We are often asked to put content in Frequently Asked Question format (FAQs). They’re a popular convention on the web, but we don’t recommend them and here’s why:
They’re too slow
FAQs are convenient for writers - they put everything in a long list; it’s all neatly organised and the ‘Q’ does a lot of work for you. But they’re more work for readers - questions take longer to scan and understand than simple headings and you can’t take any meaning from them in a quick glance.
You could read the headings of this article and work out, basically, what we think about FAQs. You couldn’t do that if this was an FAQ.
We use 'frontloading' - this is when you put the term most people are looking for at the beginning of the sentence or paragraph. If you write all subheadings with how/what/when/why (and you have to if it is a question) you can’t frontload. This means users can’t scan the words as quickly, and they can’t understand as quickly. You may not save minutes for users but you will be saving them some time.
They lead to duplication
Usually, I see FAQs duplicating content. People tell me:
“users want them in this format; they can find information faster”
If that’s true, it probably means you need to structure your existing content differently.
If a question is frequently asked, it means you need that content on your site. Structure that content clearly so you won’t need another page repeating the same information in a different way.
That problem really shows in search, where you will end up with duplicate results competing for attention. You are fighting with your own content. That can’t be efficient for you or for users.
This is actually a problem GOV.UK is experiencing. The content in our support pages is now appearing in search, so we’re stripping away all of the support content we don’t need and making it easier for users to get straight to the things they’re looking for.
They’re tonally wrong
On GOV.UK, our remit is to get the information to the whole of the UK. We have to write for everyone. The best way to do that, is to write simply and clearly and remove all duplication and superfluous text.
Twitter agrees
I saw this on Twitter from James Hupp and it perfectly shows our position:
18 comments
Comment by Robin Carswell posted on
If you have frequently asked questions, it's because you haven't answered them in your content.
All questions should be infrequent.
Comment by Stuart Barker posted on
I think FAQ is a thing of a bygone era. They served a purpose in the dawn of the internet but are an outdate concept. To be fair don't people just Google these days if they have an actual question? Quick, simpler.
Comment by Web Governance (@diffily) posted on
I did a fair bit of work on FAQs as part of an "online help" system a few years ago.
The Conclusion ... **some** user needs can be served well by FAQs.
The Focus ... the best are those that are [1] True/False, [2] Descriptive & [3] Explanatory. (See an info graph at http://blog.diffily.com/2011/06/frustrated-housewives-seek-satisfaction.html)
The Rub ... high turnover of content is needed, i.e. kill old FAQs as they drop in traffic.
Comment by Rebecca Buckingham posted on
Spot on! I have never come across FAQs that couldn't be turned into more targetted, useful and reader-centric content, and I've been editing content for a long time (way too long). FAQs go against all the principles of effective web content and are a last chance saloon - visited when people can't find the content they're looking for.
As the great man Gerry McGovern says, 'We should never classify based on the content type (FAQs) or the tool. We should instead classify based on the task the customer wishes to complete'.
Content is content and websites are, in themselves, there to answer questions anyway.
Comment by Geoff Hunt posted on
Customers already know the question, a web page should answer their question as quickly as possibly - NOT make them re-read their question. (or 10 other questions before getting to their answer)
Example: which would you rather read on the page?
FAQ:
Q: Where are the toilets located?
A: They are located on the main floor.
Non-FAQ:
The toilets are located on the main floor.
Comment by Michael Eriksson posted on
Looking at this reply, the reasoning in the original post, and the many nonsense uses of FAQs (cf. also my previous comment), I would see a problem with too many people simply being aware of the origin and original main use of FAQs.
These arose on mailing lists intended to answer questions on a certain topic, thereby saving time and resources and preventing every second ``newbie'' from asking the same questions over and over again.
This should also not be seen as a sign of lacking documentation in other regards (when speaking of e.g. a particular software): The problem, instead, was often that the documentation was too extensive for every little detail to be found fast, that the answer was in the documentation but would require some amount of experience or thinking to arrive at that answer, that some people, as a matter of course, ignore the documentation and jump straight to asking, etc.
Alternatively, on list which had no natural documentation (e.g. through being of the discuss-a-topic kind), a FAQ could simply summarize the answers arrived at during several earlier discussions started by the same question.
Problems arise, for instance, when some naive people insist on putting the lion's part of the ``documentation'' in a question and answers format, while ignoring the fact that the questions chosen are not asked frequently or at all---and that the sum of the documentation could be written down on two easily scannable A4 pages. (To be compared with the often hundreds of pages to which an original FAQ was a complement.) FAQs are simply not intended to be a ``neat'' or ``pedagogical'' format, as these naive people seem to believe, but to make it easier to find the answer to specific frequently asked questions in a sea of information, where the search for the answer could otherwise take hours.
Comment by Michael Eriksson posted on
``too many people simply being aware''
Here I intended to have a negation, e.g. ``too many people simply not being aware''. Sorry for any confusion.
Comment by Julian Staddon (@JulianStaddon) posted on
I think FAQs can be a good tool. A few people mention that the questions need to come from the user/customer and that is exactly right. We use them to support software launches and allow users to add a question to a document that alerts experts/super users to then go and answer.
Agree that reading questions can take time and agree that you need to frame them. Sometimes a simple initial heading is useful so that the reader can zone in on what they need (e.g. Excel | How do I...; Workbook | How can I rename...). Then as mentioned you need to use the FAQ to rework your content.
