Service Design in Government 2024

In mid-September I attended Service Design in Government for the second time. I headed up early for the book launch of Platformland (a book I have already read and written about) and then was around for the 3 days.

Keynotes

Rachel Coldicutt

Rachel has been “mucking around on the internet” for 30 years, with work including an incredibly millennial artefact: digitally creating the room Dawson Creek's titular character. More recently she's the unusual person in policy who ‘makes things’. 

Some points from her talk:

• There are tensions between centralised ‘Mission Control boards’ (as leaked to the Times) compared to devolution.
• Risks of moving in parallel between hype, missions, and delivery. Some things like AI can live in the hype line - and these things get handed down in a magical way. More generally, technology isn’t free - every technology choice is choosing not to do something else - and FOMO is not a strategy. 
• It’s important to hold the difference between what a bureaucrat wants compared to people going about everyday life. Similarly, policy is often influenced by famous technologists who don’t understand policy
• Service design can be be the little things - the bins, the apps for schools - sometimes it’s better to make smaller things rather than just map all of the things
• We're also moving away from universals to fragmented publics (or at least being more aware of it, maybe it was always there) groups. This is combined with younger people expecting things to come to them via social media rather than looking for it. This could also apply to government as well as people.

Book reference: The Unaccountability Machine by Dan Davies “the regrettable tendency of complex systems to have opaque and volatile dynamics” 

Co-design - KA McKercher

I've been following KA's work from afar, and in what I think is a well known moment for book lovers, I lent my copy of Beyond Sticky Notes to someone 2 years ago and still have not got it back…

KA offered a healthy challenge to the room about terminology, reminding people that 'a mission is a place that kept Aboriginal people away from their kin', important given that much of KA's work referenced First Nations people in Australia.

They also reminded the room that co-design isn’t right for every project, we may not be skilful or open-hearted enough to lead, and we can have gaps between what we want to offer and what can be done in the constraints, and some places can be one of harm. What's more, people with lived experience becoming more discerning about the “Promises about co-design” as trust can be easily lost. Often if we don’t have a bureaucracy whisperer all stuff stalls

They spoke a lot about co-design rigidities (related to Alba villamil's 'Has design become too dogmatic?':

• maintaining a duty of care (paternalism is near enemy of care and to be allowed the dignity of risk)
• doing group work (individual work is good too - before and in-between is powerful. Sometimes people may not want to keep co-designing, they may be tired and happy to hand over for work and then check)
• doing co-design (can use other means too like consultation - consultation can still be good. If we’re not good at consultation, how can we do co-design? )
• getting work done (communities want to be listened to and have solidarity, as well as just producing. Allow a first session about solidarity, let people have goals - give spaces for processing and integration)
• furthering self-determination (don’t just throw things to the community, stay involved) 

Some other things I took from the talk:

• 'the maintenance of shared will' (like a daisy-chain) - any no breaks the chain
• queering work and “Writing into the wound” (Roxanne gay)
• some creative examples of doing things, like deep work generative usability testing and co-writing, which is showing someone a thing and the facilitator editing in real time based on comments, asking for feedback and then sharing to the others. These sorts of 'tactile and playful practices' can also help when co-designers are tired
• ending things- via Cassie Robison, talking about shifts, and seasons of participation - asking more times, and checking in at point of writing reports and asking people if they want to be thanked 

In a bit of meta-review, I also admired KA's approach to explicitly defining the relationship between presenter and audience. They both gave explicit instructions both verbally and on screen as to what images could be photographed 'there is a second problem of designers copying without attribution', and instructed people to chat about the talk later after they'd had a few hours to decompress!

Book and link references: Noel et al - Pluriversal Futures for Design Education

Government 360: Whitney Quesenbery

When I got to chat with Whitney I said "I met you in New Zealand back in the 00s". I later had a moment of panic that perhaps I'd got it wrong but checking my email shows that there was in fact a UPA (now UXPA) Auckland lunch meetup in November 2007 when she came along. It was fair for both of us to not remember though, she says she had red hair back then and I had (natural) black hair in a pixie cut!

Since then Whitney has spent over a decade at the Center for Civic Design. At the beginning 'officials didn't know how to say usability let alone spell it'.

What is civic design? I liked her explanation - creating and experience that invites everyone to participate by bridging the gap between the complexities of government and the needs of people. It also requires a sincere desire to make things better: Whitney often does not know the political persuasions of colleagues, merely they’re just trying to do a good job. “If you can’t understand something, you can’t trust it”. Similarly, different people worry about different things with forms - for some making it better was enough, also other officials were more worried about someone not getting a form and sued.

