DEV Community: Victor Leung

What Corporate Finance Taught Me

Victor Leung — Sun, 05 Apr 2026 05:00:43 +0000

Corporate finance and enterprise architecture seem like entirely different disciplines. One deals with capital allocation and shareholder value; the other with systems, capabilities, and technology strategy. But peel back the surface, and they share the same underlying logic: how do you make good decisions when resources are scarce, uncertainty is high, and the people making decisions don't always have the same interests as the people bearing the consequences? Here are four ideas from corporate finance that I think every enterprise architect should internalise.

The first is the concept of the opportunity cost of capital. In corporate finance, this is the return shareholders forgo when they let the firm invest their money rather than taking it back. If a firm's investments can't beat that hurdle rate, the stock price falls. It's not enough to generate returns, you have to generate better returns than the next best alternative.

Enterprise architects face the same logic every time they evaluate a technology investment. The question isn't "does this system deliver value?" It's "does this system deliver more value than what we'd get if we spent that budget elsewhere?" A bespoke integration platform might work fine in isolation. But if the same investment could fund a commercial off-the-shelf capability that reduces operational risk and accelerates two other programmes, the bespoke build destroys value, even if it technically succeeds. We need hurdle rates in architecture. Not just TCO comparisons, but genuine opportunity cost thinking: what does this architecture decision preclude?

The second idea is the principal-agent problem. Shareholders delegate decisions to managers. When the agents' interests diverge from the principals', whether through misaligned incentives, information asymmetry, or simply self-interest, you get agency costs. Value leaks.

Enterprise architecture has its own principal-agent structure, and it's one we rarely name explicitly. The business delegates technology decisions to engineering and IT teams. Product managers optimise for their own delivery metrics. Infrastructure teams optimise for stability. Vendors optimise for contract renewal. Each is acting rationally from their own perspective, but the aggregate result can be an architecture that no one actually chose, a fragmented landscape of systems that each made local sense but collectively impose enormous integration cost and strategic risk. Architecture review boards, capability-based roadmaps, and fitness functions aren't bureaucratic overhead. They're mechanisms to reduce agency costs and align local decisions with enterprise value.

The third idea is that a safe dollar is worth more than a risky dollar. Investors demand a premium for bearing risk. A certain cash flow is worth more than an uncertain one of the same expected value. This is why discount rates go up with risk, and why optionality has value.

In architecture terms: a modular, replaceable system is worth more than a tightly coupled one of equivalent capability, even if the tightly coupled system is cheaper today. The modular system preserves optionality. It lets the firm adapt when vendor strategies shift, when regulatory requirements change, or when a better technology emerges. The monolith forecloses those options. The hidden cost of tight coupling is rarely captured in a business case, but it's very real. Every time I've seen a firm locked into a legacy platform beyond its useful life, the root cause is usually the same: an architecture decision made years earlier that looked locally optimal but didn't price in the cost of future inflexibility.

The fourth idea is that smart investment decisions create more value than smart financing decisions. The real value creation happens in what you invest in, not in how you fund it.

The enterprise architecture parallel is the distinction between architecture strategy and architecture execution. I've seen firms spend enormous energy on delivery optimisation, agile ceremonies, platform engineering, DevOps tooling, while the underlying investment thesis is weak. They're financing well but investing badly. If your architecture is evolving a capability the business no longer needs, no amount of execution excellence will recover that value. Conversely, a well-chosen architectural bet, the right platform, the right capability boundary, the right integration pattern, creates compounding returns even if the execution is imperfect. Before we optimise how we build and run systems, we need to be right about which capabilities actually matter.

Taken together, these four ideas point toward something larger. Enterprise architecture is ultimately a discipline of resource allocation under uncertainty, which puts it closer to corporate finance than most practitioners realise. The decisions we make about which capabilities to build, which platforms to standardise on, and which technical debts to carry have exactly the same structure as capital allocation decisions, with similarly irreversible consequences when we get them wrong.

The opportunity cost of a poor architecture decision isn't just the remediation cost. It's everything the firm could have built instead.

Five Years in the Lion City: Wherever the Heart Finds Peace, That Is Home

Victor Leung — Sat, 28 Mar 2026 14:27:22 +0000

New Asia Life Monthly Magazine, March 2026 Issue is published! 🎉

This issue features my contribution to the “Journey Around the Globe” column, sharing the daily life and reflections of alumni living abroad. Hope you’ll find a moment to read it 🙏

📖 Digital edition: https://online.fliphtml5.com/xhegu/sgku/#p=21

Leaving Hong Kong to work in Singapore, nearly five years have passed in the blink of an eye. Every time I return to Changi Airport and walk up to the automated immigration gate, two words appear on the screen: “Welcome Home.” Politics aside, let’s keep things light. As a Permanent Resident, this has become my second home. But the word “home” wasn’t always so certain.

