A lot of blogs or other types of sites with a lot of content use tag clouds, where trending words are shown and popular words are displayed larger. For example

Should you put a tag cloud on your site? Does it have a purpose, other than linking to more content from your homepage so search engines will find it? Does a tag cloud really have something to add UX-wise?

  • 15
    I know this is my own experience but personally I've never even touched them. Jun 15, 2011 at 23:31
  • Must admit I have never been a fan. I find it hard to find situations where its relevant to use them (there is usually a better alternative as people have mentioned above). Personally as a user I have never clicked on one but thats just me.
    – Sheff
    Jun 17, 2011 at 12:57
  • Yeah, I agree. Not even once. In some testings we did, the users don't seem to even notice them. But that depends on how it's applied and on the audience, of course. Jun 18, 2011 at 0:05

7 Answers 7


Nielsen says: "Tag clouds were a huge fad in 2009, and have actually been a fad for several years. Even so, usability studies show that most normal users don't know what they are and don't know how to deal with them."

Although he doesn't link to any studies, I tend to believe him. Tag clouds are hard to understand and hard to process visually.

If it's a need for your page, why don't you just call it "Popular topics" (or something similar) and show the topics in a numbered list? It's transparent and much easier to understand.

Edit: After the discussion with JoJo here some nice examples of alternative design ideas, all of them from patterntap.com:

enter image description here

enter image description here

enter image description here

  • 1
    I could have written that comment but Nielsen says it all really. Jun 15, 2011 at 15:49
  • I've never touched a tag cloud despite seeing them everywhere. Jun 15, 2011 at 23:29
  • 2
    A bit off topic, but I am really eager to know the answer - what's that with Nielsen & Norman not linking to studies for a while now? Did they somehow upgraded themselves from solid, bona fide scientists to people who can just say stuff and it's believed true? I am really curious, because while I believe them on this particular one (tag clouds are real noise in the user interface), I was a bit surprised about their article on how gestural interfaces are bad for the users. Jun 16, 2011 at 6:32
  • @Mariusz: Good question. I attended their Usability Week Conference last year and in some workshops it wasn't really clear if their conclusions were based on real studies. But to be fair: It was the minority of workshops and they still were very interesting.
    – Phil
    Jun 16, 2011 at 18:34
  • I've never touched a tag cloud either, but the above images are excellent examples of much more useful & attractive ways of surfacing content - though they are also nothing at all to do with tagging, only popularity. Jun 19, 2011 at 2:37

I find tag clouds to be rather distracting.

I can say that I have never clicked on one. I understand that the larger the tags are, the more popular they are, but it doesn't seem to work nearly as well as the "Recent Tags" sidebar on this site, which displays a count next to the tag, and lists them in an easily scannable list. To me, tag clouds are just ugly clumps of words that are difficult to scan.


Some types of sites benefit from tag clouds. An example was an old version of www.torrentz.com. The tag cloud showed the most popular things being currently downloaded. With torrents, the popularity is proportional to download speed, so it's a win-win situation: you get good content really fast.

The reason why tag clouds don't work on other sites, such as blogs, may be a technical reason, not a user experience reason. Nearly all the blogs you visit are built on top stock content management systems (Blogspot, Wordpress). To be flexible and serve any type of blog, these CMSs merely do tag counts to determine the size of the tag in the cloud. The number of times a tag appears is really useless in determining its worth or popularity, so these tag clouds on blogs have really marred the regular user's perception of tag clouds in general.

  • 1
    Still, what makes a tag cloud better than a list with the most popular torrents?
    – Phil
    Jun 16, 2011 at 5:54
  • 1
    People can judge the popularity of the torrent by its size in the tag cloud. Without the tag cloud, people must compare the seed/peer count. Processing numbers is slower than processing size.
    – JoJo
    Jun 16, 2011 at 17:19
  • Yes, but people have to know that a bigger font means more seeds - and average users don't understand that. If you make a numbered list they don't have to compare the seed count or the text size (just the position in the list). There are a few good examples on pattertap.com: patterntap.com/tap/pattern/10938731234dd12c794ccbf patterntap.com/tap/pattern/17940777244b3264cc0315b patterntap.com/tap/pattern/10623223724b07266185ae5
    – Phil
    Jun 16, 2011 at 18:20
  • I agree with Phil, a standard list would be more easily usable than a word cloud for something like a popularity ranking. I think what this answer is missing is the (for better or worse) aesthetic factor of the word cloud that sets it apart from the rest.
    – DasBeasto
    Jun 21, 2016 at 13:12

I think that the problem with tag clouds is that they are so often done poorly and look quite ugly. A properly done tag cloud can be visually stimulating and informative. For example look at IBM's [Many Eyes word clouds]

Many Eyes word clouds - Obama Speech 2013

  • Shame you didn't grab a screen shot, the site is no longer available.
    – Tony
    Mar 25, 2015 at 23:43
  • 1
    How is this example visually stimulating and informative?
    – Ooker
    Jan 21, 2017 at 11:55

Tag clouds are one of those things that users either like or dislike. My recommendation is that if you decide to offer your users a cloud then also offer users (like me) a way to get rid of all that ugly noise.

Smashing Magazine did a good article a few years back on them. They spoke of how humans often think conceptually, and that's what a tag cloud quickly offers - an overview of trending concepts on the site.

There are a lot of articles out there describing why it's a dying trend. 1/3 of users in a survey from 2007 were identified as sometime-users of clouds - meaning 66% of users never touched them. I don't know if use has increased since then, but they look - ahem - a little old to me these days. (My 0.02c).


I don't have any hard data on tag clouds specifically. Although, like Matt mentions, I find the tag list with count on this site's search pages to be more useful.

But the problem remains that in a community-driven website, tags in general often scatter beyond their point of usefulness. In other words, when tags are generated by members, you'll often end up with two tags that are for all intents and purposes synonymous. But by searching the two synonymous tags you'll see very different results.

For this reason, it's often beneficial to also segment your site into tightly controlled categories. Obviously this will depend on the nature of your site. But if you design your categories to have a clear separation and no overlap, then that can be very useful. The navigation-by-category UI should then be above the tag list/tag cloud as it's likely to be used much more often.


My basic problem with tag clouds is that they appear to be a solution for a problem that doesn't exist - i.e. what are they actually supposed to be useful for?! I can't think of any time that I've ever cared about what generic subjects/topics/tags are popular on a site/service/app/etc... surfacing popular articles is a different kettle of fish altogether, those are specific, tags are abstract.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.