AI, black boxes, and designerly machines

On my holiday, I started reading into some topics I ought to know more about: artificial intelligence, genomics, healthcare, and the fast approaching intersection of the above. Here follow some half-baked reckons for your critical appraisal. Please tell me what’s worth digging into more. Also where I’m wrong and what I might be missing.

1. Opening the black box

large ribosomal subunit (50S) of Haloarcula marismortui, facing the 30S subunit. The ribosomal proteins are shown in blue, the rRNA in ochre, the active site (A 2486) in red. Data were taken from PDB: 3CC2​, redered with PyMOL.
By Yikrazuul CC BY-SA 3.0, from Wikimedia Commons

Reading Siddhartha Mukherjee’s ‘The Gene: An Intimate History’, I discovered the amazing trajectory of human understanding of DNA, RNA, enzymes, proteins, the genome, and the mechanisms by which they interact. There’s no doubt that this stuff will transform – is already transforming – our relationships with medicine. Crucially this generation of scientists are looking inside a black box, where their predecessors could observe its effects but not its inner workings.

At the same time, fuelled by petabytes of readily available data to digest, computer science risks going the other way in the framing of artificial intelligences: moving from explicable, simple systems to ones where it’s allowed to say, “this stuff is so complex that we don’t know how it works. You have to take it on trust.”

When we apply artificial intelligence (AI) to healthcare, transparency is essential; black boxes must be considered harmful.

It’s not just me saying this. Here are the words of the Institute of Electrical and Electronics Engineers (IEEE):

“Software engineers should employ black-box software services or components only with extraordinary caution and ethical care, as they tend to produce results that cannot be fully inspected, validated or justified by ordinary means, and thus increase the risk of undetected or unforeseen errors, biases and harms.” — Ethics of Autonomous & Intelligent Systems [PDF]

Transparency must be the order of the day. It comes in (at least) two flavours: the first is clear intent; the second, understandable operation. Both are under threat, and designers have a vital role to play in saving them.

2. The opacity of intent

It’s a commonplace to say that technology is not neutral. I won’t labour that point here because Sara Wachter-BoettcherEllen Broad and others do a good job of highlighting how bias becomes embedded, “AI-washed” into seemingly impartial algorithms. As the title of Ellen’s wonderful book has it, AI is ‘Made By Humans’.

That doesn’t seem to stop stock definitions from attempting to wall off AI beyond the purview human control:

“In computer science, AI research is defined as the study of ‘intelligent agents’: any device that perceives its environment and takes actions that maximise its chance of successfully achieving its goals.” — Wikipedia

But what goals exactly? And how did the AI get them? The Wikipedia definition is silent about how goals are set, because, in the words of Professor Margaret Boden“the computer couldn’t care less.”

“…computers don’t have goals of their own. The fact that a computer is following any goals at all can always be explained with reference to the goals of some human agent. (That’s why responsibility for the actions of AI systems lies with their users, manufacturers and/or retailers – not with the systems themselves.)” — Robot says: Whatever

When any technology moves from pure to applied science, intent must be centre stage. If we fixate too much on the computer science of AI, and not enough on the context of its application, intent will always be unintentionally obscured.

Many discussions about the “ethics” of AI or genomics are really, I think, discussions about the opacity of intent. If we don’t know who’s setting the goals for the machine, or how those goals are derived, how can we know if the intent is good or bad?

Moreover, true human intent may be difficult to encode. In a domain as complex as health and care, intent is rarely straightforward. It can be changing, conflicting and challenging to untangle:

  • a boy was triaged on first contact as in less urgent need, but has suddenly taken a turn for the worse
  • an elderly woman wants to get home from hospital, but her doctors need first to be sure she’ll be safe there
  • the parents want to help their children lose weight, but know that pester power always leads them back to the burger chain.

In these situations, even Moore’s Law is no match for empathy, and actual human care.

3. Designers to the rescue

Design, in Jared Spool’s wonderfully economical definition, is “the rendering of intent.” Intent without rendering gives us a strategy but cannot make it real. Rendering without intent may be fun – may even be fine art – but is, by definition, ineffective.

It’s time for designers to double down on intent, and – let’s be honest – this is not an area where design has always covered itself in glory.

We know what design without intent looks like, right? It’s an endless scroll of screenshots presented without context – the Dribbblisation of design.  If you think that was bad, just wait for the Dribbblisation of AI. Or the Dribbblisation of genomics. (“Check out my cool CRISPR hacks gallery, LOL!”)

Thoughtful designers on the other hand can bust their way out of any black box. Even if they’re only called in to work on a small part of a process, they make it their business to understand the situation holistically, from the user’s point of view, and that of the organisation.

Design comes in many specialisms, but experienced designers are confident moving up and down the stack – through graphic design, interaction design and service design problem spaces. Should we point an AI agent at optimising the colour of the “book now” buttons? Or address the capacity bottlenecks in our systems that make appointments hard to find?

One of my team recently talked me through a massive service map they had on their wall. We discussed the complexity in the back-end processes, the push and pull of factors that affected the system. Then, pointing at a particular step of the process: “That’s the point where we could use machine learning, to help clinicians be confident they’re making a good recommendation.” Only by framing the whole service, could they narrow in on a goal that had value to users and could be usefully delegated to AI.

