Last week we pushed out our latest addition to new.shropshire.gov.uk, the Parking pages, occasionally known as "those bl**dy parking pages" in the office.
Just to be clear, we've got plenty of love for the Parking service, and the site we've arrived at looks and works great, but getting to this point has forced us to re-evaluate the process we use to develop our sites a little bit, and in some cases get back to good ol' fashioned basics. The cursing should have more accurately been aimed at ourselves, and how we let ourselves get into the mess we got into because we forgot to properly manage this project and our workload.
The project.
The idea behind the parking pages is to make it as easy and intuitive as possible for the people who live, work and visit Shropshire to locate appropriate parking, and quickly find other parking-related stuff, such as blue badge information, parking permits and how to pay a fine. It replaced the existing parking pages on Shropshire.gov.uk, which did look a little bit long in the tooth.
The Parking project actually started over a year ago, as a "quick win" on what we refer to now as "new.v1" (I'll explain why later), but, because of other higher priority projects, including building the optimised "new.v2" version of the site in Umbraco 7 (see, I told you I'd explain why we call it that... it's much easier than saying "new new.shropshire", trust me), it was put to one side and never completed.
So, what did we have? Most of the content we needed was on new.v1, which in theory just had to be migrated to new.v2 and releasing after a quick content review. Sounds simple doesn't it, so what's the big deal?
The big deal.
Unfortunately much of the code which runs on new.v1 is not longer compatible with new.v2, due to the difference in technology between the two versions of Umbraco. Umbraco officiandos and CMS geeks will no doubt be aware that previous versions of Umbraco use Webforms with Razor or XSLT as their primary means of dynamic content retrieval, whereas 7 uses MVC with Razor (stay with me, non CMS geeks...), which means we had to re-write some templates and functions.
Not a show stopper in itself, but then we had other stuff which works in a slightly different way in new.v2. The navigation is entirely different, for example. We use different stylesheets and templates between the two versions too. So does that all make a difference to the user experience? Yes indeed. So what we were left with after a bunch of redevelopment was a string of pages which, in all honesty and despite our best efforts to replicate the old site, didn't really hang together coherently at all.
Back to the drawing board for Parking then, and back to square one to some extent, which meant asking the most basic question of our content:
"If I were one of our customers, what would I want to see?"
We have a range of UX design techniques which we employ, although it's a topic I'm not best qualified to talk about (speak to Sophie, our UX whizz) but current analytics, theoretical user profiles and a dash of common sense all come into play, and slowly but surely something more worthy of new.shropshire.gov.uk and the Project WIP brand began to emerge.
We do have to praise the patience of Zoe during this time, who, as our contact in Parking, had to juggle her day job with constant requests from us to review the work we had done and confirm that any changes to wording or layout wouldn't land the council in trouble.
So what have we learned?
There are obviously a number of lessons we can take forward from our experiences on this project - in no specific order (although the last one is the most important):
- We weren't quite as far forward with v2 as we thought we were - A large number of the hold ups with redeveloping Parking were due to the fact that we had started from scratch to build new.v2 in Umbraco 7, using MVC, and, as we started developing, we mistakenly took for granted that we had replicated all the functionality that existed in new.v1. The venue pages, for instance, use a bit of fancy geolocation stuff which we'd never fully developed or tested in new.v2, which meant we had to rebuild some of our APIs and things - all of which took up time that we hadn't planned for. The good news is that we are now a lot further forward than we were, and future developments will be faster. Hurrah!
- Process is important - Starting with the end-user experience in mind is something we're usually quite good at, but the drawn out nature of developing this one meant we got a bit lost along the way on this occasion. We have a well thought out workflow which starts with user experience, passes through graphic design before coming over for development, and it works, so we should stick to it.
- Content is king - it's an old web cliche but for good reason: if your content is poor (or badly organised), your end result is poor. We never really took the time to review and plan the content or the structure when we first started last year, and, as we had forgotten that we hadn't done it, assumed we'd meticulously tested and planned everything and just transferred over the content as it was. It was only when it was all in and we gave it a quick run through that we realised our mistake and had to pull out the post-its and flip chart paper to plan it all through properly.
So here ends the saga of our Parking pages. We're a bit older, greyer and hopefully a bit wiser for the experience, but what do you think? Check out the finished version at new.shropshire.gov.uk/parking and let us know!