Notes:

• look for trusted messengers - doing a good project with people, who then can vouch for your work (departments copying others)
• the people who care about legislation and do close readings are often at both edges - low end with people who have experienced dysfunctional government and high end trying to unpick the details. The middle may not care.
• Making quality explicit: the team are working on a ballots and audits self-certification for excellence. League table are also helpful for things like rejected voter forms as no one wants to be near the bottom of the league table!
• watch out for accidental inconsistency. The team created consistent ‘mail, drop-off box, vote centre' materials that everyone used, but didn’t think about translation so the spanish versions were varied leading work later.
• If Elections 360 is connections to the community co-ordinated outreach, consistent voice, meaning better websites mean communities will use them; then  Government 360 is to surround yourself with many perspectives (include advocated and practical experts but don’t force consensus), get up close and interactive, extend your reach with collaborations, go where the people are  
• Vote.gov is “a single location for access to voter registration  no matter their background, language spoken, physical capabilities, or geography - the federal government can provide a fail-safe that has never before existed in this country. It has language provision is often localised based on statistics so people that move can lose access. The team worked on having consistent language and to support it - also looking at non-written languages and how to keep video updated. Text size and contrast is deliberately visible. The voting form used to have 2 pages and 52 pages to questions - now changed into a form with filter question about state, and then populates a print and send form (making fully digital is still being worked on). 
• work ahead looking at policies and infusing some thoughts relating to disabilities. This got 9 added. Some prioritisation (high/med/low), new or existing question and risk and reasons - worked as people wanted to improve. Whitney shared examples of dealing with a challenge of long list by doing concepts, and by starting from accessibility first realised that just saying the number of candidates helped
• Mix up backgrounds, explore together, work on ideas, not solutions. The team ran workshops making designers, officials and people with disabilities work together (with industrial design students as sketchers) - book was ’50 ideas for more accessible voting’  - creating a space for people who are either normally in a bubble or in fractious asking for things. 4 key ideas from the workshops became ‘LA voting system for all people’ ideas (partly as key person was there) lavote.net/vsap - also had idea of letting people do practice online, and then bring this pre-filled version as a pass onto the system. it also sped up the voting process and people just thought it had always been like that 
• I loved the phrase “Pre-bunking”: getting right information out early to stop needing to ‘debunk’ it
• Think about how to share bits of information – they have the 'bite, snack, meal, and also 'find, understand, use' - and create tools to help work with complex materials that can’t be changed (yet)
• Go where the people are - pop-up research in libraries, community colleges, and other place 
• Expand your view, bring more voices into your work, bridge the gaps between service and users, create relationships that foster trust and transparency, and support users in 360 civicdesign.org 
• Don’t ask for the big thing, ask for small steps and be OK with it not being amazingly implemented at first

Other notable mentions

I'd love to add more, and may do this later, but this is a start.

Designing truly accessible forms: Caroline Jarrett and me

Going through my notes, I went… why is there a big gap? Oh, I was running a workshop, oops.

Caroline ran an earlier version of this workshop at Ladies that UX. This time around, I added a few extra things relating to accessibility as well as QR codes for the activities.

It was great to work with such an experienced facilitator like Caroline, she picked up things like the groups being too big to enable a lot of chat meaning that needed to split them up!

Talks about games

One of my favourite talks of the conference was Jemima Higgins' Tax and Ladders. Faced with making stakeholders understand an unresearchable user group (fraudster), the method had to encourage creativity, incentivise rule-breaking, facilitate open but structured discussion. After thinking about roleplay before discarding it as too scary, she settled on games. She started with mapping journeys and using the ’snakes and ladders’ sense of tension provided with ‘choose your own adventure’ 

Making the game showed some gaps in the policy, and playing the game showed that none of the solutions worked and that the workaround would be outside the scope of the initial work. It also helped the stakeholders think in a different way, the team now have the board and use it as a training activity.

Some tips she had:

• Test the full game - make sure there are mechanics to keep things fun
• don’t overcomplicate it - try to shorten areas and keep it engaging. Doing it again Jemima might have done ‘choose your own adventure’ - and test existing mechanism  
• it can be best not to tell people beforehand, just call it a ‘structured walkthrough of the user journey’ 
• the value is in the playing, not the playback
• Games are hard to use and play - can get some buy in, other things like roleplay and legislative theatre can also work,
• Using game techniques like co-drawing can help, or if you have to use it use an existing game mechanic e.g. Policy Lab's systemic  

Interestingly other talks also mentioned games, such Anna Lay's on using AI for prescriptions. Her team made a Guess Who game to play with stakeholders and even made some data not work to show the need for good data.

Futuring and prototyping

There were a few talks about discoveries, futures and prototyping. It has made me realise that I think there's a missing conversation about using discoveries to actually explore other contexts and concepts… more on that maybe another time.