This project is a website upgrade for a real-world nonprofit organization. Besides having real users, it differs from the class project nonprofits in that 1) it does not solicit donations, but 2) it provides a large number of informational resources on the web site that need organizing. So this site is suitable for the card sort task. Also, I'm really taking the lead on this and may end up doing an implementation myself - so these documents may be needed to explain to 'civilians' what I'm doing, rather than being technical input to an experienced cross-functional development team. A lot of my time on this project is spent chatting up AJ leaders (to get direction and buy-in) and members (to get user data.)
Competitive Review and Personas
Here's the draft competitive analysis (updated with labels for the numbers, and screen shots):
Even a site like WIND's would be a step up, for Appleton Jobseekers (AJ). But it's an open question right now whether AJ can handle content updates for weekly events, which would be necessary to get to the level of South Bay. AJ would have to switch their implementation to something like Wordpress (from hand-edited HTML (!)), and sort out the leadership team roles, skills, and personalities. I expect to improve the organization scheme and nav tables when I get to that in the design phase -- and I also need to look at some job resource sites as well as AJ's Yahoo page.
A primary persona using Personapp:
Recruiting brief for card sort:
0) card sorting is very sensitive to phrasing, and can definitely be shallow. For instance, the item "AJ Mission Statement" got categorized with bureaucratic functions, whereas if it had been called "Welcome to AJ" it might have come out closer to general meeting information. Instead of trying to be specific, it's important to try to be neutral. "what AJ does" or "purpose of AJ" might have been better choices. Another example would be a participant's category "Groups" which contained all items containing that word even though they weren't otherwise related.
1) so far, the most useful tools are the Participant-Centered Analysis, which is very clear, and the Similarity Matrix, which for this particular set of cards is also very clear at the top level, as well as being very useful for drilling down into the ambiguities. These seem like better starting points for a design than bottom-up crunching through the category standardization, though I may have another go at that later. We'll see what happens at later stages of the design.
2) One reason that results are unclear is that an item might reasonably belong in two different categories; this happens in the current sort and will show up in the final design as duplication of some information and multiple links in other cases. Another reason is multiple groups of participants with different perspectives that might need to be accounted for. Finally, sometimes there simply is no strong categorizing principle and it might be more appropriate to organize on an alphabetic basis or just do the best you can.
3) More participants might be needed in the case where you are trying to sort out multiple populations, for instance, or if you have loads of cards. For the current sort, it really would be useful to run additional sorts on some subparts of the project. Which is a bit different.
There is a rough site map over in the group Balsamiq account. I have a rough spreadsheet too, but am finding the visual site map easier to work with at this stage due to the ease of structural updates. It's just too soon to decide whether something is an info block, a page, or what.
I also have a homepage wireframe in Balsamiq. Next steps are to put more white space in the layout, and create the rest of those dropdown menus.
---- news flash ----
Presentation to management team went well!! I now have some designated project buddies & authorization to proceed toward development (with of course some reviews anticipated.)
---- news flash ----
Here is a new and improved homepage, with menus more fleshed out. Next steps are to iterate on the affinity diagram, site map, and spreadsheet, to consolidate the IA; then probably get some feedback from the organization and users before proceeding to prototyping.