An Enterprise Government Approach: How We Meet the Challenges of Tomorrow

“We need the network, in order to exist in a digital age, to just work.”

Every day, organizations are challenged to anticipate, adapt and keep pace with the digital landscape. Canadians expect to use the latest digital technology to receive the services they need anytime, anywhere and from any device on a trusted and secure platform. This presentation provides an overview of how an enterprise approach will help equip Government of Canada’s departments to meet the challenges of tomorrow.

Continue the conversation with Paul on Twitter and see the slides from the presentation here.

Please see the full transcript of the talk below.

Paul Glover: [00:00:00] Good afternoon, everybody. Really, really happy to be here. I’m going to take a few minutes, as was said, and plug a little bit about Shared Services, what we’re doing, how we’re going to try and move forward. So I’ll share with you our plan. And then I’m going to share with you what I think is probably the biggest risk to that plan, that if we don’t get this aspect right, is going to be really the rate-limiting factor for how successful we will be collectively.

So I’ve got a few slides I want to share with you. So the first thing is, Shared Services. We called this 3.0, and I always like to start with why the 3.0. So if you think back in time, Shared Services has just turned eight years old, and the version one was essentially creating the department that nobody wanted. And that’s frankly, what it was. It was 42 [00:01:00] departments and we got all their toys and took them from them. It was not a fun time to me. And in addition, it was standing up a department, not something that’s done every day, a lot of different processes. So it was a pretty intensive time to stand the place up, figure out how to bring everybody back together and prove that this could work. So we spent a bit of time doing that. 

The second version, or the 2.0 version, was when we were trying to make sure that we could actually make it work, run it, service it. So it was about that service culture. It was about really trying to make sure we understood the client needs, we’re delivering on them and trying to make sure that things were moving forward.

And then I was appointed last January, so I’m about nine months in, and that’s where the three comes from. I’m their third president. It’s really that complicated. So that’s what we’re doing. And why is it going to be a different approach? And I think it’s part of that natural maturing that we’ve seen. [00:02:00] Could we create the department? Could we actually service the town? How well is that working? And the reaction is- frankly, not bad, but not great. When I talked to a lot of the deputies in the 42 departments, treasury board ministers, there’s still a sense that it takes too long, it costs too much, it’s a little cumbersome or a lot cumbersome.

But there have been some real successes there. The one thing that I will tell you- I am struck with since arriving at Shared Services- is the name is wrong. There’s no “shared” in Shared Services. And I think that’s one of the things we’re really going to have to come back to, because if we’re going to talk about enterprise, there needs to be more shared services. And I think a large part of the problems that we’re seeing today is the amount of complexity in the ecosystems that we support. There’s over 50 networks. There’s hundreds of data centers and we’re closing them and we’re still finding them eight years later. So the complexity is [00:03:00] gigantic.

And so anytime there’s an issue, it’s always unique. It’s always special. And it takes a lot of expertise and understanding of those environments. There aren’t enough standards, aren’t enough simple ways of operating. So it’s going to be important for us to really take a look at that as we move forward. And we’re a little bit racing the clock there. Some of the stuff is aging out faster than we can break it. We literally have hundreds of incidents every day, unplanned outages every single day. So there’s a lot going on in that ecosystem that we’re going to have to work to improve. So we’re trying to improve it. We’re in break-fix mode in a lot of places, and those really aren’t sustainable approaches to doing things. And because of that complexity, which should be simple, is actually really complex, beause it’s got to be over planned, got to really understand that environment, it takes a ton of time to actually figure out how to deliver those solutions. So we’ve got to [00:04:00] simplify, we’ve got to streamline in order to be successful moving forward, and I think that that will help us to serve people better moving forward. 

