Weeknotes S03E10
A productive and reassuring week of moving forward. Mostly.
Some things I’ve been doing and thinking about this week.
Shared understanding of terminology and ownership.
Last week I mentioned that departmentally we are moving towards a distributed data architecture. This necessitates some change in ways of working and ownership. Distributed data leadership, product and pipeline management. Centralised standards and governance.
Lots of my time over the last few months has been spent in conversation about these things. This week is has been standards and governance.
I mentioned our standards work in S03E01 of these weeknotes (and probably every one since). When I first moved into Health Digital I mentioned the need for this work. It was sometimes met with perceptible twitches in tone or body language. The feeling was this should be centrally owned.
I think part of this is down my language and explanation. There are a few terms that can tend to be used interchangably. So this is where my head is at.
- Policy. A statement of intent. A formally documented system of principles. Describing an approach to guide decisions and achieve outcomes.
- Standard: What the required or agreed level of quality is.
- Procedure: How a policy is implemented.
- Guideline: Similar to a policy but in relation to discretional standards.
So broadly speaking a policy is a business decision on whether to implement and meet one or more specific standards. It dictates whether a standard is mandatory or voluntary. Controls need to be in place to see if mandatory standards are being met.
Guidelines help augment this when there is discretion. Individual business areas may want to raise their standards. For example through recognised industry best practices. Where data is concerned there are many industry standards.
Note 1. This all starts with external influencers. They establish what is considered to be due care. They can impose penalties for non-compliance. Plus internal influencers. They establish the approach to consistency and balance across business strategy, budget etc.
Policies, standards and controls are designed to be centrally managed. These are best defined and developed by separate teams due to the competing dynamics and aims when balancing effort and quality. Procedures and guidelines are by nature de-centralised.
So in the context of our work in Health Digital, when I say our standards work. This is a combination of developing our guidelines and procedures. Plus linking centrally on standards and policy. Both departmentally and to the Data Standards Authority.
Note 2. This needs dedicated central teams for policy and standards. You can’t deliver global standards through local exemplar business area projects. As project specifics will inevitably bring compromise. Local projects can sometimes deliver reusable procedural patterns. Which is the way we are working.
Note3. In a world with a focus on delivery over enablement, it’s great to see recent jobs being advertised for policy and standards development. These jobs should be considered cool!
Micro and macro planning, iteration and linking up
I mentioned that we are working to make as much of our procedural and guideline artefacts as re-usable and shareable as possible. Last week in S03E09 I showed the sketched out landscape of my role in Health Digital. This was also designed to be fairly standardised to reflect similarities with other areas. This week I’ve had a lot of conversations about these possibilities.
I’ve spoken quite a bit with Emma who has a similar role to me in Universal Credit. We’re both involved in conversations with the centralised data part of the department about our roles within business areas. This focus has brought some valuable introspection. The process has made us talk more and consider how we can join up between our own teams in terms of people, procedures and products. There are a lot of similarities but also some differences. These are due to levels of maturity in different areas. Which reflects different starting points on both sides. Step one is Jason talking through his user research.
I appreciate Emma’s cautious-optimism to balance my sincere-cynicism.
I’ve also been talking to people in the Carer’s Allowance part of the business about our technology approach. Child Maintenance Group about our general approach — it was good to speak to Rich about some holistic concepts and service design. Plus the team responsible for the central data mesh co-ordination about data maturity. All validatory and reassuring.
The development of our data platform continues. This week has seen discussion and planning at a very low detailed level across the engineering practice about individual components and appraoches. Plus higher-level talks about work for next year on phase two of the platform — which is exciting. I’ve enjoyed the mix of small picture and big picture.
Stu did a great job at the Show and Tell this week. I’ve spoken before about how tricky it has been to show our work due to the breadth and variety of the initial foundational activity. This time round we’ve been able to move from talking about a collection of separate activity to showing our work in terms of our procedures and guidelines. I thought Stu nailed it.
Culture, support carrots and sticks
I’ve had quite a lot of impromptu conversations this week — I like these. It’s the biggest thing I miss about IRL working. Every chat over a screen feels formal, I prefer a phone call for some types of chat at the minute.
This week I had a meeting about contributing case studies to a data masterclass for senior civil servants in policy. This has very central and senior government sponsorship. Adam and Pete are providing case studies using their work on building capability in data visualisation and communication. It felt like a little validation of taking a user-centered approach to data — which has been hard work over the years.
When taking any approach, as with the policy and standards diagram above, the reality is there are external influencers and policy ahead of defining your strategy. User Centricity needs to be an intent before it can define the strategy. If it isn’t, a lot of work needs to go into convincing and influencing. This has been my experience within data anyway.
I made this sort of comment in response to a Twitter post this week. I later saw a comment about mansplaining. I’m not sure if it was in relation to my comment. I hope not. I don’t think my comments devalued anyone else’s experiences, expertise or views. That certainly wasn’t the intention.
Monday SLT saw a bit of an emotional check-in. It was honest and pure. There was a lot of talk about life outside of work and of social media hiatuses. I ditched Facebook some time ago and have been considering sending Twitter the same way but it feels a more difficult leap.
Over the last few months, I’ve been hyper-aware of people whose actions do not match their words. Engageing and understanding people for whom data isn’t yet an intent. But talk does not bring change. People agreeing with you doesn’t bring change. Action does. I’ve been grateful to be supported by Celine, Stu and Aaron.