In 2019, driven by pessimism about Hong Kong’s future, I decided to leave and venture out. When the offer from a Singapore company arrived, many friends looked at me with quiet understanding. It was a rare and unexpected opportunity, a rational, pragmatic choice. Yet fate always likes to set its tests: I departed just as COVID-19 broke out. On that four-hour flight, the cabin was nearly empty, the silence unsettling.

Upon arrival, I was sent straight to a hotel for fourteen days of mandatory quarantine. Confined to that room, stripped of freedom, I kept asking myself: why did I move to the tropics? Wasn’t it for a better life? Yet with the world locked down and borders sealed, that “better life” felt impossibly abstract. I hadn’t even breathed the city’s air yet, only gazing at this unfamiliar city through a glass window. Those days taught me that leaving is not a romantic adventure, but an act of courage that demands bearing loneliness and uncertainty.

New to the city and knowing no one, I was fortunate to find a few Hong Kong-in-Singapore Facebook groups. The saying “on the road, you rely on friends” rings especially true abroad. Every newcomer’s first priority is hunting for accommodation online. Compared to the subdivided flats I lived in back in Hong Kong, Singapore’s housing was relatively reasonable and spacious, at least some room to breathe. Rent wasn’t cheap, but that sense of space was the first step in settling down. Gradually, I began to find my footing in this city-state.

Friends who’ve visited Singapore as tourists often praise its cleanliness, brightness, and safety. I too wandered through museums and galleries on weekends. The Merlion, the Marina Bay Sands hotel, Sentosa, the casinos, the theme parks, all checkable in two or three days. The longer you live here, the less appeal these manufactured attractions hold. Tickets are steep, and they have little to do with daily life. For residents, these landmarks pale in comparison to the hawker centres tucked inside HDB housing estates.

Living in Singapore, one thing you simply cannot miss is the food. Tourists mostly know Hainanese chicken rice, but go a little deeper and a whole world opens up: wok-fired char kway teow, fried carrot cake, Hokkien mee, and the coconut-rich laksa, each a universe of its own. Fried tofu is a particular gem, crisp outside, tender within, dipped in sweet chilli sauce, endlessly memorable. Most of these treasures hide in hawker centres, always with long queues. When you finally reach the front, the uncle or auntie will fire off a brisk: “Eat? Takeaway?” That clipped vernacular, delivered in a local accent, sounds jarring at first, but in time, it becomes endearing.

You must also learn to order a kopi. The intensely sweet condensed milk paired with a slightly toasted coffee aroma has become my daily ritual for starting work. My most anticipated breakfast: kaya butter toast with two soft-boiled eggs, a splash of soy sauce, and a sip of hot coffee. The small joys of life are really nothing more than this. Finding familiar comfort within an unfamiliar culture, perhaps that is where the settling of body and mind truly begins.

Two festivals in Singapore left a particularly deep impression, both capable of making you feel the warmth of culture, even in a foreign land. At Chinese New Year, yusheng (raw fish salad tossing) is indispensable. The platter is neatly arranged with shredded carrots, crushed peanuts, and colourful garnishes, with fresh fish at its heart. First the dressing of orange juice and syrup, then everyone gathers around the table, chopsticks raised high, tossing and mixing the ingredients while calling out auspicious phrases: “Huat ah! Huat ah!”, “Prosper! Prosper!” The louder the tossing, the greater the fortune. Even as a first-timer, you’re quickly swept up in it, that distinctly Nanyang warmth and festivity.

As for Mid-Autumn Festival, the highlight every year is the joint mooncake dinner hosted by the Hong Kong universities’ alumni associations. That is the one night of the year I hear the most Cantonese. I am always struck by how many Chinese University and New Asia College alumni are here, many seniors who put down roots decades ago. Listening to them recount the details of their lives here, sharing the hardships and turning points of their early days, I always feel I’ve gained more than years of schooling could offer. To gather with fellow alumni thousands of miles from home, that kind of connection defies easy description, yet is deeply precious. In that moment, an unfamiliar city acquires a sense of belonging.

