Mike Gallagher

I’m currently the lead designer for the NHS App at NHS England

Right now, I’m re-making this website as a way of trying to trick myself into writing on the internet. It is a bit of an experiment and mostly weeknotes. We’ll see.

Interfaces, morality, and care

Weeknote, w/c 21 September 2025

Right now, the NHS App is mostly a website. For what feels like forever, I’ve been advocating for us to take a hard look at our approach to technology because of how much of an effect it has on our approach to design. This autumn, we’re going to make some prototypes to probe the various options for how far we can push our use of native code, which exist on a spectrum ranging from do nothing to re-platform the whole thing. Going down the latter route would be costly, both in terms of upfront resource allocation and longer term changes to our ways of working. So naturally people ask questions like, “Would that really the best use of everyone’s time?” and “What new things would that work enable?” Both are fair queries, but I believe that re-platforming is the way to go.

I can answer those questions in terms of performance, cost, beauty, tooling, or accessibility, but my primary motivation for wanting to go toward the extreme end of the spectrum and re-platform the whole thing is moral. That is a hard angle to take in an environment where reducing administrative burden and costs are the primary drivers for so much of our work (and rightly so: there is a lot of waste in the system). However, if I’m being honest about why I’m invested in this work, it is ultimately because I believe patients deserve high quality software from their public institutions, and I don’t think we can achieve the requisite level of quality without making significant changes in how we approach technology and design.


I’m not a healthcare worker in any traditional sense, but I do work as part of the wider health system in my adopted country. Everyone who labours within the NHS does so under the banner of the NHS Constitution, which says things like:

We earn the trust placed in us by insisting on quality and striving to get the basics of quality of care – safety, effectiveness and patient experience – right every time.

I don’t think that we can fully deliver on that promise with our current configuration. Most of our current design approach is taken from web design patterns, which is natural enough given how the app is built, but this prevents us from being able to fully align the app with native platform conventions. That in turn causes friction, misaligned expectations, and a degradation of trust. I want to change that.

For me, investigating how to improve our approach to technology isn’t about adding new features. It is about making the NHS App work like an app. I want to meet users’ raised expectations of software and services. I’m not so foolish or headstrong as to assume that we must make these changes to our approach regardless of the cost. The work might turn out to be too hard or too expensive, but I think the prize is substantial enough that we at least need to give it a try.


This week I was at SDinGov (for the first time!). One of the talks I attended added depth and detail to the way I think about this exploration into how we use technology. In Care for the public: trauma-informed service design, Rachel Dietkus did a brilliant tour of trauma-informed design principles and the implications of bad service design, exploring questions like:

  • Might we think of bad design as “trauma by design”?
  • If users shoulder the burden of bad design (and delivery), how do we work toward lightening the load?
  • How might we “move carefully and fix things”?
  • What does a more “hope-full” approach look like?

Dietkus’ plea for “care as infrastructure” hit me right between the eyes because of how well it elaborates my recent arguments for altering our approach to design and technology. If pursuing a trauma-informed approach is about trying to prevent our services from creating harm (intentional or otherwise), excess friction, or neglect, can we then say that attention to detail and quality in interface design is an act of care for the public?

A very practical example of how these things connect can be found in our humble back button. In the NHS App right now, the back button is a little wonky. It isn’t positioned correctly and sometimes it works in non-standard ways. Sometimes it even disappears. This happens for explicable reasons (the mixture of native and web code, the way integrations work, etc.), but it adds little bits friction to the experience over and over again, given that it is on basically every page. How do we weigh the benefit of fixing little frictions like this against adding a flashy, announceable new feature? Each annoyance caused by wonky design is manageable on its own, but over time they add up like so many papercuts. The net result is frustration, confusion, and things feeling just a little bit shit. We can do better, and according to Dietkus, we must.


