DonnaM » IA – content inventory

IA – content inventory

Why you shouldn’t start IA with a Content Inventory

Friday, May 12th, 2006

Leisa Reichelt has a blog post on Why you shouldn’t start IA with a Content Inventory

I totally disagree with this. But I work on really big, content rich sites, and they are usually in a *very* poor state, so that may differ from her experience.

I could not start a project without an inventory. I cannot imagine how I would reorganise a site if I didn’t know what’s there to be organised. A content survey is useful to get a quick feel for what’s there, but I know I’d miss some of the most interesting, most buried information. Without an inventory, I don’t really have an idea of whether I’m working with 1000 or 8000 pages and what the intricacies are within the inherent structure. It is impossible to do proper content analysis without detail to analyse.

She says that “It is also the best way to ensure that you’re *not* taking a fresh approach to how the content might be structured and related. When you’re doing a content inventory, you’re unavoidably indoctrinating yourself into the way that things are currently done”

That’s ridiculous. If that is happening, the person who is doing the IA work is just not good at what they are doing. A good IA can take the current structure, analyse it, determine what’s important, and can absolutely divorce themselves from the current structure. Even a poor site has some good chunks and it is important to know why someone has done what they have done. For me this is particularly important as I’m usually working in quite technical domains and never completely understand the domain at the beginning. I’ll just look stupid if I try to break apart something that absolutely belongs together.

Leisa’s alternative approach leads to pure top-down IA. And my experience is that pure top-down IA is just as bad as pure bottom-up IA. A blended approach is essential – where you look at ways of approaching the content, balanced by a detailed understanding of what you have to fit in.

I think the only time they are not useful is if you intend on ditching everything you have and starting from scratch.

Note: I wrote this initially as a comment, but it was so big I decided to blog it and trackback to it

Taking a content inventory

Saturday, January 28th, 2006

I’ve spent much of the past 2 weeks working on a content inventory. As a technique, it is pretty straightforward and well documented (if you’re not familiar, read: Doing a Content Inventory or any of my other posts in this category).

But I’ve done this many times and have some extra tips.

Why am I doing a content inventory?

You take a content inventory because, before redesigning a website or intranet, you need to know what you have. This is especially important if you will be migrating your content to a new structure or new CMS – at some point you need to know every single content element.

Even if you aren’t doing a big migration, an important step in a redesign project is gaining a deep understanding of the content – not only what you have, but what types of content you have and the relationships and patterns within the content.

Taking an inventory tells you what you have, and if you do it well, will also put you on the path where you have a good understanding of the content.

Can’t I get a tool to do this?

If you are using a content management system, you should be able to get it to produce a list of what you have. It may even give you some of the structure. If you don’t have a content management system, a tool like Xenu Link Sleuth can run through your site and tell you what you have.

Lou asked about tools a while back and there are some good responses on this post: Applications to Aid in Content Inventories?

What do I do with the listing

A listing of everything you have is great, but you will still need to do much work manually. Although it is incredibly tedious to look at every page and follow every link, it gives a great insight into the content.

I have done it both ways. As a consultant, I’d often get the client to produce the inventory and I would then look through it. As a contractor, I have done a manual inventory once, and a semi-manual (starting with a listing from the CMS or Xenu) every other time.

I swear, I never again want someone else to give me a finished inventory. The difference in my understanding of the content is amazing. When looking at someone else’s inventory, I get an understanding of the surface attributes of the content. With my own inventory, I gain a deep understanding of the content.

To design a truly effective content-rich site, I need this deep understanding. Having designed many sites, I can tell the difference in the quality of my outcomes. A good outcome depends on good inputs, and one of the best inputs is a good, manual inventory.

But I don’t have time to click through every page

You will regain the time easily. While you do the inventory, don’t do it mindlessly. Think about what you are finding, and make rough notes about interesting observations as you go.

After conducting a manual inventory, content analysis is much easier, as is identifying content relationships. You can produce a better high-level structure faster and it will be closer to the end result. You will also have a better understanding of the domain, which is incredibly helpful when discussing ideas with subject matter experts.

I have a big ugly spreadsheet – how do I read it?

One of the most important aspects of an inventory is that it shows relationships. If your site is a hierarchy, make sure the hierarchy is reflected in the sheet – look at the numbering system and indenting in the article listed above. This makes it much easier to understand what you have.


Doing a manual inventory hurts. On the first few days of any project, I end up with sore wrists and neck from sitting still too long and copying and pasting too much. Yes, I know I shouldn’t do this, but there is something compelling about watching the inventory build and I don’t stop as often as I should.

Some tips for reducing the risk:

  • Using 2 monitors allows you to look at the site on one and the spreadsheet on the other. It will save you masses of mouse-clicks swapping between the two
  • I have a split, ergonomic keyboard with browser back and forward buttons beneath the space bar. This too saves a lot of mouse clicks (if you don’t have one of these, and I do highly recommend them anyway, you can use the backspace key in most browsers instead of the back button)
  • Use standard keyboard shortcuts for copy & paste
  • Stop, stretch and walk around the block

OK, I’m done, now what?