4. How do you know? Show your thinking.

School exam paper. Question:

Crucially, designers are well placed to show the workings of their own (and others’) processes, in a way that proponents of black box AI never will.

This is my second flavour of transparency, clarity of operation.

How might we:

  • communicate probabilities and uncertainties to help someone decide what to do about their disposition to a form of cancer?
  •  show someone exactly how their personal data can be used in research to develop a new treatment?
  • involve people waiting for treatment in the co-design of a fair process for prioritisation?

In a world of risks and probabilities, not black and white answers, we should look for design patterns and affordances that support people’s understanding and help them take real, fully informed, control of the technologies on offer.

This is not an optional extra. It’s a vital part of the bond of trust on which our public service depends.

5. Designerly machines

Applying fifty iterations of DeepDream, the network having been trained to perceive dogs CC0 MartinThoma
Applying fifty iterations of DeepDream, the network having been trained to perceive dogs – CC0 MartinThoma

The cultural ascendancy of AI poses both a threat and an opportunity to human-centred design. It moves computers into territory where designers should already be strong: exploration and iteration.

I’m critically optimistic because many features of AI processes look uncannily like a repackaging of classic design technique. These are designerly machines.

Dabbers ready, eyes down…

  • Finding patterns in a mass of messy data? Check!
  • Learning from experiments over many iterations? Check!
  • Sifting competing options according to emerging heuristics? House!

Some diagrams explaining AI processes even resemble mangled re-imaginings of the divergent/convergent pattern in the Design Council’s famous double diamond.

Diagram showing how design moves from problem to solution in four stages, shown as one diamond after another. There are two pairs of divergence and convergence: Discover and Define, Develop and Deliver
© Design Council 2014 – https://www.designcouncil.org.uk/news-opinion/design-process-what-double-diamond

A diagram outlining a forward pass though three 3D generative systems, data is divergent and then convergent
“A diagram outlining a forward pass though our three 3D generative systems.” – Improved Adversarial Systems for 3D Object Generation and Reconstruction [PDF]
The threat is that black box AI methods are seen as a substitute for intentional design processes. I’ve heard it suggested that AI could be used to help people navigate a complex website. But if the site’s underlying information architecture is broken, then an intelligent agent will surely just learn the experience of being lost. (Repeat after me: “No AI until we’ve fixed the IA!”)

The opportunity is to pair the machines with designers in the service of better, faster, clearer, more human-centred exploration and iteration.

Increased chatter about AI will bring new more design-like metaphors of rendering that designers should embrace. We should talk more about our processes for discovering and framing problems, generating possible solutions and whittling them down with prototypes and iteration. As a profession, we have a great story to tell.

A resurgent interest in biology, evolution and inheritance might also open up space for conversations about how design solutions evolve in context. Genetic organism, intelligent software agent, or complex public service – we’re all entangled in sociotechnical systems now.

Advertisements

“Look after the water” – reflections 1 year into my work at NHS Digital

Other people’s jobs are endlessly fascinating. At a birthday celebration a couple of years ago I got talking to Johnny, a family friend who works as an aquarium curator. He told me a surprising thing about his work: how little of his time he spends actually looking after the fish. Johnny’s job is to look after the water. “Look after the water,” he said, “and the fish will take care of themselves.”

So it is with design leadership. Our designers have different specialisms – service, interaction and graphic design. They’re embedded across a wide range of endeavours, both public and professional facing. They’re the ones who see users in research, and stakeholders in show and tells. My role is not to tell them how to design; it’s to create the safe and supported conditions in which they can do their best work, individually and collectively. When those conditions come together, it’s a wonderful thing.

This week it’s one year since I joined NHS Digital to lead the design team. I’ve been thinking about what has changed, and what we have yet to achieve. As ever, views all my own.

Growing a team

Our third whole design team event took place a couple of weeks ago in Leeds. We ran two rounds of rapid fire show and tells: 14 designers showing their work in the space of 90 minutes. I was massively impressed that every designer who presented was so good at telling their story, under time pressure, to a room of 35 people.

We’re lucky to have two excellent lead designers. Tero heads our growing service design practice, while Dean has taken on interaction and graphic design for the NHS website. Design-minded product managers Emma, Ian, and Sophie join us in our fortnightly design leadership meetings. Stephen, who left a couple of months ago, was always insightful, knew his way around the organisation, and took on the unglamorous task of writing job descriptions. I miss our Friday morning coffees.

Hiring for designers in both London and Leeds has been a long journey but rewarding in the end. Over the past few months, it has been great to see the new seniors settling in. I believe we now have talent at every level, and a good foundation for design leadership here in the future. If my bike went under a bus on Chapeltown Road tomorrow, weekly design huddles would still happen in Leeds and London. I count that as a win. Having designers who talk to and trust each other is the foundation of a coherent experience for our users. The designers and I are rewarded on the same pay scales as nurses, doctors, and other NHS professionals. That’s a sobering reminder of the value each new recruit to the team is expected to add.

A year of recruitment in numbers:

  • Just over 50% of the design team are now newer than me to NHS Digital
  • Of the permanent staff, 11 are still here from the team when I arrived, 8 are new recruits, and 3 have left
  • Among contractors, it’s 1 still here, 7 new, and 3 left
  • On top of that are a dozen or so supplier staff with whom we work closely as members of our extended team
  • Permanent team, contractors and supplier staff alike, 100% of them want to do their best for users and the health service.