Life gradually settled, and I grew accustomed to the heat, returning to Hong Kong in winter, I now find it too cold. Just as I thought I might stay long-term, things took a turn. The British startup I worked for began facing layoff pressure; colleagues around me left one by one, some voluntarily, some not. I started worrying about my own position. In Singapore, your job and your visa are tightly linked, lose one, and everything unravels. Anyone who has spent time here knows that visa renewals and Permanent Residency applications are unavoidable realities, and I was no exception. With so many expats in my social circle, the topic surfaces naturally in almost any gathering. After much deliberation, I decided to apply for PR, and then came the long wait. What that waiting taught me was acceptance of uncertainty.

Meanwhile, despite a satisfying salary, I began reassessing my long-term career direction and sense of purpose. I enrolled in a master’s programme at NUS in my spare time, hoping to sharpen myself against an unpredictable environment. Returning to the classroom offered a different kind of perspective. I observed the ambition and drive of international students, and felt firsthand the fierce competition faced by local students. I made local friends, and through them came to understand the political culture, racial tensions, the pressure on young men from national service, the harsh realities of the job market, and the many contradictions and struggles within Singapore’s digital transformation journey. After years of living here, it becomes harder to pin any single label on what I see and feel. This city has its gleaming, polished face, and corners that are less modern, less orderly, less efficient than the brochure. The positive image and the complex reality coexist; no single phrase can capture it. Only by being immersed in it do you truly understand.

Over these years, the international landscape and geopolitics across Southeast Asia have shifted rapidly. What I did not anticipate was that more and more friends from Hong Kong have also found their way to Singapore, for one reason or another. What does this movement of people signify? How do overseas Hongkongers navigate turbulence? How do they find their footing in a foreign land? How do they carry themselves through an uncertain world? Perhaps these questions have no standard answers. Perhaps what matters more is cultivating clear-eyed awareness, listening more, observing more, thinking more, the most fundamental capacity for facing the world. I recall Singapore’s former Prime Minister Lee Kuan Yew once saying that young Singaporeans should go out, see the world, and only then will they truly understand what kind of place Singapore is. The same wisdom applies to Hong Kong, and to fellow students of New Asia College. Sometimes you really do need to leave, to live somewhere else for a while, and only then, looking back, can you deeply understand what Hong Kong is, and who you truly are.

Five years is enough time for the unfamiliar to become familiar, and for a person to redefine what “home” means. Living abroad is not escape, but expansion; not severance, but extension. It has allowed me to move between different systems and cultures, and in the space between leaving and looking back, to find a more settled sense of self. Now, every time I return to Changi Airport and see those two words, my feelings are nothing like the apprehension of five years ago. Living and moving between two cities, I have come to deeply understand the meaning of that old verse:, wherever the heart finds peace, that is home. Singapore is no longer merely a place I landed. It is the place where a real life was lived.

Think Like a CEO

Victor Leung — Tue, 24 Mar 2026 15:28:17 +0000

Studying how the world’s top executives make decisions, you start to notice patterns that have very little to do with industry, company size, or technology stack. What stands out is the rigour they apply to their own decision-making, the same rigour most enterprise architects reserve exclusively for the systems they design. EA work sits at a strange intersection. We’re technical enough to understand the systems, commercial enough to translate them into business value, and senior enough to influence strategy, yet the craft of deciding well rarely gets the same attention as the craft of designing well. Here’s what we can borrow from the executives who’ve made it their discipline.

The best CEOs invest heavily in what might be called the “North Star”, a compelling answer to the question: what are we actually for? Not a mission statement crafted by committee, but a genuine narrative that reframes how success is measured, influences every resource decision, and gives people a reason to behave differently. Nelson Mandela and François Pienaar didn’t coach South Africa’s rugby team to win a World Cup. They gave them a reason to play that made winning feel secondary to something larger. The team played differently as a result.

Enterprise architects face a version of this problem every time we present a target state architecture. We often lead with the technology, the Kubernetes cluster, the event-driven backbone, the API gateway layer, and leave the “why this, why now, why should anyone care” for slide 14. That’s backwards. Before presenting any architecture proposal, write one paragraph, not a slide, that answers: what does success look like for the business in three years, and why does this architecture enable it specifically? If you can’t write that paragraph, the proposal isn’t ready.

The CEOs who build durable change do so because they connect the technical to the meaningful. They find the intersection between organisational capability and a future the organisation actually wants to inhabit. Our job, as architects, is to do the same: translate systems thinking into a story with genuine stakes.