So, how are we going to do that? And that’s a lot about what this slide is all about. Thankfully, that’s not a pop, it won’t explode when I try to open it. So three priorities in the center of that circle. The first thing that I would say is we’ve got to really focus on the network and security. We all know about security, doubts are increasing. We have to make sure that we’re continuing to involve our defense mechanisms, hearing a lot about zero trust networks, end to end encryption. How are we going to make sure that we continue to have a robust approach to addressing security? Because we are literally under constant attack. The number of things that we are working- to prevent bad things from happening every day, that is staggering and growing each and every day. But in addition to that, we have to make sure that the network becomes something we don’t [00:05:00] talk about anymore. It just works. We walk into an office building and we know there’s going to be heat. Few degrees hot, few degrees cold, but basically it’s within an operating range and we can make it work. If it’s too hot or too cold, you can’t work. And you plug an outlet in, you know, there’s going to be electricity there, it works. We need the network, in order to exist in a digital age, to just work. And it’s a lot of effort now to make it work the way we want. And anytime we get it big enough, it’s automatically filled, we’re behind on Wi-Fi. We need to get to a state where we stop thinking about the network, it’s there like heat, like electricity in a room. It is that important in a digital age to how we operate and we are not approaching it like it’s that kind of basic infrastructure. And so we have to start thinking about how we make the network treated like a utility. It’s there, it’s operated and we behave like [00:06:00] it’s a utility.

So I like to tell the story about today, when we’re trying to address the network, we’d literally do it, not just building by building, but sometimes floor by floor in government complexes. And in some buildings, one of the larger complexes across the river, we’ve been to the same floor seven times. Different departments, different branches of the same department, different budget groups. Think about that. Now I’m going to tell you a true story about at my house. One of the sponsors here was deciding they were going to lay infrastructure down that street. They were going to put fiber cable, the whole street. They knocked on the door, told my wife “We’re digging up the first six feet. Very hard. Yeah. And it’s not yours, it’s ours. We’re laying fiber and we’ll put it back. You won’t notice. And by the way, from the six foot marker to your house, we’d like to dig that up and connect. If you don’t want us to, we won’t, but if you ask us to come back, you’re going to pay. [00:07:00] We don’t care if you’re a customer or not, we just want to get it done once. And all your neighbors have said yes. And it’s really expensive if you ask us back.” And so she did what all the neighbors did. And two days later, it was done. The whole street, all the houses. And I can tell you if that was us, the way we would behave, we’d dig that street up the number of times a house says “Hey, can we be a customer?” And in fact, we’d actually have to dig up some of the houses that we already put it in because no, we wouldn’t put the pipe big enough for the whole street that people would connect to, because nobody pays for more than you need. It’s just- it can’t be done in a government system. That’s killing us and we’re doing that over and over. We need to start thinking differently if we’re going to be successful. 

The second thing in that circle is collaboration tools. The world is changing. When Shared Services was created, our Act literally talked about email. My kids are like “Email? Email?” Let me do [00:08:00] phone and video conference and explain to them “You have to go to a special room to do a video conference.” And they look at me like I’m from Mars. Like my tablet: “I’ve got an open conversation with three friends, and poof, look, we’re all talking. You have to go to a room and you’re the government?” 

We have to get with the times. We have to make sure that the tools we are providing are relevant to the way people are operating and it is not email. It is not all of these individual stovepipes. There’s different ways of collaborating. I don’t think they’ve emailed a document, the way we are behaving and doing things is not the way the world is working, and we need to get to get with that pretty quickly. 

The top one is application health. So what are we going to do in that space? So the big push was for us to close data centers, close data centers, close data centers, and we were closing them down. I saw Ken here, the number is staggering. Hundreds a year, and we’re getting better and faster at it. 

[00:09:00] And I go back to what I said at the beginning: hundreds of incidents, unplanned incidents every day. So we were shutting data centers, but what was the impact that that was having? Not really the kind of impact we would want it to have, because the applications were still crap. They’re old, they’re breaking. And so we take a bad application and move it to a better data center. It still breaks. And so we had to shift focus from closing data centers to improving application and the health, reducing the risk, improving the reliability of that application. Now that means when they’re in bad data centers, power sucks, we can’t do the cooling right, they’re maxing out, we’re going to shut that data center down. But the point of doing that is not to shut the data center down. That’s not the outcome we should be measuring. How reliable are the applications we run as a government? And how often do they [00:10:00] break? And so we’re trying to shift the gears and we are going to continue to close data centers because a lot of them are at or past end of life and we need to get people into better spots.

So that’s the three priorities. How are we going to do this? Well, we’re going to put the users first. It’s not about us and the number of data centers we support. It’s the user in that application, how well it works. It’s that network that they don’t have to think about. It’s the collaboration tools that give them what they want. Put the user first. And that’s for federal public servants we serve. That also means Canadians, Canadian businesses that government departments serve. Put the user first. We want to start small and scale up. So one of the things- we’ve got a cloud offering, you know, there’s a lot of conversations about that here. We have taken an approach with what we call path finders, which is a little bit different than “someone’s going to do it and share lessons”. And I go back to that “we’re all a special flower”. And so, you know, the lessons that a department learned doesn’t [00:11:00] apply because they’re department X and I’m department Y and that’s great for them, but I’m different. It’s not going to work for us. So the idea is not to share lessons, but to lock them in as approaches, as standards in playbooks. If you go first, it’s going to hurt. It’s going to be a little bumpy. We’re not going to lie, but there’s a lot of people of expertise, and we want to learn from that. And we want to lock those lessons in as approaches, as ways of doing things so that other people follow and make better. And we will evolve it as we move forward. And that’s the approach. So to start small, learn, plan to fail, fix it, lock that fix in and scale. And repeat, and repeat. And that’s how we think because our biggest problem today is operating at speed and scale. It’s really, really hard to do with what we have to support. 

18,000 is, I think, treasury board’s [00:12:00] latest estimate and growing in the number of applications. It’s the scale of the stuff is gigantic. So we’re going to start small, we’re going to learn, we’re going to lock it in and just repeat that over and over so we can scale up. 

The other thing I think we’re going to need to do in order to do that at the other end is we’re going to really need to make sure that we’ve got the incentive structure right. And I think this is important as we move forward. It means that we’re really going to have to make sure that the financing models are correct for these sorts of things. 

An example, if you exist in one of our old, really bad data centers, it’s free, or core funded to run it for you. You move to cloud, you pay. So you have to find that that doesn’t make sense. Why is the same activity free for a department here and paid over there? Well, you say “Well, Paul, when you take it out of the data center, just give them their money back.” It’s not the way those work. That’s a fixed overhead cost. Until I close and get [00:13:00] everybody out of a multitenant data center, there’s no money to give back. So the incentives aren’t working for us. Back to that analogy about the network, and we go to the same building over and over, because that’s when people get the funding and that’s when they ask us to come in. The incentive structures aren’t working for either of us, so we’ve got to work to change that if we’re going to move forward. It means we’re going to have to do better with standards so that we have repeatable, reliable ways to do things. And it means we’re really going to have to take a look at how we deliver services. We take 600,000 tickets a year. It’s a lot of overhead. We take a lot of steps to deliver things and that’s on both of our sites. And so we have to look at how we can do that more efficiently. So that that’s kind of what we’re going to do, how we’re going to do it, and we’re pretty pleased with that, we think it’s really gonna make a difference. 

So I promised, I’d tell you, what do I think the biggest risk is to that plan and how we do it. So let’s talk about our current environment. This is a [00:14:00] very rudimentary graph. I apologize for it. I was trying to figure out how to show this to you, but it’s going to be fairly generic. So we have a user, whether that’s a citizen, a business, an internal public servant, they interface with us. Through a department, through a series of applications that then talks to a network that then talks to data in the computer to make something happen for them. And we do that over and over. Everybody sets special flower and everybody has their own unique interface to do very, very similar things. And I can tell you firsthand that that same user one wants to go back and tell another government department the same thing, or a different government department they’re probably interacting with a different interface.

