I learned quite a lot this week, and only a small amount of it about how to use a CMS.
Thursday was our Go Live day so we had three working days to complete all checks and get things ready to go. In all honesty, by Tuesday it really felt like we’d done all we could on the content side. Without SiteImprove running on the Prod server (yeah, thanks SiteImprove — 4 business days and one page crawled) we recognised we’d be flying blind a bit, but nothing was cropping up to suggest we couldn’t switch over as planned.
I’d gotten very comfortable in my self-appointed role as CMS Super Hero (honestly, at one point I even suggested I should wear a cape) but what I learned was this wasn’t what others needed or wanted. They wanted to do it themselves so they would know how to do it in future (what, no mansplaining? Who knew?) So that took a bit of adjustment and I was childishly put out by it initially, until I grew up a bit.
We shifted our focus to using some already booked-in sessions to do some group learning, picking a theme or topic so one could show the others (slowly, and answering questions all the way) about how it is done in the CMS. As a starter I’d suggested learning about the Home page, since it had a bunch of different navigation menus we’d not talked about. This ending up bearing fruit; we’d stripped tier one landing pages of descriptions to minimise the text shown but realised the impact on SEO, so moved instead to request a tweak to the template instead, which Placecube delivered really quickly.
While other problems and technical glitches were found (and fixed) it did feel like a leisurely saunter towards Go Live rather than the up-to-the-wire rush I’d been expecting. A temperature check with the team late on Wednesday, ahead of a final check-in/confirmation about the switch over) showed the team were more anxious than me, but there was a fair amount of “let’s get this over with” in the air. So we did.
A 6am start on Go Live day to switch over the domain and see if anything exploded. Which it didn’t. Initially. And then we started uncovering the problems…
One of these was partly down to me. I’d spent a quiet Sunday afternoon cleaning and checking all the (3,000+) URL redirects that had existed on the old platform, but hadn’t done a good enough job. There had been some weird re-shuffling of the URLs on my sheet, and even though I’d caught some I still managed to leave others in. This took down our MyLocal service to download bin collection calendars. On the flipside I had a really lovely online chat with a colleague from the waste data team (who I instantly clicked with) and we worked on checking a testing the new redirects to use.
Another problem with redirects put our users in a circular error, unable to actually make a benefits claim. And when that was resolved we found an in-page app designed for us as part of the development wouldn’t let us in to change it.
Both of these issues put me front and centre with service colleagues to help resolve, and I could see why content designers love working with the services they look after — we have some lovely, patient and understanding colleagues that are absolutely committed to delivering great service to their users, and are brilliant to work with.
The bigger problems on Go Live day concerned the use of the old platform as a repository for css files used by other applications. Suddenly having these files unavailable meant our planning system intro pages were almost unusable, but also that our ModGov committee papers system wouldn’t load (and this on the day of a Full Council meeting). Unbelievably, despite having the new theme already ready , with the right URL it took the best part of a day for issue to be resolved by ModGov.
As this wasn’t really a ‘website’ issue I found myself feeling distanced from the whole thing, confused why this had been missed and a bit cross that it would be associated with the new website. That needs looking at as an approach as we’re now on a hosted system with no direct access to servers.
As the volume of feedback picked up during the day (an email every minute at one point, with numerous repeat reports of the same problem) I tried to pitch in to help, and ended up taking on the infamous team inbox where feedback arrives. This laborious and brain-numbing task is allocated on a rota basis in normal times, and it was a valuable lesson in seeing what comes in. It was a great confirmation that the data captured in the feedback really helps pinpoint the issue, with landing URLs and referring URLs, as well as routing non-website issues to service teams.
The snag list put together by Placecube to focus effort where needed most was a great move, and we had to turn to this numerous times that day to track progress and re-assess priority fixes. We did find some that were more product issues than errors, but it took the heat of the moment to bring out their importance; a problem with content scheduling meant pages on a new application were exposed to Live (which coincided almost perfectly with our service colleagues sending both the wrong date and URL to their users); a lack of a URL redirect tool meant we couldn’t apply quick fixes ourselves to issues. We’ll need to work with Placecube to prioritise these on the roadmap and get them online as soon as we can.
Or at least, in a week or so as I’m finally going to take some leave. I don’t believe in bragging about long hours as a measure of hard work, but I know that I’ve worked as hard as I could for a few weeks too many and need to reset and come back to this a bit fresher, a bit kinder and a bit more patient. So no weeknotes next week while I take a break.