Occasionally I get emails asking about the (rapid) information architecture process I followed for FullCodePress last year. Steve Batey from Meld Consulting asks all the right questions and has agreed to share my answers with you!
[Steve] What were your thoughts going into the event about your role in the process? Were you concerned about how much you could do in such a short period of time, and did that affect how you approached the UxD process?
[Zef] I made it clear up front with the CodeBlacks what I expected my role would entail and what I would deliver. The team fortunately agreed. Some cross-over did exist between myself and the Peter Johnston (Writer) so we decided we should sit side-by-side and keep the lines of communication open.
My overall approach did not change much from usual as I’ve designed a UCD-process which can be scaled and adapted to fit almost any timeline (which is usually weeks to months). A condensed timeline simply means spending less-time refining certain deliverables (for example, the persona profiles I created were a light-weight version – we didn’t have the luxury to back these assumptions with field research).
Where I did cut corners was for what I call the ‘content model’- this would normally include in-depth content analysis, metadata and an ontology. All up I had about 4hrs to craft the User Experience deliverables, but had set myself a personal goal of 3hrs (one hour per deliverable). In hindsight I’ve got no idea how long I actually spent, but I think it was on target!
What methodology did you choose to use? What were the key deliverables you handed over to the rest of the team? What was the thinking behind those choices?
For my part I followed a variation on what is now probably a generic approach to user-centred design for websites – with one big exception – I didn’t have any contact with actual users!
I would usually conduct user research upfront to build comprehensive persona profiles. To validate the information architecture I would normally conduct usability testing on the paper prototypes.
What I ended up creating (with the team) is based on a persona hypothesis and a ‘common sense’ approach to the IA. We kicked-off the process by meeting with client (), learning about their organisation and then establishing their goals. This happened very rapidly in about 30min and my attempt at using sticky notes quickly degraded to unreadable scrawls on paper! We had learn all we could about this organisation fast and I think we reached a point after 20minutes where we had the gist of it, wrote up some goals and then moved onto the next stage- ‘user research’.
My usual process is to workshop with clients and create a set of draft personas on the spot, so this wasn’t much different for FullCodePress. We started brainstorming user roles and characteristics using sticky-notes and established the primary personas. I then sat down with Debbie from Grampians and using MS Word we crafted a first cut of the personas. We ended up with one primary and three secondary personas (Frank, Client; Pat- Caregiver; Kathy- Agency; Josie – GdAA Office Admin).
The first iteration of the persona set was released to the team (as print-outs). I later refined these when I had time.
I used the personas to establish tasks and primary pathways which are reflected in the sitemap. The sitemap (created in Visio) was done in collaboration with Peter who by now had a rough handle on the content. The first iteration was printed and presented to the team for feedback. At this stage labelling wasn’t finalised but it did establish a site framework which was essential for Mark Rickerby (Programmer) to build a site skeleton and linkages. As we went along all that changed to the sitemap was some swapping around of low-level content, following feedback from Debbie and once Peter began to rationalise and craft the content.
Next stage was the creation of what I call the Template Specifications (or ‘wireframes’). Created in Visio these were rapidly created by first establishing the master template (header, navigation, footer and layout grid). This base framework allowed Jeffrey Wegesin (HTML/CSS), Steve Dennis (Designer) and Ali Green (All-rounder) to really sink their teeth into the visual design and frontend coding.
The other templates I had predetermined would be the usual set of a homepage, section landing page and content pages- so all this meant was a quick analysis of the content types and ensuring a page template existed for each. In all we ended up with five templates (homepage, section landing page, content page, multiple content and custom form). The Template Specifications acted as the blueprint for the design, frontend code and development.
How much collaboration and involvement of other team members was there during the user experience design phase? How about the clients? (Did that help or hinder the process?)
The whole team viewed every deliverable (as a work in progress) and gave their feedback before it was released. A release involved me giving a hardcopy to everyone. If I created a new iteration then I gave them a new copy and took away the old version. I think we went through about three iterations of the paper prototypes during the course of the 24hrs.
As for the client, to me their involvement is always essential- I see them as part of the design team and fortunately Debbie was very articulate, open to ideas and had a really good grasp on who the site was targeting. This made my job easier and is a key reason why I think the site is a success- it meets the needs of the organisation (an outr
each to isolated and potentially vulnerable people) and for their customers (as a place they can go to get information and support).
One question we asked was- “How easily do you want users to be able to contact you?”. Debbie’s answer was- “We WANT to have personal contact with these people!”. So this guided the site to be not just information, but to provide pathways to key issues and allow users to contact Grampians via an online form, email, phone or visiting a local office.
How did you decide how much time you had to spend on each task? In hindsight, was this too much or not enough?
Prior to the event Thomas Scovell (Project Manager) had mapped out a 24hour timeline. I no longer have a copy of the plan but I think my initial push was about 4hrs all up. But this time doesn’t reflect the forward planning. I’m the type of person who has good days and bad days- I can rapidly achieve a lot in a very short time frame but then spend the next few days flailing about. Fortunately for FCP I did no flailing and just churned it out. If needed you could do basic personas, a rough sitemap and scrawly templates in under one hour on a whiteboard – but you get what you pay for! The less time you have the more likely things haven’t been thought through properly- time for reflection and refinement is important.
What was your favourite ‘guerilla’ IA tactic you were able to use during the competition?
I didn’t try anything fancy (except for the tagging-engine which we had to drop), and I used templates the team would be familiar with. The deliverables I created for FullCodePress are a base-line version of the UCD deliverables we create here every day at Provoke.
How did the rest of the team react to your role? Did they know what you were doing there?
They were fine and fortunately most of the team had already worked with a similar process and deliverables.
How closely did the final result match your ‘vision’? Were there technical constraints uncovered further into the process that forced you to revise your design? Could this have been avoided?
Looking back the site is about 90% accurate to the design- we had to drop some functionality at the last minute due to time constraints. This included ‘related content’ (created by listing content tagged with the same topic) and a site map.
During the early stages of the competition the spotlight was well and truly on the UxD member of each team: how did you feel about that?
Absolutely fine. We agreed that Thomas would take care of the business side (business analysis and business goals), while I would focus on the users.
Last question: did you have fun? Does it feel like it was worthwhile to you to be involved?
I had a great time and encourage others to give it a go – a great learning experience for all levels of experience. The WIPA and Webstock crews looked after us- good food, accommodation and great company. The fun factors surrounded the actual event with some delightful dinners, exploring the sites and killing ourselves trying to find a decent coffee in Sydney (I admit, there’s a vast improvement on 9 years ago when I lived there!).