While floating around SDinGov between sessions I had a chance to check in with some former colleagues from TPXimpact (né Futuregov). They had a booth and were handing out stickers with the slogan “we can change that”. I’ve always liked the simple optimism this phrase represents. Sure, every situation is complicated and there are lots of trade-offs to balance. Sometimes there is no right answer, but we’ve got hope that we can make things better. Not an ignorant hope, but one rooted in the experience of having done it before. The many frictions that exist in the NHS App’s user experience now don’t need to exist. They aren’t a law of nature. I believe we can fix them. We’ve done it before. The sticker is now on my laptop as a reminder.

Permalink

Cognition in the flotilla

Weeknote, w/c 11 August 2025

When people in the NHS discuss strategy, tactics, and how hard it is to change the current direction of travel, you often here phrases such as “its like trying to turn an oil tanker”, but that analogy has never sat particularly well with me. Earlier this week, Steve Messer wrote a thing that gets so much closer to what reality feels like. It is hard to reference his post without quoting nearly the whole thing, but here’s the most salient bit, logged for future reference:

Regardless of how many people are on your boat or how big it is, you need one hand on the tiller.

That doesn’t mean one person calling all the shots. Captains will listen to their crew and get feedback on the prevailing conditions. They use that information to adjust the course and turn the tiller.

[...]

Being a leader in today’s world is more like being an admiral: having influence over several boats rather than being on every one. Clear direction, clear decisions, and clear coordination.

Many hands, many tillers.

My situation is probably more akin to whatever you call the ninth in command of a small flotilla, itself part of a much greater fleet.

With contemporary, product-led ways of working, where each team is meant to be empowered and autonomous and agile and whathaveyou, trying to produce something that looks like coordinated action across a flotilla or fleet is a bizarre experience. You’ve got teams and clusters and programmes and portfolios and directorates all trying to play their part, but in the public sector you operate amidst nearly constant changes in the direction of the wind, the choppiness of the water, and the amount of cloud cover. Things tend to get blown all over the place and before you know it: mess, entropy, sunken ships.

By happy accident, I’m currently reading Cognition in the Wild by Edwin Hutchins, which, get ready: is about distributed decision making (on boats). I’m only half-way through but thus far the key concept is that, taken together, people and rules and objects can comprise a cognitive system that can be said to have its own computational power. We’re wading into actor-network theory, and going back to Steve’s post, might we think of the fleet itself as a thinking entity?


I continue trying to keep the big difficult project moving forward. This work isn’t on our public roadmap yet and it isn’t about AI or the Single Patient Record, but in my estimation it is really very important. It is also an absolute slog. If we do manage to make the thing happen, I’m not sure how much energy I’m going to have left to actually work on it.

Much of my energy right now is focussed on devising a plan for how to approach the topic, which involves identifying who should work on it, under what guidelines, and with what means. Sound familiar? Per Edwin Hutchins, it is about establishing a cognitive system – a thinking machine – by assembling and arranging the elements of a collective.


Amongst my cohort of lead designers in NHS England, we’ve discussed weeknotes and blogging a little this week. More and more people across the org are writing about their work on the internet, which is great. Between Caroline and Frankie and Irina and James and Joe and Kathryn and Matt and Max and Micol and Ralph and Rebecca and Rochelle and Sarah and Sarah and Tero and Tom and Trilly and Vero (and probably people I’ve missed; sorry!) it feels like we have a substantial public community developing around the work.

Design histories are part of this too, as are the Github issues where people document what they’re learning about how the NHS design system works in practice. Both are useful tools for shaping future decision making.

There is a sense of energy that comes from this collective publishing endeavour – so many people working in the open! There is also a sense of overwhelm stemming from the sheer volume of words being put out there – how on earth does one find the time to keep up with all of this? On Friday, Joe wrote about how he deals with this tension (he also referenced me and frankly I don’t know what to say other than it is extraordinarily nice). For my part, I don’t try to read every single thing right when they are published – it isn’t the news; it is enough to know that it is happening and that the material will be there when you need it. That and I use RSS religiously, off-boarding the act of keeping track to a machine.

All of these messages in their proverbial (digital) bottles are another aspect of trying to direct the fleet. Posting about our work online is an indirect method of steering, but I think the collective knowledge and learning that is captured is a hugely important part of group decision making. This corpus of material is another element of the cognitive system that makes up design in the public sector, multiplying and connecting the hands and the tillers.