The story I like to tell, because it’s very personal, is [00:15:00] when I lost my parents a few years ago. The number of letters I had to write to say the exact same thing to different departments at different levels of government to say “My mom and dad aren’t here anymore, can you stop sending the letters? Don’t send the checks.” It was ridiculous because I had to go to all these different interfaces that then talk to their own little network that talk to their own systems and heaven forbid they should talk to each other. That is what is killing us. And that is what will kill the plan that I just showed you at risk. And then you have more of these things and it just goes on and on. It’s a pretty generic diagram, there’s probably a lot of people in the room that will tell me what’s wrong with the arrows and how things work. It’s not the point. I’m not going to talk to you when I leave. I know it’s wrong. Okay. But the point I’m making is what should be shareable in our current environment is not shareable [00:16:00] and we are laser-focused on the departments we are in as IT professionals, that client, and making sure we deliver for that client what they need, that we’re not thinking outside ways enough, about who else is doing similar work and how do we make sure that we don’t do it over and over and over. 

And if you think about what I said, and it’s an estimate and it’s probably around 18,000 apps, if you want those to start talking to each other, “Oh, well let’s just build some nice little interfaces.” Well, let’s evergreen that crap, right? As one system changes- I run 21 email systems. One little change in one email system, got to test it with all the others to make sure it works. “Oh that? Oh, that changed something there. Okay. Well, let’s fix that now. We’ve got to test it all again.” It’s unsustainable. And this is what we’re not doing well enough as a community. What should be shareable needs to be [00:17:00] shareable and we have to make that happen. 

Now the problem as I see it is our track record in this stuff hasn’t been very good. If you look back at history, we said “well, email”. We didn’t take a commodity product, use it as a commodity product. We over-engineered it. When we have tried, our approach to enterprise has been one size fits all. So the analogy I would use is if, you know, that we’re all special snowflakes, think about that as a custom suit. Everybody has their own custom suit. Well, if we’re going to standardize, how are we going to do that with one size for everybody? It’s either really expensive, very cool fabric that can shrink or grow on demand, right. Or it’s never going to work. It’s going to be way too small for some people and way too big for others. What we need to think about is how we go from [00:18:00] custom suits for everybody, really timely, really expensive. They fit nice, we liked them, but they’re not sustainable with what we do. So how do we give you sizes so it’s good enough? So you can come in and quickly get what you need and get on? That’s what we need to do. And then when you really, really need a custom suit, because it really matters- not because you want it, but because it really matters- there’s a way to get it. 

So what we need to do to change that environment is make what should be shareable, shareable. Build it once and share it and make sure that the data talks to each other, that we can share the data. One source of truth, data lakes- we’ve got API stores, all the fancy buzzwords that we’re talking about have to allow us to find a way to identify what should be shared and make it shareable. And that isn’t something I’m going to do as [00:19:00] your Shared Services. That is not something the CIO shop is going to do at treasury board, because they’ve got this infinite source of wisdom. That is something we must do as a community. And if you look at other communities of practice, they do that. 

An accountant has a responsibility to the client, but they also have a responsibility to their profession. And they build out standards that makes sure they all operate the same way. There’s only so many ways to do petty cash. There’s only so many ways to comply with GAAP, generally accepted accounting principles, that create a behavior. And their responsibility as professionals is not just to the client, but to building, maintaining, and adhering to those standards. Are we operating that way as a community? And if we don’t, there’s a lot of money to be made satisfying individual special snowflakes, but that day is going to come too because it’s [00:20:00] not sustainable. And I can do all those things that I was talking about, back a slide, but it won’t ever reach its full potential if we can’t start to figure out what should be shared, make it shareable and then work to make that happen together. That’s going to be really, really important as it moves forward. 

