I've just finished the book.
It's generally seen as the authoritative text on what information architecture is - it's been around since the 90's and this is the most recent edition released in October, and while it's still fresh in my mind, I figured it's worth sharing some notes
### Who is it good for?
If you have any interest in the tools, the techniques and the deliverables that existing firms use to "do IA" for larger corporate clients, you'll come away learning lots here. There are lots of good examples of different variations of sitemaps, content models, controlled vocabularies, wireframes, and the processes you might follow are also outlined in detail - in fact, this book ends up feeling like the closest thing to an authoritative, 'this is how the industry does it' reference for both information architecture, but also for elements of UX and content strategy.
That said, while it definitely feels like a secondary audience, there is enough coverage about the general principles going into this for you to adapt these to more cash strapped environments, like scrappy startups or small NGOs.
I've don't have background in information sciences, and I have never worked as a librarian. As of December 2015, the categories and tags I use in this blog, and other sites or services I manage elsewhere are so poorly thought through that I've largely given on relying on any taxonomy or tagging system, and rely almost exclusively on search now.
So, what the book had to say about the value of taking a more considered approach to tagging, categorising and otherwise marking up content with extra information was fascinating.
There's a great section about the application of controlled vocabularies (i.e. sets of specific terms, with explicitly defined relationships between each other and so on) to inform the design of their Disneyworld parks, and some good examples of clever use of metadata on content within the BBC to enrich search features with 'Editors picks' and so on.
In addition, there are some good suggestions about how to go about structuring information with controlled vocabularies, so you're not stuck with the IA equivalent of a blank page staring back at you. For the first time, I think I'm prepared to do more than make up a taxonomy as I go along when writing, and it's definitely changed how I think about content professionally.
I also came across something related to this I had no idea existed, but seems obvious in retrospect. Most sectors have their own sets of language or jargon, and you might come across the odd glossary, but these also exist in more structured downloadable forms, that work as good starting points when applying some IA principles - you can find examples of these at sites, like the Taxonomy Warehouse, or the maintained by groups like the American Society of Indexing.
### Sitemaps and the roles they play
There's also some interesting material about the roles that sitemaps play - previously, I've put together sitemaps largely as hierarchic documents of pages, but I hadn't found them too useful beyond relatively static sites, that don't have much in the way of interesting interactions.
The sitemap chapters in the book helped me see sitemaps as something more than just hierarchic diagrams, but more like schematics for interactions, for conversion funnels, or useful guidelines to inform wireframes, visual designs or just going straight to HTML.
If you haven't found sitemaps useful before when working on sites before, I recommend in the very least looking over that chapter.
This is turning into a pretty long post, so I'll try to round up with a grab bag of nice ideas in the book.
There's a nice passage in the book about how important it is to try to avoid dead-ends for a user when they're using a site you're building. For example, if someone is searching for content, if no results can be returned by a search query they use, there are various steps you can take to salvage that user journey:
- you can provide hints on how to refine search queries for common issues (assuming you're not already solving these automatically)
- you can have some helpful links to where they can get help or find related content
- you can trigger events for support tools like Intercom to in the very least try to capture their intention, to help inform your content plans for the site
Another idea was the navigation stress test, which is a good check to see if your IA actually works. Here's how you do see if your site can pass:
Ignore the home page and jump directly into the middle of the site.
For each random page, can you figure out where you are in relation to the rest of the site? What major section are you in? What is the parent page?
Can you tell where the page will lead you next? Are the links descriptive enough to give you a clue what each is about? Are the links different enough to help you choose one over another, depending on what you want to do?
You can tell this book as been revised a fair few times - you can tell from the illustrations and screenshots used in a number of places. While responsive design and Omnigraffle are mentioned, you see some extremely ancient looking screenshots of wireframes and websites, that look like they date back to the early 2000's.
At first I though this was worrying - isn't there a risk much of this is out of date? But after reading the accompanying text, seeing such old looking screenshots ends up being quite comforting, as you realise the principles they are illustrating are a) grounded through years of use, and in many cases b) universal beyond just the web.
If you're looking for an authoritative reference on the subject of Information Architecture, and you want to understand how this relates to to fields like UX or Product management (I found a lot of overlap between both of these disciplines), I'd recommend picking it up - there's a good reason it's seen as the standard book for the discipline, and there's enough here outside of IA to help your thinking about the web.