(How’s that for torturing an analogy?)

Permalink

A million little pieces

Weeknote, w/c 4 August 2025

I’m working on plans for a new round of work that will explore our general approach to designing the NHS App, our design system, and how we make use of native technology. If I’m lucky, this is going to be a major part of what I spend my time on for the rest of the year. Writing about this in public sets off a certain amount of trepidation because I don’t want to jinx anything. I’m not a superstitious person, but getting to the point where we might take this on for real has been such a long road that I don’t want to take any chances of something going wrong, be they material or cosmic.


We’re getting on the design histories train. There isn’t much there yet, but we’ve set out our intention of creating a public record of why we’ve made design decisions, and much like this very website, the mere presence of the design history site is motivation to add to it.

We’re also trialing new a set of design principles that aim to clarify our definition of what good looks like. They’re an amalgamation of various sets of design heuristics, tailored to our domain and values. They still need a bit of work, but we’ve now got a few teams putting them into action and the early signs are positive.


A discovery looking into a niche component of appointment booking concluded this week and the main thing I walked away thinking was that we can’t solve any of the challenges that have been identified whilst in our current shape. As a team dedicated to working on the NHS App, we aren’t set up to solve problems that require changes to how services work outside of the app – not by ourselves, anyway. This is one of the big, persistent issues with having a team that works on an app when said app is only ever a window into a much wider and deeper system.

When the problems that make the app less than it could be exist well outside of our area of control, whose job is it to fix them? Marianne Brierley and Jane Maber cover aspects of this issue very well in their article The space around the thing: why products alone won’t transform healthcare. They say:

The NHS environment is especially complex. It’s governed by opaque, interwoven factors – structures, behaviours, legacy systems, safety protocols, policies, culture, and people. And when those forces aren’t understood, or accounted for, even the best designed product will struggle to survive rollout.

Yes, exactly. In my team we can add an extra dimension to the challenge: we can’t even responsibly design the thing without first sorting out all of the spaces that might surround its hypothetical future area. It is a very chicken vs. egg type of problem.


Cathy Dutton published an article this week called It’s time to get serious about design that absolutely nails a lot of what I’ve been struggling with lately. The basic pitch is that now – in 2025, following on from the Blueprint for Modern Digital Government – is the time to reset what we expect from design, moving away from digitising paper processes and formulaic approaches toward something messier and more imaginative. Further, this change is the only way we (collectively) will ever be able to deliver on the ambitions set out in the blueprint.

Dutton pulls out a quote from the blueprint:

We need to holistically improve policies, business processes, data, and systems rather than on a piecemeal basis.

That sounds about right for addressing the 10 Year Plan, no? The reset being asked for mirrors some of what Kuba Bartwicki describes in What’s a good design team anyway?, specifically the note that a good design team “can ship, but can also dream”.


There are stickers proclaiming “be bold” all over the place in the office. Mostly this is the doing of the prevention services gang. Is this the responsible, considered version of YOLO? Perhaps something is in the air.

I find this declaration easy to internalise but tricky to operationalise in my current role. Most of the big bold things I want to do sit outside the team’s remit (see above). Pursuing bold ideas means coalition building to assemble and coordinate all of the many little pieces required to make significant change happen. The good news is that cross-team plans across the entirety of the NHS are beginning to come into focus, however the amount of zoom in / zoom out telescoping required to keep track of everything is rather a lot for one’s neck muscles.


Speaking of the 10 Year Plan, now that we have a collective north star, I’ve been getting involved with an ever growing number of conversations in which people ask some version of “ok so how do we fix this big systemic problem that affects all care settings?” It is too early to tell whether any of the conversations and diagrams and workshops and project plans are going to solve any of the challenges they are aimed at, but there is a palpable sense of energy in the air.

None of the topics being discussed are novel, none of the ideas being proposed have never been had before. The issues affecting the health system are well known and thoroughly catalogued (and have been for quite some time), but the level of ambition set out by recent policy documents has had a galvanising effect across the organisation that is really nice to witness. What a time to be alive, eh?

Permalink

Older posts: