Weeknotes 382

What did I do this week?

Among other things…

Helped a new product team figure out their rituals.

Had a chat about website accessibility and what it really takes to make a website accessible.

Wrote a plan for running experiments to optimise web page layouts.

Figured out the technical architecture for a new product and how it’ll evolve over time (gotta love decoupled systems).

Arranged onboarding for our new service design lead, and began planning the line management transition for the design team.

Started prioritising the team’s work for December. It’s already easier than last month thanks to dropping a few projects that have been hanging around or didn’t start as expected.


I set myself three goals and achieved 67% of them.

I took a day off which meant I completed 71 tasks, averaging 14.2 a day (over five days because of course I worked on my day off). My average since I started tracking in August is 11 tasks a day.

For my objectives, 18% of my tasks contributed to one, 54% to my second, 0 for third, and 10% for my fourth objective.

I had 61 interactions with 26 people, much fewer than last week.

Managed to keep my meeting time to under 10 hours this week, but next week I already have 14 hours booked and it’ll probably go up.

What did I read?

North star

Read Amplitude’s post and ebook on north star metrics. I particularly like the checklist:

  • It expresses value. We can see why it matters to customers.
  • It represents vision and strategy. Our company’s product and business strategy are reflected in it.
  • It’s a leading indicator of success. It predicts future results, rather than reflecting past results.
  • It’s actionable. We can take action to influence it.
  • It’s understandable. It’s framed in plain language that non-technical partners can understand.
  • It’s measurable. We can instrument our products to track it.
  • It’s not a vanity metric. When it changes we can be confident that the change is meaningful and valuable, rather than being something that doesn’t actually predict long-term success—even if it makes the team feel good about itself.


Read a bit more of Better Value Sooner Safer Happier. One of the ideas I find really interesting is ‘Invite over inflict’, that change is more successful when people choose the change rather than having it forced upon them. Apart from treating people better, it shifts the discussion away from oppositional top-down vs. bottom-up.

Insight into the neurodiverse experience in UX

This is an interesting piece from Dr Maria Panagiotidi on how “properly supported, neurodiverse perspectives can bring immense value to UX through diverse skills and innovation”.

What did I think about?

Four questions for charities to ask themselves about AI

How do we use AI?

Just as introducing any new technology in responsible ways takes consideration, charities should introduce the use of AI in ways that help their staff and volunteers figure out how to use it effectively.

How do we affect how others use AI?

Charities, in fact all of civil society, have a role to play in shaping how other organisations use AI. Figuring out how to be part of the conversation around the responsible use of AI is an important role for all charities.

How do we help people affected by AI?

If a charity supports people, it will be supporting people affected by AI. Increasingly, more people will be affected by automated decision making, job replacement, deep fake revenge porn, etc., etc. How to help people facing these situations is something more and more charities will have to do.

How do we affect AI?

This is probably the hardest question of all, but still an important one to ask. Can a charity have any affect on AI models, the data they use and what results they produce about their cause? How can they provide the right information for the AI models to learn from?

Organisational unconscious incompetence

I wonder whether De Phillips, Berliner and Cribbin’s Four Stages of Competence applies to organisations? Can organisations do things and not know that they don’t know how to do them well? And how does it affect strategy? Should an organisation only develop a strategy for areas they are consciously competent in? The problem is, if a strategy relies on things that the org can’t do very well but doesn’t know they can’t do, then it will almost certainly fail.

Fewer things better

It’s easy to glibly agree with the idea that an organisation should do fewer things well rather than lots of things poorly, but obviously it’s more complicated than that. Maybe organisations should aim for doing fewer wellestablished things better and doing lots more poor quality, experimental, things to learn from. The key point here being about learning. A organisation should have learned everything it needs to from the fewer well-established things and be set up to learn as much as possible from the new things so it can figure out which will become well-established in the future.


Consistency of solution comes one step at a time. Slowly. After you’ve got all the other things in place that make it possible, things like trust, reliability, knowledge and expertise.