Enterprise Architecture: Bridging Strategy, Innovation, and Change with Tom Graves

Enterprise architecture (EA) has long been a critical function in organizations, often operating under the umbrella of IT leadership. However, the landscape is shifting. As organizations navigate rapid technological change, economic uncertainty, and global shifts—such as Brexit—the role of EA is being redefined. EA is no longer just about aligning IT with business; it is about shaping business strategy, enabling agility, and ensuring that systems, people, and processes work harmoniously.

In a recent conversation with Tom Graves, a seasoned EA thought leader, we explored the transformation of enterprise architecture and its broader implications.

Transcript

So, my name is Mike Walker. I'm with Gartner. I cover enterprise architecture and technology innovation. 

Today, I have with me Tom Graves, who is an EA thought leader who has been a prolific blogger for many, many years and a pusher of the envelope. So, welcome, Tom. Could you tell us a little bit about you, your company, your blog, and perhaps maybe a little snippet on your view of enterprise architecture? Sure. 

I've been working for rather too many years on how people learn skills is really my real focus. And enterprise architecture is the domain in which I'm mostly working in that, how people learn to make sense of change and then adapt to the change. So, we go down into the detail as well as just the big picture. 

The blog was about 1,200 articles on there, totaling about two and a half million words. So, it's been around with about 700 illustrations, graphics, I think. So, it's a bit. 

There's 11 books behind it as well. There's a bit of work. I see myself very much as a tool maker, maker of tools for how things work.

So, when I came into enterprise architecture, we started from a perspective, this is back in Australia post. We started with the assumption that any process could be implemented by any combination of people, machines, and IT. And I moved back to Britain about 10 years ago on the assumption that it would be a more mature market and discovered somewhat to my horror that it was only about IT and they didn't understand the idea of people. 

Interestingly, they didn't understand the idea of machines either. And one thing we're noticing now is, and this is where you're asking about how things are changing, the big shift is as the merging on the one side of business and IT with things like mobile, social, all those areas, and on the other side with IoT and Industry 4.0 and real-time information gathering from machinery. There's a much more of a merging where we have to understand how all of these domains work together, rather than simply saying it's only about IT. 

So, the other big shift that we're seeing and is beginning to happen now is that the enterprise architecture, which has always been kind of dumped underneath the CIO or sometimes the CTO, is actually moving out. Where we had it in Australia post was it reported directly to the executive as a whole. And that's starting to happen again, particularly as we now have the marketing CMO, Chief Marketing Officer, often having a larger proportion of the IT budget than the IT department. 

So, there's some very big shifts happening in the way that we understand what EA is about. And that came up a lot in your session on business-driven EA at the Gartner conference. Yeah, and so, you know, I agree wholeheartedly with you. 

You know, the transformation, you know, in general that EA has gone through. I can tell you from my personal experience, you know, back here in the U.S. in the banking world, you know, we suffered from a lot of the same challenges and dilemmas. Now, we see EA in a very unique position. 

You know, it's kind of, you know, we've been hoping for this, you know, that we'd get this opportunity, and now we have it. And it's, you know, guys, don't screw this up because this is our opportunity. But, you know, you know, the people is a component, the automation or the machinery, the technology is a component. 

But also, we're seeing that there are so many other aspects as well. And you were touching on this, you know, and I want to double click on a little bit here and something really topical to where you're at, you know, Brexit, the controversial, you know, event here. You know, those are some very political and economic issues. 

Now, for EAs, what do you think that means for enterprise architects? That event, what's that going to do for their job? Sure. That one event, well, everyone's strategy is just being thrown in the wind. And if we look at a lot of, I mean, I get a lot of my work with the military environment.

I started out in science and technology research in Australia. And I'm working with similar people over here. The military environment, the whole context is about dealing with change and about dealing with uncertainty. 

And there's this huge difference between the plan and the ability to do planning on the fly. It doesn't mean that we don't have a plan. We have the ability to do planning. 

And the same is true of architecture. Instead of having an architecture roadmap, we have to have the ability to architect on the fly, but always in context of the whole, always in context of the big picture. And the big picture has just changed quite radically.

