Stop using icons in data tables
medium.com
medium.com
Stop using unlabeled icons in data tables.
It says, "Norman Nielson argues that text + icon has the highest cognitive recall and lowest error rate"
Here's what the Nielsen Norman Group says about Icon Usability: https://www.nngroup.com/articles/icon-usability/
The conclusion: "Always include a visible text label. As Bruce Tognazzini once said, 'a word is worth a thousand pictures.'"
Here's the quote in context: https://www.asktog.com/columns/038MacUITrends.html
"In 1985, after a year of finding that pretty but unlabeled icons confused customers, the Apple human interface group took on the motto 'A word is worth a thousand pictures.' This still holds true."
I was wondering... is OP in fact saying it should be just words, without the icons next to them? Which is an opinion you could have, I think it's not entirely unreasonable (although as you say Nielson contradicts), but I wasn't entirely confident it is the point OP is making... or are they just talking about unlabelled icons (few if any? I'm not sure?) in those screenshots?
I still am not totally sure what OP is arguing.
Other people in the thread I see are arguing the OP is saying the icons next to text would be better off as just text without icons, the icons still make things messy and increase cognitive load. (which again, I think is worth considering at least, sure!) But you don't think it is?
I think I'm just going to put this away concluding it's a very unsuccesful article at whatever it's trying to do.
I agree icons are not always the best option, but I'd need better arguments (and bad examples) than this to stop using them altogether.
- icons really work only as a small set (e.g. checkmarks versus crosses or similar) – so the only useful application is also in a context which is already highly selective and with low ambiguity. (Familiarity helps and should be a major concern.)
- icons are useful for scanning for rows with a given property – however, to scan icon + label we must not only scan horizontally as well as vertically, but have also to scan items with varying column alignment, which isn't favorable for a vertical scan, either.
So, if it is a case for icons, say, each row has a state out of a set of three, there shouldn't be need for an accompanying text label (this can be learned from a tooltip or a legend), which is also apt to destroys most of the advantages of having icons in the first place. Having both just introduces a mode change, which isn't helpful for scanning, either. If you must have both, put them in separate (adjacent) columns. (In this context, we may note that tables are really about reducing redundancy, though.)
Moreover, to be meaningful in tables (or menus – yes, I'm speaking of recent macOS), icons should only be used to represent state or to represent actions that modify state. There are, of course, exceptions to this, but this may be still a viable starting point…
I think the reference to Norman Nielsen was confusing. At the end of the article the author likens icons to visual speed bumps in a data table. The implication being to avoid them or at the very least use sparingly.
If you have many text options, highlighting some specific ones with an icon is benefitial, as you have an easy way to differenciate it from the "less important" ones without icon. If everything has an icon, that benefit goes away and you could just as well use text.
As computer OSes became globalized, i18n and l10n began to demand that we abandon English words to describe GUI actions. I believe this was the correct decision.
So I viewed icons+text in a GUI as transitional, a learning period. I learned to recognize the icons for what they represent, the skeumorphism of it all, while the icons plus text were displayed on my screen (I mostly used Ubuntu and KDE.)
When the time was right, I toggled the setting to remove the text altogether. I was ready to spread my wings and fly through an icon-rich UX that no longer relied on words or text to convey rich meanings.
Now this means that I can make mistakes or I need to experiment in terms of "What does THIS button do?!" in order to learn new icon vocabulary. But I definitely prefer it now to the English text, and really if anyone uses a modern device proficiently, they've picked up a decent vocabulary for this already.
Many people neglect to take into account that, for example, an app in an app store is represented more strongly by its chosen icon and branding than its title! Google Play lets us search by name, but that is not very efficient when I'm searching for a particular icon by description!
It also presents challenges to verbal tech support, whether by phone, by Meet, or over the shoulder, because how do you describe graphical icons and gestures? How to translate them back into English words? UI/UX designers have a hidden vocabulary for controls and widgets that aren't common knowledge, but can really enhance understanding once we can name them the same way the pros do.
Yes, icons can be used badly. Yes, people who are still learning may lean on text labels too. But graphical UIs are mature and we all need to acquire a vocabulary for these, so that information is conveyed across national and linguistic barriers.
Or we could just translate the text?
Colors and pictures aren’t exactly free of cultural baggage, either.