So the other thing is back to user four, but not everything has to be the same everywhere. It’s back to that: not one size will fit all of us and we will need different suit sizes in the racks so you can come in and pick out different ones. I do not see, with the federal government that I service, and when we start thinking about what we’re doing now, where we’re moving forward, we just identified some big protective B-cloud contracts. We’re working with a number of provincial governments, municipal level governments who want to access that vehicle. We’re going to start sharing even beyond just with the federal public service because that makes [00:21:00] sense. And we’re able to do those sorts of things, right? So if we’re going to do that, then we’re going to have to identify what should be shared and what shouldn’t be shared. When it is unique and needs to be unique, and we have to get much better at doing that if we’re going to be able to succeed moving forward. So let’s make what should be shared shareable and what isn’t, and shouldn’t be, let’s make sure that we have a quick, easy way to deliver those things. But the key is making sure that we’re really able to differentiate those things.

So I think, how do we do that? And what have we been doing wrong now? Well, we operate in a Westminster system and that’s very vertical. It’s a cabinet, a caucus, the prime minister gives ministers priorities, ministers see people like me, deputies, and say, “do these things”. And that’s what we’re told to do when we go off and do it. But enterprise are not vertical. They’re horizontal. They cut across. And so it’s going to be important for us right now- we start [00:22:00] thinking about how we cluster things by department. 

So I was president of CFIA before I was president of Shared Services. When I left that job, I turned in all my devices to arrive at Shared Services to be given the same set of devices, because I live in a world where I deal with ministers and secret documents- and by the way, the functionality on those devices was a little different, but my job was exactly the same. I still accompanied ministers to cabinet, briefed them, saw security briefs- it was different. Why? Because the standards were set by departments. That’s stupid. We have to change our behavior. Ministers, deputies, people who work in a secret environment should have the same tools, regardless of the department they’re in. Right, health? I spent a lot of my career- if you looked at my bio at [00:23:00] Health, well, Health is not one type of employee. They have policy people doing deep thinking, they’re working on the Health Act, they have regulators, they have inspectors, they have scientists. We need to start thinking about the people and the functions. There’s an inspector who’s a transport inspector crawling through planes and trains, or a CFIA inspector walking fields or through a plant, need the same set of functionality, connectivity tools as an office worker or a policy analyst that’s collaborating with the world. Or a scientist who’s dealing with terabytes of data on a daily basis and collaborating with academia.

We need to start thinking about the experience and starting to build standards and personas, profiles to make that happen. And then let’s look out: who do we serve? We serve citizens. Back to my very personal example I went through. Okay. We also serve businesses as a federal government. We [00:24:00] regulate them. We impose licensing conditions on them. We tax them. And we have a special relationship with Indigenous. How are we servicing that? How are we making sure that we’re cultivating the right kind of environment for that to happen? And how do we think about servicing remote, rural communities? That’s a terrible picture. I don’t know why we picked a modern office building for rural communities, sorry. But it’s got to look different. Right? How we service the far North is going to be very different than how we service downtown Toronto. It’s just the cost of doing business is going to be different. We have to think about them differently as we move forward. And are there other ways to divvy this universe up so that we can start to think about what should be shareable? What are the standards we need and what is unique so that we start to cultivate and curate those things? And we have to do that together as we move forward. So that’s basically what I [00:25:00] wanted to, to share with you.

I think I got it about two minutes. The little light is still green, so I don’t know if anybody wants to call BS on anything I said. I’m happy to take a question, a reaction. But I do think we have a plan for what we’re going to do at Shared Services and it’ll work to a point, but in order for us, collectively, to realize the full potential, it means we all have to start working together. And I’ve talked a lot about the federal perspective on this stuff, but think about things like identity. Those are going to cross federal governments, and provincial governments, municipal, local governments. And there is opportunity for us to actually identify what should be shareable across the different levels of government. Think about those cloud contracts that I did in the security stuff. A small, small municipality isn’t going to be able to go in and do the security audit. It’s all the stuff we did with the experts and the specialists in that space. It’s only- I have to [00:26:00] share it with them or they can’t get there.

