I was bored yesterday on a train and challenged myself to rapid fire (it was only Bristol to London) jot down as many (hardly) 🔥hot takes🔥 that might make for interesting blogposts that might generate a bit of discussion on BlueSky.

In the end what I mainly realised is that most of my thinking crystallised in about 2017 and I feel a bit like Abe Simpson yelling at the clouds.

Anyway I wrote quite a few words so I’ve added some links and I thought I’d share them so here goes >>

---

Discoveries have become the ends not the means. They have become too long, too broad, too dogmatic. Just do enough research to make a decision to a problem not to understand an entire sector…and sometimes the insights already exist.

Will’s Three ways to run better discoveries and Sarah’s Discovery guide for Defra are both long in the tooth but have not been bettered.

---

We need tech insight much (much) earlier – even if that means prototyping in Disco. There isn’t enough thought about feasibility early enough.

Loved these posts about not being scared to start with solutions from Andrew and Steve…and this classic from Richard about knowing when to run….oh and this one about when digital IS about the technology.

---

The illusion of momentum provided by Discoveries is also high. They provide cover for risk averse leadership – discoveries beget discoveries and nobody needs to make decisions.

---

Alphas have too often just become the first phase of the build – there is not enough support for testing multiple hypotheses nor for stopping or pivoting projects early once enough has been learned.

Still think Michael articulated this best back in 2017.

---

…and again we need engineers in Alphas (data and software) and access to security people (Team Onion) to ensure options are feasible outside of the Prototype Kit!

---

We still don’t give content design enough attention – it is the fix for so many issues but also the route to identifying so many more.

---

We have to get to Live much faster and invest time, money and talent *then* – turn the current model on its head – we learn loads more once things are live and that is when we should be improving and iterating.

Matt wrote this in 2016 and it is all still true and rare.

---

Make things open, it (still) makes things better. Please. Plenty of individuals do so but official blogs have become corporate comms channels with little real insights and when there are it’s months after the fact. To be authentic we need to share the rough with the smooth and do our thinking in public – it shouldn’t just fall to individuals. Teams need to be encouraged, supported and protected so they can contribute to the Commons.

---

I actually do think AI might genuinely be a gamechager – but it isn’t yet despite the hype and it isn’t an excuse not to #fixtheplumbing. We especially still need to get a handle on the data issues (quality, governance, sharing, standards…). Want an AI enabled future – invest in the foundations.

---

The DDaT / Gov Digital and Data jobs families have become too siloed / too specialised. There aren’t enough glue people anymore. Too few generalists. Not enough empathy for other people’s challenges. I think the rise of CoPs contributed to this – which is a shame as I think they are hugely important.

---

Given it is my contention that the main purpose of Product people is making decisions we have armies of them not empowered or trusted to make any real choices other than reordering a backlog.

---

The failure of Agile©️ is evident everywhere because the point was lost in the rush to adopt the elements of the frameworks that made project managers and leadership comfortable losing sight of the spirit of the approach.

There are still people doing good agile thinking but it hasn’t delivered like it should have…and I say that as an agile advocate tip to toe.

---

… and a symptom of this is the endless battle to land agile product roadmaps when what leadership really want are Delivery plans full of false certainty – even if they ask for roadmaps.

---

A kind of Service Manual fundamentalism has emerged where the guidance included has become a dogma to be followed religiously even at the detriment for successful delivery. Now I think the Service Manual is the greatest that came out of GDS 1.0 and totally changed the game but it is a guide not a rulebook and not every team needs every role for instance!

---

…related to this is my fear that too many Service Assessments have become box ticking project gateway interviews and not not peer conversations and that an entire industry has grown around the idea of preparing to ‘pass’ them. At their best Assessments have been amongst the most helpful professional experiences I’ve had (on both sides of it) but at their worst they have felt the very opposite of the psychological safety we seek elsewhere.

---

Terence is right – it is time to put away the anti-app rhetoric in Gov digital circles. I mean Tom was also right when he wasn’t ‘appy but a lot has changed in a decade plus. I mean I still don’t think GOV.UK needs an app but I am ancient.

---