This idea stems from a blog post I read ages ago from GDS, and also a tweet from James Hupp, who said, “FAQs are a way to show you’ve thought about what your users should know but haven’t thought about your users.”
The general principle at work is that a service or user interface should be sufficiently well designed that everything the user needs to know is clear. This might mean clearer language, a different or more intelligent flow between steps in a process, smarter processing of input (e.g. forgiving someone for entering a country code in a telephone number if that’s not what’s wanted), helpful wording under text boxes, and much more.
If we really insist on eliminating FAQs then we may force ourselves to devise innovative ways to solve users’ problems. For example, if there is a process where a user fills in a form, and then it goes to some internal department for processing, it might seem reasonable to have an FAQ that explains what happens next. But what if we were to force ourselves to not have an FAQ? Then we might send the user an email thanking them for their submission and explaining what happens next. We might even send them occasional updates when their request passes through otherwise-hidden internal gates.
None of this is to say manuals, or help pages or explanations written in paragraphs are always wrong. But FAQs are too often dumping grounds for unresolved problems with the user experience. Eliminating FAQs is good thing to aim for, and if we force ourselves do this we can find innovative ways to help our users.