Intranet Realities

I work with a content management team that supports a couple of Web sites.  This includes a corporate intranet, which has aged pretty poorly overtime.  Like many, it has governance issues and has been largely used as a dumping ground for documents.  The front page became a link directory because the search and navigation was so poor.

The developers had the leeway to update the look and navigation of the site and so they did so.  I was anticipating that it was really just putting a new layer of lipstick on the old pig.  Our internal usability testing was far more positive than that, though.  Some of the feedback that we received indicated that it was attractive in ways the former site hadn’t been, that it was now no longer embarrassing to have it as the default start page, that it seemed easier to use.

Wrong Assumptions

The lead developer got through a significant amount of wireframing and development before we sought outside input.  In part it was because you can tell when a site has been developed by a committee.  It was also because we realized we were not going to fix what was inherently wrong with the site.  We were just trying to make it less awful.

The approach had been to keep some of the information widgets – local weather, local transit – on the home page because they had been there before.  Some of the feedback was that these were no longer necessary.  “Everyone” checks these things on their phones anyway, so why would they be on the intranet?

We took this input and the developer dropped these widgets off.  The long term plan is to make the site customizable, so those people who want a transit or weather widget can add it back.  So removing them at this stage seemed like a minor thing.

I’m not sure this was, in the end, the right move.  One of the other fixes we made was to improve the search experience on the intranet.  I’ll explain more below, but it mean that I was watching the search logs more frequently.  There were plenty of searches now for weather or train or buses that it seems fair to assume are because that information is no longer on the front page.

Search Words

The intranet refresh has been a positive experience in almost all cases.  The staff who we have heard from are pleased not just with the look but the choices made in where content is.  For example, the most sought after document was the daily cafeteria menu.  That is now on the home page.

But we also saw searches for content we hadn’t expected would be as heavily used.  This was content that was also not on the home page previously but still had a following.  It highlighted one of the other tweaks we had to make behind the scenes.

Our intranet search is powered by an ancient Google Mini search appliance.  It’s like having a bazooka in a water gun fight and was indexing about 3,000 pages under the 100,000 page license.  I had to hack a lot on the underlying XSLT to make it look more modern – we made the search results style reflect Google’s current results – and to make minor tweaks to the language.

One thing we hadn’t used much in the past were the Keymatch and Related Queries functions.  These enable a “did you mean” quality to search results.  I’ve been monitoring the searches each day to see what sorts of things we could apply Keymatch too.  This was useful particularly when the search was for a resource that didn’t exist – had never existed – within the content management system.  For example, furniture moves and help desk requests are made using forms.  But these forms, though Web forms, live on application servers elsewhere within the organization.

The Keymatch helped us make that connection, because it offers up a URL and it can be within the intranet or external.  Again, I had made some assumptions about what this would be used for and created some matches in anticipation of the keywords that would be used.  Many of these forms had been links on the home page and had been shifted off it.

What was just as interesting were the other search terms staff used when looking for things.  It highlighted the gap between the official corporate language – performance management or credit card requisition form or health and dental benefits – and what the staff thought of:  annual review and expense reimbursement and [insurance company name].  I’m not sure how these things were found before.  Is our new layout and navigation hiding things that were findable before?  Or have these always been hidden since these terms don’t return the relevant document?

We’ve ended up adding a lot more matches to ensure staff get to where they are wanting to go.  In some cases, it means returning more than one match since some teams use the same document name but mean something quite different.  In some cases, they mean a resource or database that is not typical content at all.

More Content

We have too much content.  It’s a dumping ground and I know we’re not different from many, perhaps most, intranets.  But I ended up having the search engine crawl our public Web site as well.  This seems counter intuitive – we’ve got too much content there too, since it suffers the same issues as our internal sites – but it was clear that the intranet and, particularly, the search were being used for things that didn’t exist on the intranet.

Many of the searches are looking for staff names.  This makes complete sense and we are now investigating how to hook our search up to the LDAP directory for staff.  There is a database driven staff directory but it’s not crawlable for search.  If staff miss the link on the front page, they won’t find staff.

Other searches are looking for information that is easy to find on the public-facing Web site but isn’t duplicated on the intranet.  A search will now return results from both the intranet and the Web site, with the former results weighted more heavily.  This has enabled staff to find resources without knowing exactly where they were.  There’s no knowing if the results are more relevant.  I am interested to see how the search results change over time.

It’s been an interesting experience.  A project that I thought was merely a superficial one has had a deeper impact.  It hasn’t fixed any of the underlying problems we have with the site but it has rebooted interest in the site and appears to have engaged staff.

David Whelan

I improve information access and lead information teams. My books on finding information and managing it and practicing law using cloud computing reflect my interest in information management, technology, law practice, and legal research. I've been a library director in Canada and the US, as well as directing the American Bar Association's Legal Technology Resource Center. I speak and write frequently on information, technology, law library, and law practice issues.