In a redesign project, you will return to your inventory a lot.

Before starting to analyse the content (which I may write about in a few weeks), I highlight key pages – those that have a lot of inward links, are frequently used, are linked directly from key pages or are otherwise particularly important. I sometimes highlight topics or document types with colour to more easily spot them.

I later use the inventory to mark progress – either in the analysis, mapping to a new structure or migration.

Content Inventory – day 11

Sunday, August 25th, 2002

Yippee…I finally finished the big content inventory.

(I actually finished it on Tuesday but was away for a conference and have been writing an assignment since then).

I ended up listing around 5800 pages. This still isn’t a full listing of all pages – where there were consistent sets (eg meeting minutes), I only listed them once – but it is still a really useful list. All I have listed so far is link name and url. Now I have to start going through and identifying what type of documents exist. I also have to pick some of the documents (or small groups) to use in a card sort exercise to determine a structure for the new system.

Then later, of course, I also have to find out who owns every page and whether it is current or complete ROT (redundant, outdated, or trivial). That will be an interesting exercise. I imagine at least 50% ROT.

Content Inventory – when is enough

Wednesday, August 14th, 2002

Peter wrote an entry on his blog wondering whether a content inventory is worth the effort. I initially said yes, but it got me thinking.

How do we draw the line between listing out everything and getting a good overview. What would have happened if I were working on a 20,000 page site. At 500 pages per day, would it have been worth spending 40 days just to list out all of the pages? I suspect not.

Maybe this is complicated by the fact that I’m not only designing the new site (so need to know all of the content types), but implementing a new content management system. To do this well, I need to know what content we had at the beginning, and what we have done with it (moved, deleted, rewritten).

What do you think? Where is the line?

Content Inventory – day 9

Tuesday, August 13th, 2002

I continued the content inventory today. I finished a section of the site that has a structure resembling a pile of wet spaghetti. There were more than 200 links, but linking to around 30 individual documents. What a mess!

Now that I’ve finished that mess, I think I may be on the home run…

Content Inventory – day 8

Wednesday, August 7th, 2002

I’m still plugging away with my content inventory (4200 pages so far), and it really is driving me nuts.

I’m keeping going by telling myself how useful this damn list is going to be. Things I can use it for are:

  • identifying different content types
  • identifying who owns each piece of content
  • noting content that is out of date
  • assessing what will happen if I delete pages (as I have a list of all links and where they point to)
  • keeping track of content during the transition from old to new system

But the most important use of this list is to help us to make decisions based on some real content, not just vague ideas about what we have.

Content Inventory – day 5

Tuesday, July 30th, 2002

After a few days’ break, I started back at the content inventory. 2200 pages done, and I think I can see an end in sight (I’m going to take a guess that there will be around 3500 pages to map).

Some interesting stuff:
- the deepest page so far takes 9 clicks to get to
- the page with the most links has 130

The numbering system in Jeff Veen’s spreadsheet makes counting this very easily.

Content Inventory – day 4

Thursday, July 25th, 2002

1700 pages. Groan…

Content Inventory – day 3

Wednesday, July 24th, 2002

Day 3 – 1,267 page titles and urls recorded, and there is still no end in sight.

Tips for today
- learn as many keyboard shortcuts as possible
- using IE’s accessibility options, ignore colours so I can ignore where people have changed the visited and unvisited links

Getting boring…

Content Inventory – day 2

Tuesday, July 23rd, 2002

I’m on my second day of a content inventory. 641 page titles and URLs listed, and I still have no idea how much more work I have…

Content inventory tip – clear your browser history before you start and use your visited links to show whether you have already listed a page. Don’t list it twice, but cross-reference the primary location. This will give a clearer idea of the number of pages and extent of duplication.

Content inventory – day 1

Monday, July 22nd, 2002

Today I started to do a content inventory for a site I’m redesigning. I have no idea how long it is going to take me, or how big the site really is.

I’m using Jeff Veen’s spreadsheet, and it is pretty useful so far.

I procrastinated all morning (wrote a to do list, tidied my desk, read some blogs, chatted), then worked all afternoon logging page titles and URLs – 317 of them.

Work along with me over the next week and see how long it takes me to go crazy…

**I didn’t really clean my desk**

Doing a content inventory

Thursday, July 18th, 2002

I am about to start to do a mind-numbingly boring content inventory for a Intraweb redesign. It will probably take me all week and drive me insane, but it is absolutely essential to understanding the content in real depth.

I have been looking out for some references, and have found some good help.

From AdaptivePath: Doing a Content Inventory (Or, A Mind-Numbingly Detailed Odyssey Through Your Web Site)
BoxesAndArrows: Re-architecting from the bottom-up
WebTechniques: Taking A Content Inventory

I have sometimes wished that a tool could run through the Intraweb and generate the list, but that would defeat the purpose of me understanding the content well, wouldn’t it?

I may have some useful insights once my brain returns…