Transcript - Paul's speech at ThoughtWorks Live on 'Simpler, Clearer and Faster Government Services' (March 2016)
Paul Shetler: My name is Paul Shetler. I am new to government. I’ve been in government here for like about the last seven months or so. Before that I worked for government in the UK for about 18 months, on the digital side, and we were doing a lot of really interesting stuff there, which I hope to talk about a little bit today, so lessons we learned. Obviously, we did some stuff with ThoughtWorks as well, so a little plug for our hosts. And before that, I spent a lot of time working in Internet and primarily in the financial services sector since 1993.
So I’ve been doing a lot of different things, and I’m hoping that this will be of interest. So the idea is yeah, ‘Simpler, clearer, faster government,’ things which you don’t usually associate with government, and yet there is this thing called the Digital Transformation Office, which now reports to our Prime Minister, who feels very strongly about these matters. And we’re trying to make it so that, when people use government services in Australia, they can actually get stuff done. And people might say, ‘Well, gee, why is there this thing called the DTO? What is this thing?,’ right? Because we are - probably - I’m not sure about the exact numbers, so I might be exaggerating just a little bit, but I think we’re the smallest government agency in Australia at the Commonwealth level. We’re about 50 people, and we’re split between two offices, one in Canberra, and on in Sydney. Sydney, we got about 18 people, Canberra we got about 30, something like that. We started out about a hundred when I joined, and we brought those numbers way down, because we wanted to make sure we were just digital specialists, and we’re now starting to ramp that back up.
But we have this huge, huge, huge remit. And the reason we were formed was because of that top number there [gesturing at slide]. One in eight Australians over the age of 14, every month, are going to be trying to access government services, and over half are going to have a significant problem when they do that. So you’ve got to keep in mind - I mean, I’m sure we all know this - that, when people access government services, they’re not going online to order a book, or to get an Uber, or to, you know, get some take out dinner, or anything like that. It’s typically because they need to do something that only government allows them to do. Whether that’s starting a business, or something to do with a benefit, or their health - typically very very serious things, which is why they’re in the hands of the state. And yet, over half are going to have serious problems getting that done. Now, if were a private sector firm with those kinds of numbers, well, we’d probably go out of business, wouldn’t we, right? So that’s why we exist. We exist because those numbers are unacceptable, and they have to change. That’s why we exist.
And we think if we fix that, in addition to having happier users, and in addition to being happier ourselves, because we’re also users of government services - we’re people too, after all - that we could also probably save an awful lot of money. Because at least my personal thesis is that part of the high cost of government services is the high cost of doing things incredibly poorly. It’s all the manual processes, all the paper, it’s all the remediation, it’s all the broken links between systems, it’s all the broken links between different aspects of the service that somebody’s trying to use, and it’s - in some cases, as well - it’s the failure demand, which is driven from a digital channel even. So we know we have some digital channels that have such high error rates that we have to have call centres, right? Think about that. Just think about that. I’ll let that drop in there for a bit. And we know those things cost a lot of money. So we also think that, on top of having happier users, and being happier people, and, of course, we’re in government because we actually want to do something for people - that’s why we’re public servants - so we would also be happier professionally, we can also save the government money. It’s a really good thing.
So I think one of those things, I think before we go too deeply into how we might do that, I think it’s a really really good idea to clear up some basic misconceptions. And these are misconceptions that people inside government have, but I think also not only in government. I mean, when I was in the mid-nineties, and I was working for Reublic National Bank of New York, and we were opening up our e-business subsidiary, which I was heading up, you know, we had our executives saying, ‘Wow, wouldn’t it be great! Won’t it be amazing, Paul! We’re going to be able to have a website, and people will be able to interact with us, right? Wow! It’ll be social. People can talk to their banker.’ And people thought that was an amazing thing. People thought people actually wanted to talk to a banker. And sometimes even now you hear, ‘People want to interact with a brand, right?’ People don’t want to interact with a brand, and they don’t want to interact with a banker, and they don’t want to engage with government agencies. It’s the last thing in the world people want to do. People just want to get shit done. That’s what people want to do. They just want to get things done. And typically, when it comes to government at least, the things they want to get done are much bigger than the little interaction they may be having with any particular government agency. So somebody might be interacting with the Tax Office, or with Department of Industry, or with the licensing authority at a state level, or something like that. And so the person in the agency might say, ‘Yeah! They want to interact with me. They want to engage with me!’ Sort of like the blind man with the elephant. Actually, the person wants to open up a business, might want to open up a business because they want to have a good life, or all the many reasons you can have for opening up a business. They probably just don’t really want to get that particular permit, or have that particular interaction with the Tax Office. So it’s really important that, when we think about why people’s experience of services are so poor, let’s think about what people are actually trying to do in the first place. Again, a lot of the conversation here today is about technology. And technology might be an enabler. Technology, in our view, is only part of the solution, and most of it comes down to service design.
