These weeknotes come to you belatedly, as I was on holiday on Friday and spent the weekend celebrating my 30th with friends. It’s still a great practice though, helping me tie up the week and reflect on what happened. Makes me wish I’d been doing it earlier, there’s so much from building eCasework I could have shared. Thank you to the weeknotes folks for your kind words and encouragement!
Four Things That Happened
What started out as a packed day ended up being free; meetings moved around giving me time to think. It’s important to have that time to process things, to actually do the work of a PM.
In the afternoon we had an incident review – a meeting covering the causes and solutions of a technical incident – which has given us more reason to work on the publishing pipeline this quarter. As mentioned in my last weeknote, I’m keen we take on that work, and the consensus in the room was that we needed it now. With a major national event on the horizon, ensuring the pipeline can reliably deliver content is key.
There’s varying levels of experience in the team and I’d been planning on holding a mapping session followed by some problem sorting. After that we could implement some request tracking to measure the pipeline, to find out where the bottlenecks were, and then jump into making changes. Tech leads and senior devs thought that my approach was sound and that we needed to build collective understanding – it was comforting to hear my thinking validated, in some part. Many in the team know how to tweak things to fix them, but don’t know why to tweak those things or by how much.
Andy Dudfield, Service Owner at ONS Digital, popped in to chat to the GOV.UK product manager community about the ONS website and how it had been transformed over the last couple of years. I’d originally invited him in to hear about all the work they’d been doing on data, to help us think about data.gov.uk, but there were lots of questions from the community about the general website too. It was super pleasing that everyone was so engaged, there was lots of food for thought on the next steps for GOV.UK.
Andy’s presentation and the work at ONS Digital – kicked off by @jukesie’s work, which Andy was keen to point out – gave the PMs corroboration on our thoughts about user personas, DataOps, SEO, the definition of the term open data, and how to engage generalist audiences.
After that we chatted about the campaigns platform and how ONS might use it for the upcoming census. It’ll be good to have their needs shape the product alongside upcoming EU Exit work, to help avoid feature-bloat and keep the platform’s utility more general.
Thinking about how we define a product’s success in the public sector can be tough. Programmes and transformation are designed to effect change, but a digital product is propelled by metrics. A Product for Good meetup on defining metrics for social impact provided great insight from the tech for good community.
How can you turn qualitative comments into something you can measure quantitatively?
Matt Kepple from Makerble has done lots of thinking on scaling impact measurement and trying to measure in the short term, where indicators are more attributable to your efforts. He said that changing what people do, think or have are the three types of impact story or behavioural change that you can influence. It touched on thoughts I’d been having back in February and raised in our 52 Weeks of UX lean coffee about measuring emotional impact. I’ll be able to develop that blog post more now.
On Thursday we set aside two hours to map out the publishing pipeline and its problem space. It was my first time leading a discovery-esque approach to a live ‘product’ (arguably the platform is an assembly of products), but it went well. Our experienced developers shared knowledge amongst the team, gaps were filled in by others, and the new team members brought new ideas to the table. Following on from a tech intro the previous week, I was able to qualify my understanding of how it all fits together, which was pleasing.
After that we had a discussion on what problems we’d been facing and why, and the conclusion was that we can’t say for certain: we need good metrics to help expose where the bottlenecks lie. Though I had been directing conversation that way to a degree – the Build-Measure-Learn loop is central to hypothesis-driven development – it was great that Paul, previously the programme’s tech lead, called for this. We’ll also be able to take a look at work Elliot had done a couple of years back on request tracking, which will not only inform us but could help rebuild trust in the publisher community. ‘Getting two birds stoned at once,’ as Ricky says.
Excited to see how conversations have developed when I get back from holiday next week.
- Stop making new things. Start making old things work.
- Reach out and meet people. We have a great community in gov.
- There’s other communities you can learn from too!
- Oh heck, I love this job.
What I’m Reading
- Putting users first is not the answer to everything
- Ten Reasons Not to Measure Impact – and What to Do Instead
- Why you need “never” in your product roadmap
- 5 things Nathan Barley taught us about Service Design
What I’m Cooking
Should have made notes of these, I can’t remember everything we made. To those of you who asked for the aubergine recipe, I’ll put something together (though it was technically made by my gf!)
- Texas BBQ chicken with chargrilled sweetcorn rice salad
- Marinated garlic courgetti with Dijon mustard dressing and rocquefort