DuPont’s CEO Ed Breen has a habit that every architect should steal. Before committing to any major decision, he asks one question obsessively: what does the downside look like, and can I live with it? Not the expected outcome. Not the best case. The downside. Jamie Dimon at JPMorgan goes further, describing decision-making as toppling dominoes, the first is easy to push, but you need to trace every subsequent one before you commit. If any downstream outcome is company-threatening, regardless of its probability, you choose a different path. Full stop.

This is, in essence, risk architecture applied to strategy. And it’s something we rarely do systematically in EA work. We model the happy path. We document the target state. We sometimes note risks in a RAID log. But we seldom sit down and ask: if this migration goes badly, which scenario would we actually be unable to recover from? And are we designing for that scenario, or just hoping it doesn’t happen? For every major architecture initiative, it’s worth running a “reverse architecture review”: map the failure modes first, not the capabilities. Define which failure mode is unacceptable, and make that the constraint your design must satisfy before you optimise for performance, scalability, or cost.

Adobe’s shift to cloud subscriptions is the canonical example of this discipline executed well. Shantanu Narayen’s team spent hours in the boardroom stress-testing pricing models, unit economics, and the rate at which perpetual license revenue would decline. They weren’t confident, they were prepared. There’s a significant difference, and it shows in the outcome.

Among the most underappreciated habits of elite executives is how they manage investment portfolios. They don’t release budgets on an annual calendar cycle. They release them against performance milestones: prior investment must produce demonstrable results before the next tranche is unlocked. This sounds obvious. In practice, it is routinely ignored in large technology programmes. Platform teams get multi-year budgets approved upfront. Workstreams are funded annually regardless of delivery evidence. Architecture decisions made in year one persist into year three because no one built in a formal checkpoint to ask: is this still the right call?

The discipline here is separating “continue” from “complete.” At each milestone, the question isn’t whether the programme is done, but whether it should continue. The bar is different, and it keeps teams honest about value delivery. DuPont reviews major programmes one year after completion, something EA teams almost never do. Without retrospectives, patterns of poor decision-making compound silently across initiatives.

Danaher’s Larry Culp offers the sharpest framing: he consistently asked what a new CEO with no emotional attachment to the portfolio would do. EAs should ask the same of their architecture. Is this still the right platform, or are we maintaining it because we built it? Thinking like an outside investor, rather than an internal custodian with sunk costs to defend, is one of the harder mental shifts in the discipline, and one of the most valuable.

Research consistently shows that only one in three strategies is successfully implemented, and that 72% of the obstacles are human and cultural, not technical. Peter Drucker put it plainly: “Culture eats strategy for breakfast.” Enterprise architects tend to believe that if the architecture is sound, adoption will follow. It won’t. The most technically elegant architecture I’ve seen fail did so not because of a bad design choice, but because the teams on the receiving end didn’t trust the process, didn’t understand the rationale, and didn’t feel ownership over the outcome.

Satya Nadella’s transformation of Microsoft is instructive here. He didn’t lead with the technology strategy. He led with a cultural reframe, “Growth Mindset,” rooted in Carol Dweck’s research, which gave the organisation a way to think about failure, learning, and collaboration that made the technical changes feel coherent rather than imposed. The cultural shift was designed before the technical one was deployed, not as an afterthought.

For EAs, this means the real deliverable of any major architecture initiative isn’t the target state diagram. It’s the shared understanding, across engineering, product, operations, and leadership, of why the architecture makes sense and what it enables. Culture put in place through genuine consultation, not broadcast from above, is what makes technical transformation durable. If that shared understanding doesn’t exist, no amount of documentation will substitute for it.

The most effective executives share a consistent discipline: tracking technological and market shifts early enough to invest before they become conventional wisdom. DSM’s Feike Sijbesma describes a practice of reading across disciplines, including subjects apparently unrelated to his industry, and building networks across business, science, and society, specifically to surface insight that siloed thinking would miss. Netflix’s trajectory makes the same point structurally. The DVD-by-mail service was never the end state; it was the cash-generating beachhead that funded the bet on streaming. Reed Hastings didn’t wait for streaming to be proven. He invested into uncertainty, absorbed the short-term criticism, and held the position.

Enterprise architects are increasingly expected to play this role, not just designing for current requirements, but anticipating the architectural implications of AI adoption, regulatory change, data sovereignty shifts, and the ongoing fragmentation of the vendor landscape. The temptation is to anchor too close to present-state thinking: what does the organisation need today? That question is necessary but not sufficient. A better question is: what is becoming true in the industry that our architecture is not yet positioned for?