Designers work best as part of multi-disciplinary teams. After a year here, I’ve had the privilege to see a few of those teams go through the delivery cycle from discovery, through alphas and on to release private and public beta versions. It hasn’t always been straightforward. Some teams have got stuck. Some things have stopped when we realised they would not achieve the outcomes we hoped for. But it does feel that teams are getting slicker at this – learning about user needs, and learning how to work together as true multidisciplinary teams. I’m fortunate to be part of a senior leadership team with brilliant product, delivery, technology and content leaders too.

As a design team, we have access to two larger communities of practice. NHS Digital’s Digital Service Delivery profession includes design along with user research, product management, delivery management and content design. We’re also part of the amazing cross-government user-centred design community, giving us access to Government Digital Service (GDS) training and community events. I especially appreciate my meetings with Lou Downe and the other government heads of design.

Here are some things I’ve learned…

Power is a big theme in health and care.

My focus has been with teams designing and delivering for patients, carers, and families – users who don’t work for the NHS or social care. There’s good evidence that people in control of their own health and care have better health outcomes. But I can see from our research how people’s power is diminished – by illness or disability, by social circumstances, and (though we don’t mean to disempower) by us, in the way we design and deliver health and care services.

Digital information and service have the potential to make people more powerful again. This can only happen when people can get them, trust them, understand them, decide with them, and act on them. For people to take power in the NHS, we need to work across the whole system, a partnership of patients, families, professionals, and service providers.

Sometimes we need to diverge before we can converge.

A healthy tension between divergent and convergent working should be part and parcel of any design approach.

Soon after I arrived a year ago, I worried that, in some areas, we were trying to converge prematurely on solutions that had not been tested against a wide enough range of user needs and contexts of use. To reach the required quality, we had to go through a phase of divergence in which teams went off to solve their own problems, while sharing their work and looking out for common patterns. (Patterns, by the way, are never designed; they can only emerge when teams are empowered to work independently, but transparently.)

Now, we’re back to a phase of convergence around design for the NHS website, led by teams explicitly tasked with redesign and standards creation. I have greater confidence that we’re building on firmer foundations this time, because we’ve tried more things, and understood more user needs.

I’m a design system sceptic (but we’ll probably end up with a design system anyway).

Everyone seems to be talking about design systems right now. I caution that explicit efforts to create a system can easily tend towards design for its own sake, disconnected from validated user needs and contexts of use. I hope we can keep ours rooted in reality by rotating designers through the overall redesign and standards teams, in and out of squads working on specific health condition categories and services. Nothing should get into the design system without being researched with users in multiple need states and contexts.

Rather than all swarming on the same problems, we need to conserve our energy and focus. Like birds flying in formation, each team can take a turn to lead on a design challenge, before falling back to let others fly ahead for the next stage.

Critical optimism is the order of the day.

When I wrote my 6-month update, the always-perceptive Stefan Czerniawski noted:

There is a sweet spot in any job, or more generally in understanding any organisation, when you still retain a sense of surprise that anything could quite work that way, but have acquired an understanding of why it does, and of the local application of the general rule that all organisations are perfectly designed to get the results they get.

Since then I’ve tried to bottle that feeling. Healthcare certainly proves Dr Deming right: that the same system can at once be brilliant at some things and terrible at others. There can be world class care, medical and technological innovation side by side with the shocking failure waste that comes from poorly designed service. Being a permanent member of staff, I feel a responsibility to work with the grain of the system, while retaining a sense of urgency to make things better.

One of my objectives is to grow the whole organisation’s commitment to human-centred design. This definitely feels like a multi-year commitment, but I’m confident that we have director-and-above-level support for improving the quality, consistency and accessibility of digital services for NHS patients and professionals. Our head of profession Amanda has been the definition of an empowering manager. Our portfolio director Alan has an exceptionally user-centred vision for someone in a such high-profile delivery role.

Update on some things I committed to do at the 6-month point:

  • Develop my own capability – I am investigating leadership courses that might be right for me. Ideally, I’ll do something that brings me into contact with a more diverse range of health and care leaders, not just the ones focused on digital.
  • Reflect and plan – I started by block booking Friday afternoons as a meeting-free zone in my diary. Clearing emails and weekly reporting always swallowed them up. Now I’ve blocked out the whole day. I don’t always keep to it, but it’s a good reminder of the value of meeting-free time.
  • Listen better – There’s a bit almost at the end of David Marquet’s ‘Turn the Ship Around‘ video, in which he says you will fail repeatedly at giving control to your team, but get up and go again. That’s where I feel I am with my coaching practice right now. After some conversations, I come away kicking myself. When it goes right though, it’s so much more rewarding to hear a colleague solve their own problem than to hear myself offering my solution. Ultimately this is the only way that a design capability is going to scale.
  • Influence more – Lots more to do here. There are so many opportunities for improvement that our small team will never be able to address them all. By sharing standards and setting clear expectations of good practice, we can multiply our impact and give power to the many other people across the health and care system who want to make a difference with design.
  • Say no to more things – One of the adjustments in moving from a micro-business to a biggish organisation was appreciating that lots of things get done even if I don’t do them. Every week or so, I look through my to-do list for the things I really ought to delegate, and the things I’m just never going to do. Adding a “Not Going To Do” column in Trello has done wonders for my sense of productivity.
  • Say yes to more things – I’ve been lucky enough to be invited to some brilliant events and conversations. A recent highlight was working with Victoria Betton and Lenny Naar to deliver a hands-on user-centred design session at HIMSS e-Health week. In October, I’ll be speaking at Interact London, and the conference theme is “Intelligence in Design”. Fingers crossed I’ll have something intelligent to say.