Comment by Alison Jack (@sadfangirl) posted on
If you need FAQs your content is wrong.
Anytime anyone comes to me asking for an FAQ section I've explained to them that if the same questions keep being asked they need to review their existing content and fix the problem, not add another section.
My favorite instance of FAQs is for services that are new and the project team "anticipates" the questions that might be asked. I wish I could read minds like that.
I don't even stay on a site with FAQs now, because I know it's not been designed for me as a user.
Comment by Michael Eriksson posted on
I have been on the Internet since 1994 and have had more than my share of dealing with FAQs. None of what you say above forms a strong argument against FAQs, some of it may even be seen as directly wrong.
However, there are many problems with FAQs that result from using them incorrectly. Most notably, with some potential overlap with one of your early points, it is a deadly sin to simply write the information that the author wants to bring over in the form of questions and answers: FAQs should base on Questions that actually have been Asked Frequently---the very point of them is that a reader with a frequent question should be able to get to the answer of that question fast, without reading a long ``classical'' documentation, and without having to ask a physical person. In this manner, they form a very valuable complement to (not replacement of!) other documentation.
Comment by Jim Balea posted on
Not to point out the irony, but this post itself IS a FAQ... (it even begins with "We are often asked..")
Comment by Rick Yagodich posted on
Which would make it the only FAQ in living memory based on questions that are really asked…
Comment by Sue Waller posted on
Rick I have to disagree this is not the first or only FAQ that is based on a question that is really asked, we do it all the time. KPS provides a tool for http://www.canterbury.ac.uk/faq/Home.aspx and their FAQs dynamically change throughout the year (go to their site each month and you will see that their FAQ changes) in line with the questions asked by their students. If you have a good natural language search you have a much more powerful way of allowing your customers to ask questions and actually provide the answers from the information that you have in existance, in many cases without the need to repurpose the information.
Comment by Tim Blackwell posted on
hmmpfffs at https://www.gov.uk/support/about-govuk
"What is GOV.UK?
Where can I find GOV.UK?
Who built GOV.UK?
What information can I find on GOV.UK?
Have you got printed brochures with the information on your site?
Is the information on this site up to date?
Has the website been tested?
What does ‘detailed guidance’ mean?"
...
...
though to be fair, the above faq doesn't answer the question I most frequently ask myself about .GOV.UK ie "how is search *meant* to work and why do I find it so alien?".
Comment by Roo Reynolds posted on
Thanks Tim. We've recently improved the support pages and no longer use the FAQ format there.
Comment by John Dineen (@john_dineen) posted on
Hi Sarah - great post. Thanks for highlighting an area that we've been looking at for a long time. The issue with FAQ's is not whether they add value or not - I think that they do - but how they have traditionally been constructed. In my view, your post focuses almost exclusively on the symptoms of poorly implemented FAQ systems rather than any inherent reasons why structuring content in a QA, QA format is a bad idea.
Let me explain...
The two biggest problems with FAQ's is that (typically):
1. The FAQ's do not live in context. They normally get their own page - the "FAQ section" that tries to answer too many questions across too many areas. This separation between the FAQ's and the content that they relate to, delivers a poor experience for the user.
2. Who defines the FAQ's? Too frequently, users have limited input into defining the FAQ's. We've all seen examples where FAQ pages are really just a marketing exercise. Without the user defining the FAQ, can you really deliver the answers that they require?
In my view, a well thought out FAQ system includes the following:
1. Micro FAQ's, distributed across your site. Each page gets its very own FAQ that is continually evolving.
2. Community driven. Make it easy for your users to ask the questions that you need to answer, reply to those that you need to enhance and vote up others to FAQ status.
3. Support multiple channels (i.e questions that come in via Twitter /or Facebook get included in your FAQ) so that they can be recycled to subsequent visitors. Distribute the questions to those best placed to answer.
4. To avoid reading long lists of Q&A's, support FAQ search / prompt relevant answers as the user types. Support multiple media types (text, video, audio, interactive polls, etc.)
5. Encourage good question structure by limiting character input, tagging, etc.
Writing simply and clearly is best practice for all sections - FAQ's included. So the issue then is not with FAQ's as a format but really with how people have misused them.
Comment by Brian E Smith (@briesmith) posted on
FAQs are an effective way of communicating information and they are increasingly "understood" by consumers. The difficulty in using an FAQ effectively comes about when you can't frame your search query so that it matches a previously answered question in which case you get a 0 matches response which is frustrating or your question hasn't been asked previously..
An answer to this is to structure an FAQ page so the original questions are collected under a framework that makes sense and which matches whatever it is the FAQ is about. Such a framework would also identify those areas of product usage, or whatever, that haven't been queried yet allowing the FAQ maintenance team to create suitable responses in anticipation of later queries.
This means more work for the FAQ maintenance team but a better outcome for the FAQ user.
Comment by Stefan posted on
"They are convenient for writers" is spot on, but for the additional reason that it gets them off the hook of really understanding user needs (so I think it's generally worse than James Hupp's tweet suggests). Or as I put it in a post a few years ago, Easily Answered is not the same as Frequently Asked.
For a contrary view, Nick Halliday has a very good recent post which is almost persuasive.