Now, I have a standard way of describing a lot of this, which is to say that architecture is really about one very simple idea, which is that things work better when they work together on purpose. What we've just hit now is a situation where they don't work together. They're not working together at all. 

And in fact, across the country and across many other areas, we've got a really clear indication that things are not working together properly at all. Now, that's going to have massive roll-on effects. Just as a simple one, a lot of the banks are worried that the passporting system that is used for finance into Europe, which allows the finance system to trade directly into Europe without any change in regulation, suddenly that's about to be lost. 

And that's going to have some very severe effects on how finance operates, to the extent that London is likely to lose its position, or is at risk of losing its position as the financial centre of the world. It's going to move to a second tier level if it's not very careful. So, these are really major changes that's going to affect everyone. 

At the moment, it doesn't much change the IT, not until things start to move. Once they do start to move, every single regulation is going to change. Every single piece of governance is going to change.

It's going to have absolutely enormous effects, most of which people didn't think about until it actually started to hit home what it really is. So, the point here is that strategy is about choices. Enterprise architecture intersects with that. 

It's about how do things work together in relation to those choices. So, it becomes, if you like, orthogonal to it. It's about the choices that we are building, so that we're providing information into those choices, and we're taking information out of those choices. 

And there's an interaction that loop that has to happen with it. And under these circumstances, that has to happen pretty fast. Yeah, absolutely.

What I love about what you're studying here is you're not saying that EA owns the strategy. You're not saying EA owns innovation. No, absolutely not. 

They're orthogonal. They're orthogonal. And I'll take it to the next step, and I'll say, you know, EA is really here to help give you the information you need as a senior business decision maker to make a decision.

You know, those options that we've been talking about, you know, EAs are helping to facilitate that conversation, almost as if a management consultant would. So, you go into, you know, McKinsey or Booz Allen, and, you know, the EA has a core part of that skill set, and they have an opportunity to bring that vital information to their stakeholders, and more specifically, you know, their senior leaders. And what's important about this is it shifts the conversation from a, you know, let's align with the business, which doesn't make any sense, to, you know, we're figuring out, where is my business going? Because we're not aligning to the business. 

We are the business. We're all part of this thing. And yes, there are going to be implications from a regulatory perspective that's going to have to be implemented from a technology perspective. 

However, we've got to figure out, what is the blueprint for our company in the next two years after Brexit, if it actually happens, or whatever variant occurs? You know, what does that blueprint look like? What are the what-if scenarios that we need to plan for and mitigate risks around, right? So, I agree with you wholeheartedly there. Shifting just a little bit, Tom, you know, we just met up at the EA Summit, and, you know, as a Gartner person and as a speaker, sometimes I don't have the luxury of walking the floor, talking to as many people as I would like. You know, what would be interesting is, you know, what were some of your impressions of the EA Summit in general? You know, was it valuable for someone like you and what you do for the typical EA practitioner? What were some of your thoughts? Right. 

One of them was that the sheer number of people is much larger. If I go to an open group conference, for example, there'll be perhaps 200 attendees. I think you had, what, 500, 600? And also, most of them are directly in a business context. 

They're not just IT people. They're also, they're either very much engaged with IT in the business, or they're coming straight from the business. I talked to people from pharmaceuticals, for example, another one who's dealing with the vast changes that are happening in air traffic control over Europe, and across the whole world, as the entire system is due to change. 

I talked with people in a number of different industries. In fact, one of the problems I had was I didn't get to really many sessions at all, because I was too busy talking, if you put it, walking the floor. And that was probably, and I remember someone who said that the most valuable part of a conference is the bits between the sessions. 

I actually found myself entirely between the sessions, and got a lot of value out of that. The one conference, one session I really went to right from the start was actually yours. And that was really good to see how much you've helped enterprise architecture move and drive forward. 

There were other Gartner consultants of whom I'm not so sure, but to give an opposite extreme, one of my favorite Gartner consultants is Bart Packer guy down in Brisbane, I think, in Australia, who is probably one of the world's leaders on understanding the people side of enterprise architecture. He's a real thought leader. That's someone I very much look to and look up to, as one of the most forward thinkers I've ever come across in the space. 