Electric woks or eating together? Time for human-centred designers to care about the community

Mick Ward is sick of people trying to sell him electric woks. As chief officer leading transformation and innovation for social care in Leeds, he sees a never-ending procession of providers claiming to solve enduring human problems with expensive, complicated, isolated, digital solutions.

Mick believes we’d do better to start with people and their communities, with their strengths and how they can work together to make things better for themselves. Communities like Seacroft in east Leeds, where the LS14 Trust asked a simple question: “What would happen if we spent a whole year eating together as a community?”

“You can have the healthiest greens on your plate, but if you eat in isolation every day this might not always be good for your long-term wellbeing.” – LS14 Trust video

A couple of weeks ago I was privileged to be on a Leeds Digital Festival panel with Mick, Howard Bradley from the LS14 Trust, and Roz Davies from the Good Things Foundation. The event was organised by Victoria Betton from m-Habitat, who has also written up her impressions of the event.

On the agenda, I was there to be the “digital” voice in the conversation as a counterpoint to Mick and Howard’s advocacy of asset-based community development (ABCD). But I also accepted the slot on the panel to listen and learn, because I’ve long had a hunch that ABCD contains much that could improve my practice.

While we digital designers talk a good talk about focusing on people, I can’t help thinking our processes are still too often tilted in favour of electric wok solutions, and too rarely towards things like eating together.

In my contribution to the event, I offered what I hope was a critical description of the principles of a human-centred design process, as set out in the international standard ISO 9241-210:2010. I talked about the good things we always try to maintain:

  1. an explicit understanding of users, tasks and environments
  2. users involved throughout design and development
  3. design driven by user-centred evaluation
  4. a process that is iterative
  5. addressing the whole user experience
  6. multidisciplinary skills and perspectives

There need not be a gulf between human-centred design and ABCD, but often, by sins of omission, there is.

The problems start with the deficit-based way we often talk about “user needs.”

Inherently, a user need is a deficit, a thing a user lacks, a gap that we service providers claim to fill with our special expertise. Human nature makes it so easy for us to slip into electric wok thinking: this person is hungry; we make electric woks; what they need is an electric wok.

When the user protests that she never eats stir fry, many of our community double down on this deficit-thinking, by asserting that “people don’t know what they need.” The Henry Ford quote about a faster horse is trotted out, or something about how Apple don’t do user research (He never said it; they do.)

Human-centred design theory emphasises that we don’t take people’s stated desires at face value. We say no to that market research staple, the focus group. Instead, we uncover latent needs using ergonomic and ethnographic observations of actual behaviour (“Saturday, 1:27am: Participant orders takeaway chicken chow mein.”)

Rushed or done badly, such approaches render the research participant little more than a lab rat. The experimental subject’s only stake in the transaction is a shopping voucher to thank them for an hour of bemusement that they’ll never get back.

Empathy is essential in any human-centred design process. The trouble is, we often get it muddled up with sympathy.

When we see someone in pain, or with problems, or less fortunate than ourselves, our instinct is to help them. That’s a brilliant human thing. It’s mark of a civilised society that we have a safety net, no questions asked, to pick up a person when they’re knocked off their bike or floored by acute illness.

Cartoon man on stretcher
Still from ‘Your Very Good Health’ – Central Office of Information, 1948

Once the initial crisis has passed, however, sympathy must give way to a fuller understanding of the person and their capacity to recover. True empathy means feeling their hopes for the future, the things that make them resilient, knowing which activity they’ll enjoy the most to rebuild wasted muscles.

The factors that make someone strong are so personal and so varied that they are often forgotten in the focus on what’s commonly wrong. And in the name of equality, “not everyone has capacity” becomes a reason to ignore the assets of those who do. It’s then only a short step from fixing the problem to fixing the person, applying the faulty logic that if we are well, then making them more like us will make them well too. True empathy takes people as they are, not as we wish them to be.

It doesn’t have to be this way, but human-centred design has become, by default, individualistic.

There is a rich heritage of more social strains of service design and participatory design. In recent years, these have been drowned out by digital user experiences, where the context of use is invariably a person alone at a computer or on their personal mobile device.

Reacting against the phoney seance of the focus group, we prefer one-to-one usability sessions and depth interviews. To drive out ambiguity, we write user stories in the singular: “As a user, I want to… so that…”.

When we over-rely on these methods, we miss the plethora of relationships beyond the individual user and service provider. When we think about inclusion and accessibility, we fail even to ask users whether they consider it more “independent” to complete a task alone with assistance from a service provider, or by sharing it with a family member or friend.

Finally, as a questioner at the event pointed out, the language around this stuff has always been problematic.

We borrow the words of marketing “activation” as if people were machines waiting to be switched on. I work in a portfolio dedicated to “empowering people”, but who are we to give power in the first place? How about “stop disempowering people!”