Another thing I’d like to get out of the way before we go too much deeper is - I really hear this a lot - and it’s very annoying - that, you know, on the one hand, ‘Wow, Paul, you know, everything you’re saying makes sense, but what we do is just so incredibly big, right? So big, so big and so incredibly complicated. Well, you know, if it was a little bit simpler, a little bit smaller, Paul, then what you’re saying would make sense. But actually, it’s just enormous, you know.’ What people need to think about when they say that is, when you look at the total number of transactions, even for very large agencies like the Tax Office or like the DHS, things like that, compare that to a small-to-medium-sized bank, and you’d be about right, in terms of total transactions. They’re low. They do not compare to even an hour on the NASDAQ, or an hour on the Swift Network where I used to work, or anything like that. The transaction volumes are low, and the complexity is really not any more complex - I’m not saying Amazon’s not complex - but it’s not more complex than what Amazon has to deal with. In fact, it’s probably less. And much of that complexity is created by ourselves. After all, it’s government who creates the Administrative Orders that divide things up in particular ways, that creates the policies, that creates the legislation. So these are all things which are within our remit to change.
So what does that mean? So today, if you, say, wanted to start a business, or if you wanted to come into Australia like I did? So I’m obviously not from here. You don’t have - particularly when you’re accessing a government service online - you don’t really have a picture in your head of what government wants you to do. You don’t really know where to start, and you don’t really know where you’re going to end, you don’t know what the logical sequence of steps is, you have no real idea of whether you’re being compliant or not - it is fairly complicated, stuff is spread out amongst multiple different agencies. And so it becomes really complicated. It becomes kind of scary. It’s one of the reasons - another reason why the idea of engaging with an agency probably is not a very good one, because often times government appears capricious, because you don’t actually understand what government wants you to do.
And it’s not just me saying this. You know, we had - when we built our prototype for gov.au, we did a huge amount of ethnographic research with users, and what people were telling us was that they’re worried. They don’t really know what they’re supposed to do, they’re worried about making mistakes, they do mistakes as well, and often times it becomes so overwhelming that what they wind up doing is that they just pay for people to deal with the complexity of government for them. So if you want to think about whether a service is well-designed or badly-designed, you can think about whether you’re willing to pay somebody to deal with it so you don’t have to. That’s one way of considering the quality of service design. And here’s a great quote from a guy that we spoke with. This guy’s a serial entrepreneur. He’s opened many many businesses, and we were talking to him about, ‘What is the experience like?’ And he was saying, ‘Well, if you can afford it, you know, pay an expert to deal with government, because it’ll bury you, and it’ll distract you from your own business.’ And so what we - let’s roll this back a little bit. So often times, we will hear the phrase, ‘Channel shift,’ as one of the things you want to do in government, or other businesses will talk about channel shift, and when I was at Ministry of Justice, obviously we were looking at shifting people from channels like telephony or coming into courthouses or things like that. But one of the other channels that I was personally very interested in was shifting people away from another channel, which is called ‘Solicitors and lawyers.' Again, the system is so complicated you have to go to a lawyer to figure out how to do those basic things like getting a lasting power of attorney. And I think the same thing here, right? Australia, I think we have the highest number of independent financial advisors per capita. It might have something to do with the complexity of the services. I don’t think it’s that Australians are trying to be like incredible financial wizards. There might be other reasons for that. So our view is, we have to do a lot better, absolutely have to do a lot better. Australia’s going to compete on the world stage, in an economy that’s increasingly where goods and services are bought and consumed and delivered online, we have to do a lot better. Government should be an exemplar.
