in which dates are shown to users, and converting dates retrieved from a database to an easier-to-read format for presentation to users can usually be done with a simple function call. Indeed, it’s often a good idea to store dates in databases as Unix timestamps, i.e., the number of seconds that have elapsed since Jan. 1, 1970, but it would virtually never be a good idea to show dates to users as “1724796066” 😕]

Have something to say about this article?

Reply in the comments below, or comment on the post of this article on LinkedIn.

Want more tips like this? (Like, A LOT more?)

Until 09-04-24 (just kidding 😆)... Until Sept. 4, 2024, readers of my blog can get 25% off my Practical Charts On Demand course​ (6.5 hours of video in 45 lessons) by using the coupon code newsletter-04 at checkout!