> Colors and pictures aren’t exactly free of cultural baggage, either.
I remember being really confused that I had to click on a picture of a frying pan for "search".
If I were in E.U. I would be afraid of needing to learn enough of 2 dozen assorted languages vs. learning an icon interface once. It would seem that human languages are not as inclusive or universal as icons.
You can do all the l10n you want, but who translates all of them and keeps current?
Your "app" icon says nothing about the code quality of your app and coming with such ideas: " the user will press the icons to see what they do" means: uninstalling.
"Do you want to delete all your files ? [Yes] Remind me later[]"
I haven't found a good general solution to that. Hover doesn't work on mobile. A legend map is hard to locate. "Expert mode" introduces new problems.
Sometimes I'll just use an unlabeled button and make whatever it is undoable, so that users can just click and discover. But that's hostile to completely new users.
This is a lesson to be learned by people who want minimalist UIs.
Is this just a very strange typo or is it a person unrelated to Jakob Nielsen and Don Norman?
All the example table images seem fine, and have no captions saying whether they're supposed to be examples of good usage or bad usage.
So either I have no idea what "bad" examples of icon usage are because the author doesn't show any, or the author thinks some or all of them are bad when, to me, the icon+text+color examples seem great (and one figure caption indicates icons+labels are best)?
Yet the author continues to argue against icons and to use text instead? But never says whether icons+labels are actually better than just text, so we should use them in combination?
I'm baffled. For an article arguing for greater clarity, the article itself couldn't be less clear.
In a data grid or table the relative cognitive load of the page is already very high. Adding iconography to the table body content is often unnecessary and increases visual noise, processing requirements, and generally reduces readability/scanability.
I've always felt that icons in this context are a risk or liability instead of a strength. I decided to info dump my findings to my team then published it as an article.
I probably could use a good editor to help me next time!
Everyone would expect that you show example of good and bad UX for each point that you are making.
Making me see confirm by myself with example that doing X is clearly looking nicer to understand than doing Y.
Your first image has zero icons in the rows. It has album covers but those aren't icons.
Your second and third images show very usage that combines text, color, and relatively standard icons like checkmarks, X's, or in-progress. These are good and if you're trying to suggest these reduce readability, scannability, or add noise, then I'm frankly baffled.
Your third image also shows profile images, but again those are not icons.
So what are you arguing against? I can't ever remember coming across icons in a datagrid that "added to the relative cognitive load". And if you're arguing against checkmarks or X's, I don't think your arguments hold up.
But even with "real" icons -- like, I've seen icons to show if a software package is for Windows or Mac or Linux. If a row is a TV show or a movie. If it's one file or an archive. If there's a PDF file download attached. An alarm icon for something past due. But these all seem totally fine and helpful. They're generally linked to a major feature of the platform that everybody understands, and help scannability.
Without clear examples of what you're arguing against, I'm frankly completely lost.
What if it's an icon with a simple shape? How does that compare to noising up the table with long phrases and repetitive words? Is the cognitive tax if icons a lot higher or just a little higher? What if it's an app where the user will be using it for hours, so they'll quickly learn what the icons mean and will appreciate the space they save?
Is a tick icon really that big a deal in place of "Task completed"? Or a pencil instead of "Edit"? Sometimes you don't have a choice because of lack of space too. There's always tradeoffs to make. Obviously try to avoid icons that are hard to guess though but sometimes that's not always possible.
I can't say I've ever felt tired looking at icons in a table, but when designing I have had the experience of replacing wordy repetitive text with some intuitive icons in a complex table and it suddenly looking less intimidating.
And I'm fine with a bit of cognitive exploration to figure out a green check and red X scheme rather than see a whole table column filled up with words like "active" and "inactive". The former allows more columns on screen at once. Horizontal scrolling is a worse impediment to assimilating information from a table.
Personally, my brain "page faults" whenever it has to interpret an emoji, which makes most use of in-line icons far worse than the text they represent. I expect few people have this problem, but I also expect that I'm not the only one with it.
Our users are context switching across dozens if not hundreds of digital experiences a day. Forcing memory recall is a tax. The question is always "whats the ROI?"
IMO color and words go just as far as an icon without relying on net new visual language.
As per your comment on horizontal scrolling, I couldn't agree more. Horizontal scrolling is booty. However, depending on the job to be done you can avoid overly wide tables with customizable columns, expandable rows, hover states, and strategic truncation.