And we think right now we’re in a good - a very good point in time to be able to do this, because people have been talking about this kind of stuff for a very very very long time, but the economics have changed. So you heard about fast cycles, and small teams, and multidisciplinarity, and cell structures, and autonomy, and all this kind of stuff. And that flies right in the face of the way most large organisations work. So it’s not just government, pretty much any brownfield organisation is having to deal with these kinds of things. They can be banks, they can be insurance firms, they can be brokerages, they can be universities, they can be other institutions of higher learning, they can be law firms, they can be government or whatever. Pretty much every organisation has accreted all kinds of heavy governance, which, of course, fly right in the face of autonomous teams, and flies right in the face of self-direction, or tight feedback loops. All those things have been put there because the IT required to support their business in the first place used to be incredibly expensive. You know, to put in a new system that was going to serve customers would probably require perhaps a data centre, a number of servers, so on and so forth, huge capital expenditures. So of course you’re going to wrap that up, when you have to make this huge, upfront huge upfront investment, without really having done the user research, or really knowing exactly what all the problems are you’re trying to solve. You’re going to wrap it around huge layers of governance. Which are going to stop all of your agility, all of your ability to be Lean, all of your ability to be very responsive, all of your ability to continuously deploy things. Like, everything very tightly wrapped up. Something comes along: cloud, which means I can now get all of my IT on-demand, everything becomes an operational expenditure, I get rid of - pretty much get rid of the idea of CapEx, turn it into OpEx - and I can now take a lot of those resources that I had spent, and all the organisational effort I had spent in terms of trying to manage risks through boards and committees, and I can focus it on the one hand around service design - which I couldn’t do before really, because it was just too expensive, it was viewed as a luxury - and, on the other hand, I can use the risk mitigation which is already built into Agile and Lean techniques to lighten the organisational load of governance. So that’s why we think we’re in a great spot. And I was asked to talk a little about technology, that’s about all you’re going to hear today.
So there are - I think I mentioned - the digital estate of the Australian Government is really complicated. There are at least count - and I know this number is low, right? This is not the most recent number. At least count, it was 1,524 websites at the federal level, 4-5,000 at the state level, and I don’t know how many at the local level. That’s what I mean - that’s what I’m trying to get at when I say people don’t really know where to look. You can ask Google, ‘Where do I, you know, how do I start a business? or whatever,’ and Google will give you a steaming heap of links. Without organising them in a particular way, with no particular user journey. It is complicated, it is messy, and whenever government has a new idea, it pops up a new website. And there are many CMS vendors who encourage us to do so. So we thought, you know, maybe radical idea here, maybe - maybe we could make things easier if we took those 1,524, and we reduced them to one, that that might make a difference, right? Just consolidating information in one place might make a difference. And we think it would, we think even just having things - we think just that one thing, applying a common look and feel - because, again, those 1,524 different websites also, the only thing they have in common is that they have nothing in common in terms of navigation, colour, or anything else. It is a brand new user experience every single time, it’s an exciting thing, right? We do think that to do this - we have to go beyond that. Conway’s Law, we don’t really just want to replicate the structures of government inside the one website. And so we think again, again, people don’t want to interact with your agency, they want to get something done, so what is that thing that they want to get done? Let’s organise content, because most people who are trying to do something complicated with government, like opening a business, actually still most of them are not serial entrepreneurs. Often times, in fact in many cases, it’s people over the age of 50, 55, who are opening up new businesses. They’ve not done this before. So how can we lead them through that so that they don’t have go through something like that.