In what direction will the new partnership of the professions and the 'consumers' work to carry out their purpose of meeting the medical needs of the people adequately, everywhere?
Extract from 1948 booklet ‘The New National Health Service’

I take consolation from the fact that a 1948 leaflet on the new National Health Service places the word ‘consumers’ of healthcare in scare quotes, as if our founders knew the word was unsatisfactory, and that sooner of later someone would come along with a better term. 70 years later, we’re still working on that.

I ended my discussion by posing two related questions:

  1. How might we move beyond purely transactional models of provider and consumer to more fluid configurations of actors, in which all contribute to and take from the service according to their needs, wants and abilities?
  2. How might we (especially those of us charged with making digital services at national scale) recognise that service is co-created and co-produced in communities, and provide platforms for those communities to discover, express and meet their own needs?

We can stick to our principles of human-centred design, but we need to broaden their interpretation.

ABCD reminds us to consider user assets at least as much as user needs.

User research should include everyone as equals, helping them to beneficially articulate things they do know at some level, but have not yet consciously considered. Only then can they become active participants in the co-design of solutions that suit them.

Asset mapping is a common research activity in the ABCD world, but Mick from the council is very clear: the asset maps aren’t for him, they’re for the community, to realise what they already have. And when they’re made in a participatory way, the assets they surface are very different from the usual libraries and sports centres that turn up on maps made by the service providers.

I was recently challenged about user needs in a learning context, where people literally “don’t know what they don’t know”. Yet learners do know many other things that are highly relevant to the design of their learning, such as what they know already, how they will fit learning into their everyday lives, and what they hope to achieve with their new knowledge and skills.

The whole user experience is situated at least as much in places and communities as in individuals, devices and service providers.

Beyond the place-based work of community development, there are some promising developments in the digital world.

The always insightful Cassie Robinson at Doteveryone is thinking with Citizens Advice about collective action:

Collective action is a strand of work we’re committing more time to over the coming months at Doteveryone, discovering other opportunities and contexts where collective action can play a role in scrutiny, accountability and influencing change. As part of this work we are also looking to civil society organisations to take a role in empowering the public and their audiences to take collective action in directing the impacts of technology on our lives.

Projects by If’s new report with the Open Data Institute considers some of the many instances when organisations deal with data about multiple people:

Services that allow data portability need to consider social relationships to ensure they are respectful of people’s rights. It’s also important that services don’t make assumptions about how groups make decisions about moving data: instead, they need to allow people the time, space and awareness to work things out for themselves.

Users must be involved throughout design and development in more than one way:

  • as participants in user research specified by the Government Digital Service
  • as senior stakeholders such as patient leaders in some NHS organisations
  • as fully fledged members of a multidisciplinary team, for example by bringing experts by experience onto Care Quality Commission inspections.

While many organisations employ people in one of these modes, very few yet combine all three. This means false conflicts are set up. User researchers complain that consultations are conducted with “proxy users” instead of the actual people who will use a service. The most committed service users, with much to contribute, can be told their experience disqualifies them because “they know too much”. In truth, we need them all!

If we want fewer electric woks in our future, we’d better stay open to unexpected outcomes.

Howard described compellingly how the LS14 Trust works to “hold spaces” where people can explore and create at their own pace – “laptop in one hand, cup of tea in the other”. They start conversations on people’s own terms, asking “what do you want to change?”

As a question from Victoria highlighted, we must always be aware of power imbalances in these spaces. People will be inhibited from contributing fully if they feel they should say what the most powerful people in the room want to hear, or if, on the basis of their past experiences, they don’t believe their participation will really change anything.

And Mick shared a set of questions that ABCD practitioners use to check the impact of their interventions:

  • What will be enhanced?
  • What will be restored?
  • What will be replaced?
  • What might this mutate into?

A great set of questions to ask when designing almost anything.

Now that’s what I call doing not talking

Doteveryone CEO Rachel Coldicutt’s Medium post, ‘What if tech conferences don’t matter that much?’ landed just as we were wrapping up Leeds GovJam 2017. Here’s me posing awkwardly with the awesome volunteer team who made this year’s event happen…

DAGw6ZBU0AAqpUE.jpg
Leeds GovJam volunteer team: Sharon, Tim, Emma, Kathryn, Tero, Lisa, Matt, Angela, Marie and Si

Every one of these people could, and should, stand on a platform and drop pearls of wisdom to an attentive audience. I would definitely pay to go to that conference.

But what the volunteer team did this week was more valuable than that: they modelled, coached, and mentored. They created a platform for others to perform.

IMG_20170517_123457 (1).jpg
Leeds GovJam 2017 teams at work

Our jammers made many things:

  • 7 prototypes, tested and iterated multiple times within 48 hours
  • new connections across the public, private and third sectors
  • commitments to keep on working differently…

DAH0n3DWsAAfSCu.jpg

At the jams, we have a hashtag for that — #doingnottalking.

In shared light: why making thing visible makes things better

“In Elizabethan amphitheatres, like the 1599 Globe Theatre, performances took place in ‘shared light’. Under such conditions, actors and audiences would be able to see each other… This attention to a key original playing condition of Shakespeare’s theatre enables the actors to play ‘with’ rather than ‘to’ or ‘at’ audiences. Actors therefore develop their ability to give and take focus using voice, gesture and movement.” — Emma Rice to Step Down From London’s Shakespeare’s Globe, Playbill, Oct 25, 2016

Some rights reserved - Phil Glockner
Phil Glockner, the Original Starbucks

Early, too early, one morning I blunder into a railway station Starbucks for a coffee and croissant to take onto the train. I’m the only customer. I place my order and shuffle along to the end of the counter where the barista will hand down my drink.