So, how are we going to figure out how to work together, better? Here, as a federal family, but also how are we going to do that with provinces, with territories, municipalities, to figure out from the user perspective? I didn’t just write letters to the federal government when my parents passed. I wrote them to the provincial government. I wrote them to the city. There is a shared experience we need to cultivate and curate in order to really do that. So enterprise needs to be custom-fit as it moves forward. 

Anybody? My work here is done. McGill, an academic. Oh-oh. 

Audience member: [00:26:40] “I think you can start with the digital IDs. Like, I’m kind of fed up when I go on any government site- I work for government. You have to fill in the same information every time.” 

Paul Glover: [00:26:57] So I completely agree [00:27:00] with you that identity- everybody will say is an obvious place to start. What I would say back to you though, as you said, “I should”. And I think it’s “we should”, working together. Find a way to make sure that we are collectively taking a common, standards-based approach to identity so that it is shareable. 

How many of you are working on applications right now? And it starts with that sign-in, and you’re saying “what can we do to make that shareable? Who’s already got something that I can use, and I don’t need to write that?” And let’s take that to the central agency and demand it is standard in that space. So I agree with you, it’s the obvious place to start, but we have to do these things together. Otherwise, somebody will set a standard and the backlash- I can tell you, if I set the standards, you’re going to tell me I did it wrong. It’s just physics. But if we work together [00:28:00] and if you tell me what you need, either way, I got to procure something, I’ve got to curate it, I’ve got to find a way to deliver it. I don’t really care what it is because it’s got to be better than 50 networks, 21 email systems, hundreds of outages every day, 600,000 tickets a year. So let’s do it together. 

Anyone else? Right behind you. 

Audience member: [00:28:24] “All right. So you talked a lot about, sort of different priorities, sorry, like different things that you want to do as Shared Services, but are you worried about being- doing so many things at the same time that you have issues with resources? So I guess, like, what would be your priorities in the next, say, five years or 10 years so that you can manage resources effectively? Thank you.” 

Paul Glover: [00:28:45] So that’s a great question. I think part of the problem is when things are this broke, there’s a lot of things to work on. And if you think about it back to the beginning, that version one, it was 42 departments coming together as one. We got the [00:29:00] best of breed and we had people who didn’t even know what they had. They were happy to hand it off. So we’ve got a very mixed bag of things, not just for Shared Services, but departments overall that we work with. So I think from a priority point of view, everything moving forward is dependent upon a network. Right, a cloud-first policy without a network that’s going to support workloads in the cloud is not going to be a great experience for us. The network, we have to start talking- stop talking about get it done and get it fixed. And we need to make sure that in a digital age, we’ve given the employees the tools they need. So it’s that collaboration and I do worry very much about the application health. We’ve got a lot of really old applications in departments, and just moving them to enterprise data centers, moving them to the cloud- there’ll be marginal benefits at best. We need to start re-engineering a lot of those things. So those would be the three priorities. 

I think collectively, though, we’re going to have to [00:30:00] work together and I don’t think in order to build the standards I’m talking about, we need huge heavy processes where everybody’s got to agree “Who’s got money now that’s building something?” And how do we ask them “While you’re doing that piece-back to identity, can you take a little bit of time, talk to some people and start to build out some standards?” You look- there was some pretty impressive companies- won’t name them- but they just set some rules. If you’re building something, it has to be shareable to anybody in the company. And any data you collect has to be accessible to anybody in the company. And if we start to do it that way, change our behavior, then it’s many of us working to build out these standards, to evergreen them, move them forward. 

But we can’t- you’re right, too much all over the place, we will just scatter. But that’s why it’s going to be important for us to work together, to say, what do we want shared and let’s work together to make that shared. 

I think I’m getting the hook. All right. Thank you very much. [00:31:00]