So we had some ideas, and we came up with this idea called ‘Transitions’. So, over the course of nine weeks, we were - at the end of last year - we created a prototype for a number of Ministers, showing how we could start reorganising services such that they met what the user was trying to do, the actual user needs, instead of replicating government structure. And here’s this [gesturing at slide]. We have this idea of this thing called a ‘Transition,’ and it’s sort of similar to the idea that people have sort of life events. Our view is that that’s not quite the right word, but it’s similar enough an idea. And the basic idea is that we organise - we organise large chunks of content around what people are trying to do, we ask people a series of questions like, ‘What is the name of the company? What kind of company is it? Where’s it going to be located? What kind of things will you be selling?’ And so on and so forth. And that comes back with a checklist. So we ask a few simple questions, we then get a checklist of activities, that checklist of activities contains links to existing services that are already out there, or could be new services custom-built for this. But the point is, we don’t require government to redo everything so we can start organising things properly. And we lead them through that step-by-step. And those services can be at the Federal level, those services can be at the State level, those services can be at the Local level, because most important things that people do actually do cross agency and jurisdictional boundaries, and people shouldn’t have to have a degree in constitutional law to do things like open up a business. And those are those transformed services, existing transactional services and so on over there [gesturing at slide]. It’s actually an extremely simple pattern, and we’re hoping to be able to release it to the public over the next couple weeks, so that anybody can look at it and tell us - I mean, give us your feedback, and say, you know, ‘Does this work? Does this not work? How should we think about this slightly differently?’ But, thus far, when we’ve been showing it to users, and again, we’ve been getting pretty good feedback. People have been telling us that they think it’s going to work. And an example would be this person [gesturing at the slide] - these are real people, by the way - who we have spoken to earlier, and told us again about the difficulties that he had had particularly in terms of bringing people into the country. And if anything, you know, this would take a huge amount of call centre inquiries. Again, you know, failure demand, it would get rid of a lot of that failure demand. And there’s a lot less chance that they’d have to call or visit the service centre, less need for an immigration agent - very good thing. So we think we’re onto something here with this, and that’s how we plan on redesigning government services.
So now in the course of doing this though - that sounds really easy, I guess, in some ways. It’s not. But we’re asking agencies to do three things they have never done before. And this is, I guess, what makes digital different - aside from the fact that things are being consumed and delivered over the Internet. The first one is, we’re asking them to start with user needs, focusing on what the user is trying to do, or, ‘What do they need to do?’ Not what our own internal structures are, not necessarily putting what our own policy is first, or anything like that. But focus on what the actual needs of the user are, and how that traverses your organisation. Then, we’re asking them to deliver the very smallest improvement, or thing, that they possibly can, and to do so very quickly, so really really quickly, so human timescales, not geological timescales. The timescales, you know - because these things happen in people’s lives. It’s really important. But once we’ve got them out there, to iterate them in real time, based on what we’re learning from users. So the whole idea of continuous delivery is an integral part of what we do. Integrated with user research, get it out there, and then keep on constantly improving it. And one of the ways you do that is by building services in an agile way - so we’re Scrum, primarily, Scrumban, basically - with multidisciplinary teams. So all the teams inside DTO, and all the teams who want to pass a Digital Service Standard Assessment, they will build their services with multidisciplinary teams. They will have a product manager, they will have a delivery manager, they will have a user researcher, an interaction or service designer, a number of developers, a web ops engineer, and probably a tech arch in there. They may have a business analyst and some other people from outside as well. But they are two-pizza-size: between eight and ten people, and they are truly multidisciplinary. And that’s just how we work. It’s the standard cell structure.
And all services that are being developed by DTO, and by the Federal Government, at this point, have to pass the Digital Service Standard. This might look familiar to those who are actually familiar with the Digital Service Standard in the UK, and there’s a reason for that, because this was shamelessly lifted at first from the UK Standard. The UK standard initially had 26 points. When I came in, we released an Alpha of this - this is now in Beta - we released an Alpha of the Standard, and we brought it down to like 18. We’ve since done further rethinking, we’ve consolidated, simplified and tried to focus it more, and brought it down to 14 points. These 14 points: for a service to go Live, all of these have to be met. And that includes making it open source, that includes starting with user needs, all these points. But the most important is point number 14, which is to make sure that the service is simple enough that users can succeed for the very first time, without any help: no IFA required, no lawyer required, nobody to hold your hand required. You should be able to do it yourself, right? And if you can’t - if the user can’t - then we won’t release it, we won’t allow it to go Live. So we’re really proud of this. And actually just earlier today, I was here discussing some of the work that we’re doing with some people from the City of Melbourne, and their digital team now has adopted the same Standard. And so we’re starting to see this Digital Service Standard adopted across government, and I’ve also heard that a couple of private sector firms are doing the exact same thing. We think this is a really good way to judge the quality of your development processes, as well as what you’re actually delivering.
