It's been more years than I can believe since I last wrote about a Shropshire Council website migration for Project WIP.
We've been planning the brand new shropshire.gov.uk (now known as next.shropshire.gov.uk) for some time, but the move from Umbraco 7 onto the considerably newer and shinier technology of Umbraco 11 has given us the opportunity to completely rebuild the council website on a brand new framework which we could develop from the ground up - and you just can't rush that sort of thing.
We began our latest migration at the beginning of 2023, and in writing this sentence I've just realised that it's May already (and May is traditionally a whacking great five months into the standard 12-month structure of a year). I raise this because we're nowhere near being almost halfway through the process and neatly getting it all done within the year. If it was about 2am on 1 January we'd be just about on schedule given what we've done so far.
There are a multitude of good reasons for this tardiness: the day job, the size of the existing website and the need to properly review everything, the fact that Shropshire Council is undergoing huge cultural and structural change towards which we're contributing substantially...I could go on but it'll start to sound like excuses.
Thinking positively, we (the Shropshire Council Digital Services content sub-team) are now familiar with the backend of Umbraco 11, which, in using a content block-based approach was a radical departure for us, but one that we're very much warming to. This familiarity is speeding things up, and our confidence grows with every service that we move across. It helps hugely that we work so well with colleagues in the web development wing of our team, with their patience and ability to interpret our garbled requirements an invaluable aspect of this gargantuan project.
Generally speaking, Umbraco 11 has given us a nice balance between remaining fully accessible and responsive (up to 80% of our audience now use phones to access our sites) and allowing us more flexibility in terms of design and presentational options. It allows us to present the 'drier' stuff as it needs to be presented (I won't name specific service areas here as no-one wants their content described as 'dry', do they?) but also gives us the scope to produce draft designs that we can 'sell' more convincingly to services that require something fluffier for promotional or commercial needs.
Because of the size of the task of this migration it's an incremental approach, and that means that we're essentially running two sites concurrently. That adds complication in terms of remembering to ensure that where a page has been done on the new site but is not yet live, it's updated at the same time as its equivalent live page on the current site. A spreadsheet covering the project helps us to stay on top of that, as does minimising the time a new section remains in a beta phase.
Formulating a fail-safe redirection policy has also been a challenge. When a section on the new site has been made live, we've settled on redirecting all child pages on what has now become the old section to their homepage, and redirected that homepage to the new equivalent homepage. This should hopefully be sufficient to avoid too many broken link experiences or navigational confusion. We're also currently thinking about SEO optimisation - how to ensure that the new pages are prioritised in a search, and the old ones disappear.
The principles of how we approach the actual writing and editing of our pages remains the same as how I've described those principles here before ('How we write' and 'A capital idea') but over the course of time and multiple edits, often made in a hurry, those principles can get compromised with deleterious effect on the quality of the product we're putting out. This migration gives us a chance to completely overhaul that side of things, with one of the processes involved being to check the written content of each page, as well as physically moving the page (if it warrants coming across). For the most part, our checks usually involve ensuring that a page doesn't have any 'New for 2007!' flyers attached to it, and if practically possible converting any attached pdfs that are still valid into actual pages to make things easier for the reader, and to push up our accessibility credibility at the same time. That process is a whole other blog post on its own!
The 'if it warrants coming across' aspect is a key part of the process, because one aim is to reduce the number of pages we host as much as possible. We know there are pages we can ditch, but without going through the site page-by-page we don't know whether that's scores of pages, hundreds, or even a quarter of the site (we've got 3500 pages in total on our main corporate website). This migration by its nature is precisely that page-by-page process, so it allows us to do just that granular review.
Something we're determined to keep at the forefront of our minds is to write primarily for Shropshire residents. We'll cater for councillors, the media, and internal council audiences of course, but it is residents we'll be addressing first and foremost when writing, editing, migrating and culling pages. To this end, that culling and streamlining is vital, because the more compact and efficiently presented the website the easier it'll be for our residents to navigate. They tend to be on our site to either find something out quickly and easily, or to carry out a transaction, again with the minimum of fuss.
We could get offended that the average time people spend on any of our pages tends to be incredibly fleeting, but to meanderingly browse a local authority website for pleasure is very much a niche pursuit! We get that, which is why we place big green buttons strategically throughout the site to allow people to do what they need to do as quickly as possible by clicking on those buttons. Once they’ve spent a few seconds doing that, they can then go back to shopping sites to spend hours looking for a perfect pair of sandals for their holidays.
On that note, I'd better get back to it. I assume the next one of these posts about migration, several years down the line, will be done by AI rather than me - ah well, it was fun while it lasted.