What happens next in the customer experience is critically important. We know that Starbucks knows this too, because of a leaked 2007 memo from chairman Howard Schultz, in which he bemoaned the commoditisation of his brand:

“For example, when we went to automatic espresso machines, we solved a major problem in terms of speed of service and efficiency. At the same time, we overlooked the fact that we would remove much of the romance and theatre that was in play with the use of the La Marzocca machines. This specific decision became even more damaging when the height of the machines, which are now in thousands of stores, blocked the visual sight line the customer previously had to watch the drink being made, and for the intimate experience with the barista.”

As I said, it was early, much too early for an intimate experience with a barista. And in any case, the barista was still learning the ropes. I guess first thing on a shift, when there’s one customer and no queue, is a great time for some coaching from the supervisor. This is what I heard him say:

“You have 23 seconds for the milk… Oh, and relax. You can’t concentrate when you’re stressed.”

23 seconds! That’s what removed the romance from my coffee.

“You must either make a tool of the creature, or a man of him. You cannot make both. Men were not intended to work with the accuracy of tools, to be precise and perfect in all their actions. If you will have that precision out of them, and make their fingers measure degrees like cog-wheels, and their arms strike curves like compasses, you must unhumanise them.” John Ruskin, The Nature of Gothic

Some things in this carefully commodified service experience were never meant to be seen by the customer. When they do burst into view, it feels wrong, uncanny.

In this post I want to explore the reasons for that uncanniness, and how we might play with it to develop new service opportunities. Is it really so obvious what should and should not be visible to the user? What’s the impact on users when a component slips out of sight? And how might we make service better by keeping more things, more visible for longer?

The line of visibility

Some rights reserved - JP Swizzlespokes
JP Swizzlespokes, Experience Design wk05 #whiteboard

The line of visibility is a well-known concept in the fields of customer experience management and service design. To use, like Howard Schultz, a theatrical metaphor, it divides the service blueprint into front-stage activities seen by the customer, and back-stage ones unseen by the customer but nonetheless essential to the delivery of the service.

In the coffee shop:

  • Front-stage: the theatre and romance of taking the order, writing the customer’s name on a cup, grinding the beans, making the coffee, presenting the coffee to the customer
  • Back-stage: the operational efficiency of managing rosters, training staff, timing operations, replenishing stock, and so on.

At first glance, the allocation of activities to front or back-stage appears uncontroversial. In reality, it is much murkier, and deserves more critical attention:

  • A restaurant might make a show of fresh food preparation with an open kitchen on full view to the diners, but still have a room behind the scenes for the freezers and dishwashers.
  • Recently, after returning a hire car, I was given a lift by a new member of staff. The conversation we had about the rental company’s graduate scheme made me warm to the company and more likely to return.
  • Much has been written about the 8 simple words on the underside of the machine on which I’m typing this now: ‘Designed by Apple in California. Assembled in China‘.

Visibility and the value chain

I’ve been thinking about visibility in the context of whole value chain maps. In his mapping technique, Simon Wardley arranges components from the most visible user needs at the top to the unseen at the bottom:

screen2bshot2b2015-02-022bat2b20-25-38
A value chain – wardleymaps.com

In this interpretation, visibility is said to recede as we traverse the network – the more “hops” away from the customer, the less it needs to concern them. But is that really true?

Invisible things can have very visible effects. Amazon’s recommendation engine is deeply buried in the company’s infrastructure, yet customers experience its insights and biases every time they use the site.

Visible things may get up to all sorts of unseen activities. What if that camera or video recorder in the corner is participating in a distributed denial of service attack right now?

Invisibility and commodification

barcoded wood

Why is it that some things naturally seem to merit visibility while others have to hide themselves from view?

I think it has to do with commodification. To turn something into a commodity is to take it out of its context, to make it fungible so that it can be substituted, traded and transferred. In an example by the philosopher Andrew Feenberg:

a tree is cut down and stripped of its branches and bark to be cut into lumber. All its connections to other elements of nature except those relevant to its place in construction are eliminated.

This is what people are doing when they commit metaphorical sleights of hand such as “data is the new oil“. They take something that has deep meaning to an individual and, by aggregation, transform it into something that can be traded without further challenge or debate.

The logic of commodification prohibits the end user from interest in, or influence over, anything but the surface-level components. Before we know it, any breach of the line of visibility feels illegitimate. From Fairtrade foodstuffs to the employment rights of Uber drivers, demands to deepen visibility into the supply chain come to be seen as “political” incursions in the supposedly rational domains of technological production and economics.

Consider the much-maligned EU cookie directive.

Unregulated, the behemoths of the attention economy would place all their tracking of users below the line of visibility. “Users don’t need to know about that stuff,” they’d say. “It’s technical detail. Nothing to worry about. Move along now.” The Jobsterbedunners might even hold up web users’ continued browsing of sites in such compromised circumstances as some kind of “revealed preference” for covert tracking.

But people who care about privacy have a different opinion on where the line should be drawn. Their only option is a “political” intervention to drag the publicity-shy cookie blinking over the line of visibility. Now Europe’s internautes can take back control, every time they visit a website. Say what you like about the implementation, but we Brits will miss those privacy protections when they’re gone.