So those people are there in Gartner. I met another Gartner colleague, who's just moved over from Volvo, literally the last couple of weeks. So he's very much finding his feet. 

And a couple of others as well that I messed up with and talked with. But the main point is that the quality and the caliber of the people there, particularly the participants as much as the speakers, is very high. And obviously being Gartner, the presentation quality is very high indeed. 

So we would expect that from a large corporate. The big shift that we're suspecting is likely to happen is that one of the practical issues you have in Gartner is that people tend to focus in single areas. And enterprise architecture and architectures in general, the whole point is that we bridge across multiple areas.

Taking, going back to strategy for a moment, we are one of the very few disciplines who goes not only across, but also all the way up and down. We can go and talk with frontline people in the way that a business consultant probably wouldn't. And that's a very big difference. 

So we can collect stuff from the frontline, from the floor, and bring it up into the strategy space in a way that just doesn't happen elsewhere in other disciplines. Another way of looking at this is that we're a bit like, we are essentially a quality environment where the quality is about how things work together. So from a business perspective, actually enterprise architecture is everyone's responsibility, not just those who call themselves architects. 

Our job is as facilitators, coordinators, translators between domains to make sure that architecture happens. That's something I saw quite a bit beginning to happen in the Gartner conference, probably more than I ever have done in an open group conference, for example. I've seen it. 

The one place where I have seen it is another funny little conference called Integrated EA, which is about enterprise architecture and the defense and high risk space. That's because the military and high risk people tend to be aware that things have to connect across a hole rather than hide out in IT alone. Yeah, it makes a ton of sense. 

I love that, and I promise this wasn't scripted at all, but the point that you brought up around EA being a mile wide and an inch deep, taking that viewpoint, and then also having to pivot off of that and also go deep when they need to. How we structured the EA summit over the past, it's been an evolution. It was very, very EA mechanics centric, best practices of this or that, but we really started to expand this over the past couple of years. 

Really, it's been about three years to focus on not just EA Prime, as I would call it, where it's just us navel gazing at the EA practice itself, but also looking at the other activities that EAs will need to understand and perform. Bringing in folks like Frank Bojendijk. You mentioned his keynote in your blog. 

He's in our information management research area, or Al from our IoT area. Lots of individuals from outside of the EA research area bringing them in. Then having myself and my partner in crime, a Gartner fellow named David Curley, we focus on technology innovation and what that means to the enterprise architect. 

That's the next big wave here. The status quo just isn't going to work. That's why I brought up Brexit as a good teaser here.

Yeah, we've got these technologies out there, but there's a reason for all this. There's an intent. We have to also understand the social, the economic, the political, climate change as an example. 

All of this impacts the blueprints in which we're going to create as enterprise architects. I'm glad you said that and pointed that out because that was a very deliberate thing we wanted to do. Frankly, it was a bit of a test. 

We didn't know how the audience was going to take it. Year over year, we're seeing an increase. It's a slower burn in Europe, but we're seeing more substantial growth in the US as far as the EA Summit, as far as attendees.

You walk the vendor floor, as an example. There weren't just EA tools there. We were inviting more and more folks in from other spaces to come and talk about innovation. 

Hopefully, next year, we can get companies like, I don't know, Liquid Hub or Spigot or what have you, the innovation management tool sets. Bring those folks in there. You're spot on.

There's another one. One thing that came up in your session, which was really important, was this refocus around purpose. One of the problems is that when it was thought of as being IT strategy, the strategy and the purpose had been defined entirely elsewhere.

Often, I must admit, I've literally heard a CEO once say, our strategy is last year plus 10%. In fact, our job on that session was to get his team to understand what a strategy actually was and what it looked like. By the way, just to come back to one comment you made earlier about enterprise architecture being a mile wide and an inch deep, it's actually not that really at all.

It's that we are a mile wide, a mile deep, but thin. We cover the full depth, but the real thing that we have to build is actually a social network. I'm thinking of Helen Mills at Australia Post, who's one of the best leaders I've ever come across. 

