Examples of Help Systems that Provide Users with Multiple Entry Points?
May 13th, 2011 | Posted in Uncategorized 11 Comments »
In my Organizing Content series, I’ve been exploring the idea of adding metadata to help topics so you can sort them into different arrangements for different audiences. For example, you could add metadata tags such as “popular” or a specific role or a business goal and then provide entry points that arrange topics based on that metadata.
Since I’m presenting on this topic at the Summit, it would be nice to have a few examples of help systems that actually implement this idea. 
Despite my asking around, I can find none. I have a small system I’m creating for one of my projects, but it would be nice to point to a larger trend. Anyone have any examples of help systems with different entry points?
Do you know of any examples of a help system that provides users with many different entry points into the content?
Sponsors
Tags: Information Architecture
If you liked this post, keep updated with new content: Subscribe to I'd Rather Be Writing.
Both comments and pings are currently closed.
3841 Subscribers
Mark Baker (4)
Jen (3)
Richard Hamilton (3)
DiSc (2)
Jenny (2)
Sarah O'Keefe (2)
Aneesha (1)
Barrie Byron (1)
Catherine Crane (1)
D. Renn (1)
I think it depends on how you define entry points. I remember working on a help system where the architecture called for help at the field level, a unifying page level, and then a task flow. Those could also be assembled into a larger content flow that includes concepts and reference material. Each level of help could point to related content, if needed. The idea was to take the modular information and assemble it for as many needed contexts as we could anticipate.
It made content development interesting and challenging for the writers, but considering the complexity of the product, was worth the extra effort.
At my company we have a simple version of what you’re talking about. We have a User Manual and an Admin Manual. When we moved from DITA to MadCap Flare, we decided it would be much simpler for the webhelp embedded in our flagship product to open a single help system with two entry points, one for Users and one for Admins.
Both menu options open the same help, just at different start pages. Whichever one you open is the one that’s expanded in the TOC, while the other is collapsed (but still available and searchable).
So far it has worked out well for us. We just have to build one help file instead of two separate ones.
We’ve been using role and “business sector” tagging for a while. So far we have tried creating landing pages for each role/sector by pulling content dynamically based on metatags and there are plans to incorporate this in to the sites search engine as well.
Our solution is custom built because we couldn’t find a help/KB system with this functionality out of the box. Unfortunately the system is under lock and key (or rather password) so I can’t provide any functioning links. I came close to an out of the box solution while playing around with Joomla and it categorization and tagging tools, you might want to have a look at that.
We are currently looking how we can improve on our current system while moving to SharePoint 2010.
Another site that might be worth looking at, if you have access, is the developer network at SAP. They also have a fairly complex Role based layout and system.
Great question, Tom. I’ve been thinking around this point too.
The problem I ran into is that it’s not possible to sort the pages returned by the metadata search. For example, it’s easy enough to add tags or labels to various pages, and then to create another page that shows a metadata search returning those tagged pages. But in most CMSes or wikis, the search returns the data in alphabetical order, or by date last updated, or by popularity. Typically, we want it sorted by an arbitrary order such as “Introduction”, “Getting Started”, “Adding your Content”, and so on.
For that, you’d need the CMS to understand your need and allow you to define the order of results in some way. I haven’t seen a platform that can do that yet.
Like you, I’d be interested to see a solution where this is working.
Cheers, Sarah
Thanks Sarah. You described the exact problem I encountered with this method — lists in seemingly random orders. I put together a sample here.
The best semantically structured wiki I’ve come across is the Semantic Enterprise Wiki. One of their metadata attributes is position. However, if you have 5 different arrangements, it seems like the position in one arrangement would differ from the position in another. I’m still trying to figure out the best strategy for that. Maybe a custom content curation would be the only solution? You could create a position for each arrangement, but that seems more tedious that just manually doing it.
Thanks for the links, Tom! Having position as a metadata attribute is a good start. As you mention, there may be a problem when you’re trying to put the same chunk into two different arrangements.
This brings to mind a similar sort of thing in software development, when developers create “servlet filters” for an application. I had a hand in writing the docs for a particular use case, so this sprang to mind.
Developers can specify the location of the filter in the application’s filter chain. They can also specify a weighting, which is then used to decide which order to place the filter in the chain for filters which the same ‘location’ attribute. Example: http://confluence.atlassian.com/display/PLUGINFRAMEWORK/Servlet+Filter+Plugin+Module
Something like that may help with the problem of sorting our information chunks.
I wonder if the Semantic Enterprise Wiki allows the two levels of position/weight metadata too. Worth investigating, when I have time. Let me know if you find out anything more.
Cool discussion.
Cheers, Sarah
Yes, Semantic Mediawiki does allow multiple levels of weighting. For example, I can query all topics that are popular, and then sort by role as well as other values. However, there’s something wrong with my installation of the extension, so I couldn’t demo this feature unfortunately. Still trying to troubleshoot that one.
Here is one example
http://publib.boulder.ibm.com/infocenter/wpdoc/v6r1/index.jsp?topic=/com.ibm.wp.ent.doc_v615/install/installingwp.html
Look at examples on
– setting up a single server
– setting up a standalone …
Some examples of similar topics with slight variations. Using keyword search returns filtered results.
Fromme, thanks for the example. Filtering by country and language is a good start down this path.
Tom,
I have been thinking about something alone these very lines, but in a completely different – and I believe overarching – scenario. Mine was not to do with a help system, but a revision of navigation models as used on the web in general. The start point for consideration was architectural requirements for an in-house knowledge sharing system (your average intranet being a cumbersome middle-management ego-fest).
As you suggest, it is a system that has search as one of its primary interfaces, and the system then needs to be able to bring the resultant elements together in a logical, contextually structured way. No easy task.
As Sarah mentioned, it is the CMS that can perform this task that is lacking. While I do not claim to have the entire answer, I believe one of the fundamental principles is that the system will need to analyse new data as it comes into the repository, locating it within multiple thematic ordered hierarchies (with author override each).
Rick, thanks for joining in this discussion. A robust CMS can help you apply metadata to topics and manage how the topics appear based on the weighting of the metadata. I wish I had more experience working with content management systems, though. I’m not sure the average technical writer has access to configure major systems like this.