Dan’s Weeknotes s07e18

Week ending 2020–11–27

Dan Barrett
Web of Weeknotes

--

Oh, hi

Hey friend. I appreciate you stopping by.

I had one of the best working weeks in a while. Looking back at my previous weeknotes (see episode 17) I feel like I achieved more last week and I felt more motivated. I had some ideas and came up with solutions to problems.

Having said all this, I still ended the week feeling there was a mountain of things outstanding.

What was good?

  • Tom (chief analyst) and I had a good retrospective and planning session with Lee (delivery manager). The retrospective was particularly good. We dedicated a decent amount of time to it, had an open conversation, supported each other, and ended up with some practical actions.
  • Following on from the metrics work I mentioned last week I did the next stage of development and I was satisfied with where I got to. I thought it was a solid piece of work.
  • This article via my colleague Rebecca (Director of Customer Journey) was good to see. I enjoyed the perspectives and the turning ‘private sector knows best’ on its head.
  • I put together a presentation on data strategy for the data platform team and it went really well. I really enjoyed the session, it helped me arrange a variety of thoughts, and the team gave me really positive feedback. I’m going to repeat the session with another team soon.
  • I thought this blog post from my colleague Elizabeth (principal energy policy manager) was excellent:
  • I think I’m a fairly good and constructive coach. It was good to hear that a small intervention I’d suggested had been picked up, realised, and made a big difference.
  • Big fan of the Royal Borough of Greenwich digital strategy led and shared by Kit Collingwood:
  • Pooja (data scientist) has been working on incoming search terms from Google. My position is that this unstructured data could provide us with new insight into our clients’ experiences and needs, along with wider trends in society. It’s a good candidate for an exploratory data science project and I’ve enjoyed working with Pooja on it. There are around 24,000 unique search terms bringing people to the Citizens Advice website each week. Google search console gives you the top 1,000. Pooja has accessed all of the data using the API. Often the work that Pooja’s doing — which features some heavy duty maths and natural language processing — is at the very edge of my ability to understand. But I’ve learned things and really appreciated Pooja’s patient, expert explanations. The aim is to get a stable pipeline for collecting, transforming, and reporting on the data before Christmas. One of my main conclusions from the work is how interconnected the ‘advice’ domain is. This means it’s relatively hard to distinguish one need from another. But we’re getting there. Even without this data science piece, incorporating incoming search terms into our regular data discussions as a team has been excellent. At the very least, it adds some interesting context each week. At best, it’s helped to advance our practice and fill in some gaps in our understanding.

This story is absolutely wild¹:

What could have been better?

  • I just want to get through more. I think Monday was particularly weak.
  • Business planning is here and I feel on the back foot. It’s critical this year that I have some success in this.
  • I am not reading enough blog posts by other people.

What are you looking forward to next week?

  • Kicking on with some of the things I’ve started.

Thanks for reading. Hope you’re doing ok. There is good in the world. Stay safe. Fight racism.

Footnotes

¹ I do love non sequiturs

--

--

Head of Data Science at Citizens Advice. These are my personal thoughts on work.