Matt Knight
mattinwales.bsky.social
Matt Knight
@mattinwales.bsky.social
New here.
No budget and no structure chart. Hmmm.
November 21, 2025 at 8:20 AM
This feels relevant (if you didn’t see it when it was published)… heywoodquarterly.com/the-end-of-t...
The end of the generalist? - Heywood Quarterly
Tamara Finkelstein, Head of the Government Policy Profession, asks whether it’s time to consider the end of the Civil Service generalist.
heywoodquarterly.com
November 21, 2025 at 8:09 AM
Ah, the memories. Does HoHo still have the sodding lethal spiral stairs and the security folk who’ll rugby tackle you to the ground if you try and walk up them with a cup of tea? Took me months to find the back stairs that everyone else used.
November 7, 2025 at 6:21 PM
My experience is that the definition of private beta in the std allows a lot of room - it can be time limited, user number limited, etc. IMO, critical factor with (private and public) beta is that there’s an alternative to fail over to (online or offline) if something fails in scale up.
October 22, 2025 at 5:18 PM
Agree. And I think the standard should support that. One of the questions I used to ask teams as a lead assessor was to show me part of the service that had iterated multiple times, what had been thrown away, and why. But an assurance process can’t replace the role of leadership and risk appetite.
October 22, 2025 at 5:14 PM
“If you're not embarrassed by the first version of your product, you've launched too late” - Reid Hoffman.

But this requires genuine psychological safety and cover for the people doing the ‘scrappy real world test’, which is difficult in government.
October 22, 2025 at 5:09 PM
Apologies - last link should be

www.iterate.org.uk/the-value-of...
The value of user needs – iterate
www.iterate.org.uk
October 21, 2025 at 7:47 AM
(Two things I wrote that make similar points in more words:

1. www.iterate.org.uk/why-user-res...

2. www.iterate.org.uk/a-governance...)
Why User Research is always worth it – iterate
www.iterate.org.uk
October 21, 2025 at 7:21 AM
(3/3)

- the most radical point of the standard is that, unlike every other governance standard, it gives a voice to the real end user. Policy makers on their own, technologists on their own… are really bad at guessing what users do. User Research is the radical democratisation of service creation.
October 21, 2025 at 7:17 AM
(2/3)

- Discoveryitis can’t be blamed on the standard: there is deliberately no assessment to exit Disco and most of what gets called Disco is actually Alpha
- Testing things by building production quality code is always tempting to technical people and always inefficient and always constraining
October 21, 2025 at 7:14 AM
As someone involved a tiny bit in creation/revision of the Service Standard, I acknowledge both the creeping encroachment of Goodhart’s law and the fact that it has its roots in an attempt to prevent government in 2013 from self-harm.

But…

(1/3)
October 21, 2025 at 7:10 AM
You saw through my cunning plan… would still make a decent travelogue.
October 13, 2025 at 9:31 AM
Reposted by Matt Knight
Passmore Edwards funded institutions for me.
October 12, 2025 at 9:20 PM
That does look more achievable on the numbers front as well.
October 12, 2025 at 10:07 PM
Carnegie funded public library buildings is a tempting one, personally.
October 12, 2025 at 8:35 PM