She got headhunted to go to one of the major banks and suddenly discovered that nothing worked at all. What she'd forgotten was that she'd spent 20 years in Australia Post building a huge social network. She knew everyone who mattered, everyone right from the ground level all the way up to the top across the entire space. 

Although she didn't know necessarily very much, she knew exactly who did know. That gave you the debt that you needed whenever you needed it. When she got to the bank, she finally realized that the reason it wasn't working was because she didn't have that network. 

She set out deliberately to build this network. I think it took her two years of what someone described as an awful lot of coffee to build that network. Once she built it, she got this network across the space where she knew exactly when someone said this crucial question. 

We've got three key questions we say as enterprise architects, which is, I don't know, but I know someone who does. It depends, and I know how to find out what it depends on. Just enough detail, and I actually know or can find out what is the right level of detail that's needed here.

Those three responses to questions are crucially important. They're actually foundational to the whole discipline. The idea that we deliver solutions, that isn't our job.

That's solution architects' job. Our job is to connect. Very much more at the enterprise architecture level, it's very much more about connecting people than about connecting technologies.

Absolutely. That's a fantastic point. I don't know if it was the session that I delivered at the beginning of the summit or another, but essentially, I talk about the notion of IQ and EQ, the intelligent quotient versus the emotionally intelligent quotient.

Again, this is my personal opinion. There isn't Gartner research behind this, necessarily. My equation for successful EA is 70% EQ, 30% IQ. 

IQs keep your job, but EQs don't keep your job. Like you said, this is all about bringing people along a journey. That's all it says.

We can use lots of fancy words, but at the end of the day, we're bringing the organization along. We're helping them think through their extremely complex problems and breaking it down in very simple terms so that we can make an informed decision. Also, keep in mind, these people that we keep saying, the business and these guys don't know what the hell they're talking about because I keep telling them all these great things, and they just don't understand. 

They don't want to understand. Guess what? They don't understand. They don't need to understand EA.

What they need to understand is how to increase shareholder value, or if you're in the public sector, how do I protect the citizens or best utilize tax dollars if I'm a nonprofit? How do I help solve the malaria problem in Africa? We need to shift the track to, okay, yes, EA is a means to an end. All those fancy tools we're using, that's a means to some end. Let's focus on the end and not the means here. 

I guess I don't know. Yeah, which leads into another point which you said about because we both mentioned tools at this point. At the Gartner Summit, I spent most of my time talking with vendors, toolset vendors, because to be blunt, we have a very serious problem that none of the so-called EA tools actually support us doing EA. 

They support a solution architect's job, but they don't help us doing an enterprise architect's job. There's a famous model that came up in the design space called the squiggle, where if you like, the front end of something is it goes crazy like that. When you first start, it goes absolutely crazy, and then it slowly settles out until it comes out almost sort of level and flat. 

Well, almost all of the EA tools that we have will only work at the point where everything's become stable, but at that point, we've already handed over to the solution architect. It doesn't actually help our job at all. Our job is to help what happens in the messy part, to deal with the translation. 

So, we need to know who is who. We need to know what their drivers are. We need to know what the technology drivers are. 

We need to know what's happening at the bottom because frankly, what's a piece of research I saw recently which said that 100% of people on the front line, that's the squiggle. That's right, yes. So, the people on the front line know 100% of the problems that are actually facing.

By the time it gets to the execs, they know 4% of the problems that are actually being faced at the front line. A major part of our job is to simplify, to bring things together, to get that information from the front line, and present it, and filter to the extent where we understand how it intersects with strategy, both present and probably future, and then build it towards, bring it into a form that very time-poor executives can understand. It's not that executives don't get it. 

It's that they've got so little time that a major part of our job is filtering, and we'd better understand strategy well enough to understand what to filter and how. The tools we have at the moment don't really help us with that. They help us doing nice diagrams, but that's the easy bit after everything's already been decided. 