I’ll talk another little bit on methodology, how we actually work. So we have this thing called the ‘Digital Transformation Programme,’ a very, very imaginative name. And it’s all about transforming our services in a very, very stepwise manner, again, to be compliant with the standard, but also to minimise our risk at every single step along the way, and to make sure that we do start with user needs. The very very first step is a standard Discovery. And at first when we were doing these things in the UK, we’d say, ‘Well, what’s in Discovery?’ And it was like, ‘Well, that’s when you discover things,’ and we weren’t, you know, terribly, terribly prescriptive. We’re realising to kick this stuff off in Australia, we need a lot more directive in terms of what we’re expecting from Discovery, what we’re expecting from Alpha, what we’re expecting from Beta. We’ve already published guidance on Discovery, we’ve published guidance on Alpha. But for Discovery, we’re really insisting on a very strong one. So a full understanding of user needs: what is the user trying to do at the time that they’re using your service? The overriding objective they’ve got going on in their head, but also, what are the other interactions, what are the other things happening in their lives? What are their interactions with government, or with other people, trying to get this thing resolved? What is the point of the policy intent behind the various things that the government is offering to let them do this? And what are all the different technology constraints? Now, at the end of that, you should have a pretty good understanding of the problem. You should have no understanding of the solution. If you think you have an idea of the solution from this, you’re all wrong, and all you’ve been doing is validating your own preconceptions. Go back and start again. What you need to do is understand the problem. Alpha is when we start to test out ideas on what a solution might be, right? That’s where we start testing our hypotheses about different kinds of prototypes: paper prototypes, prototypes in code, and so on and so forth. We see what works, we see what doesn’t, and this is where we start trying to figure out what the MVP might actually be. Again, very heavily focused on user research, but here we’re actually building things. Everything we build here can, and probably should be, thrown out once it’s been built. It shouldn’t be viewed as just a precursor to Beta. This is something which is testing still. We’re just trying to understanding. Beta is where we actually have to find an MVP. We talked about that smallest thing you can do. And we release it out to the public. It’s production, it’s real, people are using it to solve their problems. Live is when the Beta is used to decommission the existing channels, the existing services. Some of that might be a while, because sometimes decommissioning things takes a long time. But we’ve timeboxed all this as well. So everything here we’re talking about being done within 20 weeks, which is a bit generous, I know, but for government, that’s pretty fast, and we’re hoping to cut that number down, right? So that’s how we work. And again, that’s sorted of modeled after the UK, but put a bit more stringency around it.
And it’s delivered in these things called digital Delivery Hubs, very imaginative name. We have one in Sydney, we have one in Canberra. They are essentially co-working spaces. So at DTO, we like to say, we are not - people sometimes refer to DTO as the start-up, we’re like, ‘No, we’re not the start-up. But what we are is an incubator.’ We actually do think of ourselves as having the classical function of an incubator: we provide a co-working space, we have standards on how work should be done, we provide space to other government teams to come work with us to deliver their digital services, we provide them with coaching, so we actually have what’s called a digital transformation manager working in each team to, on the one hand, sort of push them to meet the Standard, on the other hand to see all the different things that are happening organisationally that will block them from doing the right thing. Because when you’re trying to design from the outside in to meet user needs, that’s not usually what most organisations have grown up culturally or - that’s not what most organisations really are ready to do, in many ways. There’s lots of internal impediments to doing that, whether they’re procurement, HR, all kinds of other things standing in the way. So the transformation manager helps to remove those. To some extent, what we’re saying to guys is, ‘Here’s a minefield, go run through it, find all the obstacles, or the bombs, and defuse them’. We also provide the teams there with researchers, with designers, with developers, and so on and so forth. The teams in these hubs go back to their departments, and they form the nuclei of new digital organisations in each government agency, so that is sort of the people part of the digital transformation we’re talking about.
And this is where I give the shameless pitch. So we are always on the look-out for really great digital talent. And I know sometimes people don’t like using the word ‘Great talent,’ or ‘Really top people,’ I’m shameless about it. I am looking for the very best product managers, the very best delivery managers, the very best developers, technical architects, designers and so on and so forth. So if there are people here, or if you know people here, who really want to make a difference, and who really want to work in a truly user-centred, truly Lean start-up, in a truly Agile way, do let us know, because we do think we are changing the way that the Australian Government works, and we want anyone who is interested in that to join us. And at that, I will leave it alone. I’m told I have to get off the stage.