Shared light

What if there was another way to realise value? One that didn’t depend on enclosing the value chain by making it opaque to end users?

To Feenberg, decontextualisation is “primary instrumentalisation” the first part of a two-step process:

The primary instrumentalisation initiates the process of world making by de-worlding its objects in order to reveal affordances. It tears them out of their original contexts and exposes them to analysis and manipulation while positioning the technical subject for distanced control…

But the story doesn’t end there. There’s a crucial, secondary step where visibility has to be re-established:

At the secondary level, technical objects are integrated with each other as the basis of a way of life. The primary level simplifies objects for incorporation into a device, while the secondary level integrates the simplified objects to a social environment.

Through this secondary instrumentalisation, this resource integration, users tell us what they want technology to be. Think, for example, of the camera-phone as a concept worn smooth by countless buying and use decision over the course of a decade. This part of the value creation process cannot happen in strategy and planning; it can only happen in use.

Premature commodification would close down such possibilities just when we ought to be keeping our options open. Co-creation, on the other hand, places the service user, the service designer, and the service provider on the same side – and all of us play in all those positions at one time or another.

We maximise value when the interests of all the actors are aligned, when asymmetries of knowledge between them are reduced. To borrow another controversial theatrical analogy, co-creation flourishes in “shared light” when actors and audiences can see each other equally.

  • Not only do we see the coffee being made, we see the staff being trained.
  • We are no longer passive recipients of the recommendation algorithm, we can understand why and how it behaves.

Some service design patterns

Here are just some of the patterns that play with the line of visibility. By making things visible, they make things better.

Seeing over the next hill: We meet much of the most valuable service when facing a change or challenge for the first time. But unless we know what to expect, it’s hard for us to make decisions in our best interests, or to trust others seeking to support us. Deliver service so that people can always see over the next hill, so they know what to expect, what good looks like, and who they can trust to help them along the journey.

Provenance: People can take reflective pride in where their things come from – and be repulsed by a supply chain’s dirty secrets. Design like they’re watching. Document the journey and make it part of the service. My Fairphone may have been a little pricier than an equivalent smartphone, but it comes with a story of fair materials, good working conditions, reuse and recycling.

Individualisation: Service is intrinsically full of variation. When we treat its delivery like factory mass production, we make it inflexible, unresponsive, and ultimately destructive of value. Anticipate variation, embrace it and celebrate it. This will likely means fewer targets and processes, more self-organising, empowered teams. Be like homecare organisation Buurtzorg, which prioritises “humanity over bureaucracy” and “maximises patients’ independence through training in self-care and creation of networks of neighbourhood resources.”

A last word from actor-network theorist Michel Callon in his afterword to Feenberg’s ‘Between Reason and Experience’:

“Keeping the future open by refraining from making irrevocable decisions that one could eventually regret, requires vigilance, reflection, and sagacity at all times. Politics, as the art of preserving the possibility of choices and debate on those choices, is therefore at the heart of technological dynamics.”

So we think we’re a user-centred, agile team…

BqKr2cXIAAEE1S8

Yeah, we’re user-centred!

  1. Who are the people most present when our service is delivered?
  2. Where are they, physically and emotionally, at that moment?
  3. What could each of them put into our service, and what could they get out of it?
  4. What are we doing to better answer questions 1 to 3 this week?
  5. How are those people taking part in the design and development of our service this week?
  6. How is our service changing because of their participation?

Yeah, we’re an agile team!

  1. Who decides what our team should work on next?
  2. Who chooses how our team should do our work?
  3. What competencies does it take to do our work? Do we have them all?
  4. How are we accountable as a single team for the work we do?
  5. How are we making our work visible to ourselves and others this week?
  6. How are we inspecting progress and making things better this week?

(Because solutions don’t scale, questions do.)

Most of government is mostly service design most of the time. Discuss.

Without exception, everyone I meet in the public sector wants to help make their service better. Most of them are in some way frustrated. The domain is massive and the activities disjointed. People engaged in any given service – from users and frontline workers down to managers and policymakers – can go for months on end without coming into contact with each other. On the rare occasions they do meet, they generally do so with mutual incomprehension.

This is not exclusively a government problem. I know from my time as a product manager in the private sector that a similar malaise affects all big organisations. But when it happens in government the impact of poor service is graver. This is service delivered with the authority of the state. As users we cannot take our government custom elsewhere. Neither can public service providers cherry pick their customers like the private sector does.

Whether we realise or not, most of government is mostly service design most of the time. If we fail to acknowledge this, we’re doomed to short change our citizens and fall short of our policy goals. But when we wake up to the potential, we find proven tools and techniques for designing service. Applying them can and should be everyone’s business.

We only have to look at definitions of “government” and “service design” to find a naturally good fit.

Exhibit A: The Institute for Government’s Whitehall Monitor summarises the business of government under the following headings:

  • the resources available to government (ministers, money, civil servants)
  • how government manages them (through arm’s-length bodies or contracting), what it does with them (passing legislation, answering requests for information) and how it measures what it does (major projects, permanent secretary objectives), and
  • what impact all of that has in the real world and how the public and international studies rate government effectiveness.

… which is nice because, Exhibit B, the Service Design Network defines its members’ practice thus:

“Service design is the activity of planning and organizing people, infrastructure, communication and material components of a service in order to improve its quality and the interaction between service provider and customers.”