I certainly would prefer those strategies over relying on a unique icon language that isn't part of the dozen or so immediately recognizable icon schemas already familiar to users.
Imagine for example a set of icons that are monochrome, open-ended glyphs comprised of a single stroke with line weight similar to that of the text. This could complicate visual parsing greatly due to high visual similarity to text.
On the other hand, a 16px checkbox control with subtle gradients, shadows, and depth cues looks absolutely nothing like text and is filtered out by the brain almost automatically (unless of course the checkbox state is pertinent to the user's intent). Same goes for a 16px colorful icon with shading like used to be ubiquitous in desktop operating systems.
Not sure how this is for accessibility in terms of colour selection, but I’m sure this could be fine-tuned.
The key problem is that most contemporary web design does not follow any idioms. Idioms are conventions of design that are universally understood. Skillful use of idioms makes it much easier to parse what is going on on a given page.
Where we are with most applications is that they try to define their own idioms, i.e. their own icons, their own navigation patterns, etc. But this is very arrogant because they're assuming that the user has the time to build that familiarity with all those idioms. This is never the case.
Every day I use web applications from nominally mature companies, and they have totally different icon sets for the same actions. This is immensely distracting and hard to read. Every company sees an opportunity to define their own icons, when what they should be doing is using the exact same ones as everyone else because that makes it easy to understand.
https://essays.johnloeber.com/p/4-bring-back-idiomatic-desig...
It is like security expert who tells you to use same password everywhere. “If you use TOTP with it, you will be quite safe”. Yes, but recommendation reveals lack of knowledge .
There are two different "modes" of looking at a table. One is just "superficially glancing" (like you look at a picture), the other is more like reading, because you look for some specific piece of information. When glancing at a table, the one with more colors and icons looks more pleasing, while plain text looks boring. But when reading the table to get information out of it, the "boring" one (if it's well done) is probably more efficient and a better experience.
And here is the problem: when presenting designs to stakeholders, 100% of stakeholders just glance at the table and always prefer the more colorful table with many icons — they don't realize that actual users, who need to get information from the table, would be better off with the "boring" table.
…What? I suppose, if the OP feels he can give you imperative commands, so can I. I'll spare you the pretentious text elaborating the command — it carries as much substance as no text at all, anyway. And, seriously, if you don't use icons in data tables, you should try and see how it feels. For me personally it's the magic trick that often makes all the difference between an unreadable wall of numbers and a vivid, crystal clear picture.
And wouldn't colors be just as helpful?
Anyway, they should design for the majority. Or use some configurable approach, like with CSS, to make everybody happy. Opinionated is so last year.
Another non-negligible advantage of icon is, that it is language agnostic. Not everyone is fluent in language they have to use.
I have never found text plus icon useful in a table that has colour.
> When we pack high-density information into a data table or a complex dashboard we are increasing the visual entropy of the entire system. Forcing the brain to decode intricate, non-universal shapes in a tiny 16-pixel footprint, creates a “cognitive tax” that users pay en masse every time they scan the table.
This is not how the brain processes icons. Icons have their meanings embedded through related patterns. If you rely on common icons such as x, checkmark, plus, shopping cart, arrows, three dots, etc. the user already associates certain meaning/behaviors with these icons (and will be surprised if they act differently in your application). Introducing new icons is more challenging, because you have to convey their meaning to the user. At no point does the user have to stare at an icon and try to "parse" it pixel-by-pixel.
In all of OP's provided examples, the icons are supported by both labels and colors, which removes all ambiguity. You can argue that the icons are pointless in this case, but they certainly do not increase any cognitive load. Is it harder to identify a tiger than a lion, because his fur is covered in stripes?
Stop using images that I can't zoom in to.
I imagine this could similarly be an issue with screen readers, but haven't tested it.
For actual data work, any sort of "rich formatting" is no bueno and icons are great for quick reflexive categorization for information-dense habitually used interfaces. They just take a slightly slower learning curve.
Nobody with a brain is reacting to this (me excluded).
Grow up.
And no, medium is not the last site in the universe to host at.
Oh, but it gets better:
> We have been sold a lie that icons are the universal language of the modern interface.
... placed directly below a button bar that contains only icons.