I often describe those diagrams as decision records, whereas we need help with the sense-making and the decision-making before those things come out. So, what do you think are some of the key... So, what do you think are some of the top three capabilities that modern EA needs from our toolset bench? Well, the first one is, if you think of the kind of things that we need support on are things like, imagine a mix of Evernote, OneNote, a live whiteboard, audio recording, the ability to capture photographs from anywhere, the ability to capture conversations, links to emails. It's to maintain a thesaurus, not just a dictionary, but a thesaurus, because everyone has different definitions. 

We need to know who's holding which definition. We can't talk about, you know, breaking the silos and so on, because silos develop around a language. So, when we talk about finance, when we talk about HR, when we talk about compliance, when we talk about IT, when we talk about marketing, each one of them has a different language, and they've got different... Everything is different in each of those spaces. 

So, they develop naturally. The issue then is how to build bridges across. Things have to work together in the way that they are, rather than necessarily as we'd like them to be. 

So, the people are saying, just get rid of the silos, just do everything agile. It doesn't work. The same thing, for example, I know there's this whole thing about Gartner's bimodal IT, and I've been one of the people critiquing it. 

What you've got in bimodal IT is a very simple version of what's actually needed, which is governance of governance, where you are choosing the mode of governance dynamically according to what the thing is doing. So, for example, a lot of shadow IT, we actually want it to happen, because that's how we get development happening. But as soon as it starts becoming business critical, especially to another group, that's when we must have some other form of governance that starts to come in.

So, we've now got clear criteria. When we look at the Gartner concept of pace layering, that's another type of criterion that we need to apply to governance of governance. And if we start looking at it like that, that it's governance of governance rather than one or the other, it starts getting more usable. 

This whole thing about how things change over time, as things move from the edge towards what we might describe as a backbone, the core. And often you've got the classic heavy-duty stuff. The reason you run waterfall is not because it should do, but it's because you've got so many damn stakeholders that if you break it, you're going to have hell to pay. 

That's what waterfall is really for. Agile is when you can only use agile when you haven't got many stakeholders. Yeah, I think you're right. 

And I can tell you that, at least from a Gartner perspective with bimodal and pace layering, a lot of these concepts, I'll just call them that for right now. Sure. What criteria is that? Pace layering, I look at that, and maybe this is an oversimplistic representation. 

I might have Gartner colleagues come after me after this. But I look at this as just a way to segment a portfolio. This could be a portfolio of projects. 

This could be a portfolio of strategies. This could be a portfolio of opportunities or ideas. You can put anything in that portfolio. 

But what's nice about pace layering is very simple concepts, run, grow, or transform. And we put policy around how we're going to handle at each criteria what we will do, the type of investment we will make, the people, the amount of people, the amount of investment, and being very – and I repeated this several times in a lot of my sessions, which was we have to be deliberate on what we're doing and put in a lot of emphasis around being deliberate. We don't want to just strategize for strategy's sake because our strategy might just be – it could be a poor strategy, and there's lots of companies that say, look, we're willing to roll the dice and just be incrementally 10% better. 

Yeah, we know these disruptive startups are coming in our space, but we're going to choose to go down this path. Bimodal is similar. I think oftentimes – and especially that we see in the press – bimodal gets constructed as – or construed as kind of the end-all, be-all. 

That's the blueprint for IT. And that's the furthest thing, at least from what my team is saying, from a strategy and EA perspective. What we're saying is this is answering how we're going to do something. 

It's not answering why we're doing the things we're doing or what we will do. It's, hey, look, how we're going to implement this is we're going to have two speeds. One speed is, shoot, we've got to react to Brexit fast. 

So, mode two, get going, guys, because we don't have time to go through a waterfall, and we're willing to take the risks. So, again, that's where the deliberate behavior comes in. What we do beforehand – I'm going to cut across you here gently. 

What we do beforehand, the role of architecture is to identify what has to change slowly, what you cannot change fast, and what you can, and therefore understand the risks in relation to the fact that you can't change it fast. Understand the risk in that we can change it too fast. A lot of, for example, taking the example right now of LinkedIn, they're randomly changing their – what feels like randomly changing their comment system to a point where it makes

(This file is longer than 30 minutes. Go Unlimited at TurboScribe.ai to transcribe files up to 10 hours long.)