We may call it many things, but service design happens all the time at every level of the government stack. The problem is that when done unconsciously it’s just not very good. All of the following contain random acts of design by default…

  • Users work their way around complex government processes, even if it means hiring costly experts like lawyers or accountants to do it for them.
  • Frontline staff hack the process just so they can serve their customers better. Visit any contact centre to see tattered papers, sticky notes by screens, Dymo-labelled folders and trays put in place to expedite information from one part of the process to another.
  • Good managers manage with ingenuity to sort out shift patterns, holiday rotas and flexible working so that their people can do their best work – in spite of policies and processes that treat workers more as resources than as human beings.
  • Entire, organisations-within-organisations accrete with baroque titles such as “change management” to drive through discontinuous re-structures that fracture working relationships and frustrate any long-term organisational learning.
  • “Policy” is a Platonic conception perceived to exist on a higher plane where users are always rational, processes run smoothly and every day is a sunny one. By the time we descend to the grubby depths of “implementation” it’s already too late.
  • Our democracy itself still runs on rails laid in Victorian times, as if the population were barely literate onlookers and the parties the houses of a minor public school on a bad-tempered match day.

Everywhere there are gaps. With their 1985 Gaps Model of Service Quality, Parasuraman et al. posited 5 types of them.

gaps model

All 5 gaps are endemic in public service. Design shouldn’t just be used to paper over them: it can eliminate them altogether. As Tom Loosemore said in his Code for America talk last year describing the Government Digital Service’s approach:

“We don’t talk about policy and implementation or policy and then delivery. We don’t think of them as two separate things. Even thinking how you fix the gap is a category error. What we are doing here collectively, with policy people in the room, is digital service design.”

(Disclosure: It’s my privilege to work as a contractor for GDS, though like everything on this blog I write this in a personal capacity.)

Let’s look at those gaps again.

  1. Gap between what customers expect and what managers think they expect. We seek and expose user insights, not just at the start or end of the process, but throughout. There’s good evidence that everyone on the team should spend at least 2 hours every 6 weeks observing primary, qualitative research. How about we make that a prerequisite for Permanent Secretaries, council CEOs, and everyone else not in direct day-to-day contact with service users?
  2. Gap between management perception and service specification. Even when we understand what users need, we have to get better at translating that insight into a vision of the service. We can use powerful formats such as user stories to tie requirements back to users and their goals. The best specifications of all can be real working prototypes. Making prototypes is easier than ever.
  3. Gap between specification and delivery. Alpha and beta versions are what we use to close this gap. They help us understand the ins and outs of delivery even as we refine our designs.
  4. Gap between promise to customers and what’s actually delivered. Ever been sold a Tesla only to find it’s a Sinclair C5? In the words of this tweet, “how could we get Britain voting on prototypes rather than promises?”
  5. Gap between what customers expect of service and how they actually perceive it. This yawning chasm is the cumulative effect of gaps 1 to 4. It is also the main driver of disappointment and distrust in public services. One bad experience loops back round and poisons our expectations of future interactions with government – a downward spiral that we need to disrupt.

How can we make government better? By accepting that first and foremost everyone’s a designer, and that we all need to develop a design thinking sensibility.

Besides relentless people-centricity – intellectually and empathetically understanding users, tasks and environments – service design practice has some distinctive characteristics:

  • Service design is visual. This doesn’t mean you have to be great at drawing – but it does demand working with more than words. When we draw pictures and diagrams we engage a different part of our brains and spot things we would miss through written specification alone. Making those assets visible can feel scary at first. That’s worth it though, because they change the conversation into something much more constructive than any amount of finessing verbal positions and semantics.
  • Service design is multidisciplinary. ISO 9241-210, the international standard for human-centred design acknowledges that no one discipline has a monopoly on design. Rather, “the design team includes multidisciplinary skills and perspectives.” This is a challenge to a silo-ed way of working, but small “two-pizza” teams in startups and internet giants like Amazon and Google prove it can be done.
  • Service design is holistic and integrative. Good designers of any stripe look at the big picture – what is the user need? what is the policy intent? – but they don’t stop there. They also dive down into the details and forge novel combinations of components. They hold multiple, potentially contradictory, strands in tension, zooming in and out between the reasons for doing something and the details of delivery that will make it succeed.
  • Service design is iterative. Whenever I read an account of Apple’s development process I am struck by the number of versions and iterations their products and services go through. They create and test many ideas before narrowing down on a handful to develop further. Just at the point when lesser companies would settle and launch they throw all the cards in the air and create yet more new combinations.
  • Finally, service design treats time as a material. There’s a place for thinking and working fast, and one for being slow and considered. A food bank user needs assistance before the next mealtime while a retiree of 60 needs to think what money they might need aged 100. Big service providers get stuck too easily in the middle of Stewart Brand’s pace layers. Service design helps them to be more supple.

Pace Layering

Don’t believe me? Try this stuff out for yourself. In June I’m taking part in the Global GovJam. It’s not a ‘designers’ event, just people designing together. For 48 hours we break down silos between local and national government, the NHS and social care, public, private and third sectors. We challenge people to communicate their ideas through doing, not talking. They make prototypes and take them out to potential users when they still feel incomplete. And the buzz as people realise how much they can achieve in so little time is amazing. Come and join us in Leeds or in dozens of other cities around the world.