Emoji are a graphical shorthand for emotional states, jokes, and nuances of language, so it’s particularly problematic when your friend sees a different emoji than the one you sent. Here’s why your messages might not be coming through like you intended.

How Emoji Work: A Code for Every Smile

We, the end users, only see the graphic fruit of the emoji system. Beneath all those millions of smiley faces, hearts, and tiny piles of poop people send every day, there’s a detailed—and standardized!—code system designed to ensure everyone sees the same thing.

The backbone of emoji is shared by the very text messages their embedded into: Unicode. Unicode is a computer industry standard, dating all the way back to the 1990s, that ensures all the world’s writing systems and symbols will be correctly displayed across electronic devices. The entire standard contains more than 128,000 characters across 135 modern and historical writing systems, including symbols.

When emoji were in their infancy in the 1990s,  telecommunication providers in Japan hijacked some unused entries in the Unicode system to correspond to facial expression emoji. The practice wasn’t standardized at the time but, over the years as emoji grew in popularity and were adopted for use outside of Japan, the Unicode Consortium got involved and emoji were standardized by linking specific emoji to specific codes. In this fashion, just like the capital letter A in Latin script is linked to the code U+0014, the basic smiley face emoji could be forever linked to code U+263A.

How Emoji Fail: Design Differences, Expanding Standards, and Old Phones

Given that every single emoji has its own unique, sandardized code, how exactly does it fall apart?

Designer Interpretation: Not All Smiles Are Equal

First, it might help to think about emojis like letters. Yes the Unicode standard ensures that U+0014 is the capital Latin script letter A, but what font the letter is displayed in has a big influence on how we interpret it. Some fonts are utilitarian, some fonts are styled after fantasy script, some fonts are silly, and what font a designer chooses for display changes how we see something as simple as a letter.

The same concept is in effect with emoji. Unicode may say “U+263A is a basic smiley face!” but how that particular basic smiley face looks is up to the people who developed the platform you’re using to send and receive messages.  The following are real world examples of how designers at different companies interpreted “Smiling Face”.

From left: Apple, Google, Microsoft, Samsung, and LG

While there are some differences between the faces—some have rosy cheeks, some are smiling so hard their eyes are crinkling up with happiness—the general message is pretty clear. It would be difficult to interpret any of these symbols as anything but a happy face.

But other symbols, even when it seems like they should be straightforward, aren’t so clear. Here’s what U+1F62C, the “Grimacing Face” looks like across different platforms.

From left: Apple, Google, Microsoft, Samsung, and LG

Apple and Google’s interpretation of it both have a bit of an uncanny valley vibe, like two robots trying to imitate a human grimace. Microsoft and LG seem to have captured the true spirit of a grimace in that their emojis look like they’re actually grimacing at something bad, like a kid about to get hurt or awful news. Samsung, on the other hand, managed to interpret “grimace” as “smile with a sneer as if you just drugged your enemy’s drink while they were in the restroom”. If you send that emoji from your LG phone with the sense that you’re saying “Oh jeez, that’s awful!” a recipient on Samsung phone gets treated to that creepy “I know where you live!” face.

Updates and Old Phones Introduce Hiccups

In addition to the headache of different styling, there’s an additional wrench in your ideographic communications: an expanding emoji library combined with old and infrequently updated phones. If you have a new phone and your recipient has an old phone, or vice versa, there’s a chance that the emoji won’t match across devices, even if the devices are from the same developer.

For example, in early versions of the emoji Unicode set, the emoji for “dancer” was either a gender neutral stick figure or a little cartoon man dancing. Later that same emoji code was shuffled around in a revision of the standard, so on devices using the newer versions, it isn’t a stick figure or man, but a woman in a red salsa dress. Here’s how that one emoji can vary wildly depending on the age of the phone and the platform.

From left: Apple, Google, Microsoft, Samsung (new), Samsung (old)

Depending on which platform you’re on and which platform your recipient sees it on, your message might be “I want to dance with the lady in red!”, “I want to dance with Dora the Explorer!”, or “I want to dance Generic White Man in Yellow Pants!”

RELATED: How to Change Emoji Skin Tones on iPhone and OS X

Speaking of generic people, updates aren’t the only thing that causes trouble. Old phones running old standards trying to parse out codes they don’t even recognize can cause some really awkward hiccups too.

Back in 2015 when Apple rolled out a plethora of new emoji, people applauded them for including a bunch of emoji that displayed diverse skin tones, family structures, and so on. Unfortunately, when people with the updated versions of iOS sent the new emoji to people with older versions of iOS, rather than just display an empty placeholder (as in common on other mobile operating systems or in communications between platforms), the older versions of iOS tried to do a very strange job translating the new emoji.

Rather than display some blank placeholder, the older versions of iOS translated all the new skintone-themed emoji into the white version of that emoji plus an alien symbol. For a diversity-themed update, that’s more than a little bit awkward.

Similar, albeit not quite as awkward, situations can arise when old phones attempt to translate new emoji. At best, there is just a blank placeholder, at worst a benign message can become an insulting one.

Emoji Fu: Skills to Save Face

So now that you know how emoji work and where things can fall apart, what can you do to decrease emoji-related communication snafus? Although increasing adoption and compliance with emoji standards helps everyone, here are a few simple tricks you can use.

When it doubt, skip the emoji altogether. The newer and more specific the emoji, the higher chance that it won’t get parsed correctly on the other end. Not only that, but research has shown that how people interpret emoji varies wildly even when they’re looking at the same emoji and even more so when the face differs across platforms.

Apple’s grinning emoji is a little too similar to their grimacing emoji.

The chart above, from research conducted by the GroupLens research lab at the University of Minnesota, highlights how the should-be-the-same emoji illicit different reactions. Further, in the same study, even when looking at the exact same emoji, there was a 1 out of 4 chance that the viewers would disagree about whether the face was positive or negative. While each new revision of the emoji codes tends to make the emoji across platforms more similar, it’s something to keep in mind.

If you are using emoji, however, you can decrease the chance of awkwardness by favoring older and established emoji. The simple smiling face, heart, thumbs up hand gesture, and the like, have been part of the emoji code for years now and there is very little variation between them.

Finally, if you want to really do your homework (or you’re that worried about making a wrong impression with a potential romantic partner) you can always hit up the numerous resources that catalog the emoji codes like the Emojipedia—a service so thorough you can not only review old versions of the emoji code but they also, thoughtfully, warn you when a particular emoji is known for problematic display across platforms, like the aforementioned grimacing face.

Profile Photo for Jason Fitzpatrick Jason Fitzpatrick
Jason Fitzpatrick is the Senior Smart Home Editor at How-To Geek. He has over a decade of experience in publishing and has authored thousands of articles at How-To Geek, Review Geek, LifeSavvy, and Lifehacker. Jason served as Lifehacker's Weekend Editor before he joined How-To Geek.
Read Full Bio »