The goal isn’t prediction, it’s positioning. An architecture that can absorb an AI-native workflow, a new data residency requirement, or a major vendor consolidation without requiring a full re-platform is one that has done this work. Building a deliberate horizon-scanning practice into your working rhythm doesn’t require a formal framework. It requires the discipline of reading outside your stack, maintaining networks outside your organisation, and setting aside time each quarter to surface what’s changing before it becomes obvious.

The thread connecting all of these behaviours is a single disposition: acting like an owner. It means making decisions as though you are the full proprietor of the enterprise, unbounded by political constraints, undistracted by short-term optics, accountable to the long-term value of the whole. Enterprise architects rarely think of themselves this way. We think of ourselves as advisors, as enablers, as the people who make the technical trains run on time. That framing keeps us safe, but it also keeps us peripheral. The executives who drive lasting change refuse to be peripheral to their own organisations.

The architecture function is most powerful when it occupies the same posture: opinionated about outcomes, not just about designs. Willing to say “this is the wrong direction” and defend it with evidence. Accountable for what gets built, not just for the blueprint that preceded it. The lesson from the best executives isn’t a set of techniques, it’s a stance. And it’s one worth consciously adopting.

Beyond Buy vs. Build, Your AI Sourcing Strategy Needs a Third Option

Victor Leung — Tue, 10 Mar 2026 14:23:11 +0000

The "should we build or buy?" question has haunted technology leaders for decades. But in the age of AI, that binary framing is dangerously inadequate. With organisations now pursuing an average of ten AI initiatives simultaneously, and 43% running between five and ten at any given time, the stakes of getting your sourcing strategy wrong have never been higher.

Having spent years as a solution architect working across core banking platforms, cloud-native infrastructure, and now portfolio architecture, I've watched organisations struggle with this exact tension. The reality is that most enterprises don't need to pick a lane. They need a spectrum. The "buy, build, and blend" framework goes beyond the old dichotomy and offers the clearest mental model I've seen for how architects should be thinking about this problem.

The traditional buy-vs-build framing implies a clean boundary: either you purchase commercial off-the-shelf (COTS) software, or you write it yourself. In practice, that boundary barely exists anymore. Between pure configuration of a COTS product and building a custom application from scratch, there's an entire spectrum of options: extending vendor products via marketplace add-ons, building custom integrations with low-code or pro-code, creating automations and connectors between related apps, and, increasingly, building AI agents and custom UIs on top of purchased platforms. This "blend zone" is where most real enterprise work happens today, and it maps neatly onto a spectrum from undifferentiated to differentiated business capabilities. The architectural implication is straightforward: reserve your engineering capacity for capabilities that genuinely set you apart, and lean on vendor ecosystems for everything else.

Before diving into AI-specific considerations, it's worth anchoring on five factors that should drive all application sourcing decisions, because they apply with even greater force in the AI context. The first is criticality and business value, is the technology your core value proposition, or is it a tool to solve a business problem? If AI is central to your product offering, the calculus shifts heavily toward build. If it's an operational improvement, buying or blending likely makes more sense. The second is risk and internal competencies, which forces an honest assessment of your IP exposure and vendor lock-in risk alongside a candid look at whether your organisation actually has the skills to build and maintain what it's contemplating. In AI, this question cuts especially deep, the talent market is ferociously competitive, and the gap between having a few data scientists and having production-grade ML engineering capability is vast.

The third factor, total cost of ownership, is where organisations most frequently deceive themselves. TCO for any application stretches across four phases: go-live costs (design, development, testing, initial licenses), current annual costs (both recurring operations/support and nonrecurring maintenance), future costs (operating cost variations, predictable upgrades, potential enhancements), and decommissioning costs (particularly data retention). I've seen too many build decisions justified by comparing initial development cost against multi-year licence fees, while conveniently ignoring the ongoing maintenance, adaptive maintenance, and eventual decommissioning burden. The fourth factor is partners' abilities, their capacity to execute and the completeness of their vision, which matters enormously in a market where AI vendors range from research-stage startups to hyperscaler platforms. The fifth is opportunities: whether you're deploying your internal capacity on the highest-value work, or burning cycles on problems that vendors have already solved at scale.

If the framework tilts toward buying for undifferentiated capabilities, it's worth understanding why organisations still hesitate. Survey data on COTS challenges tells a revealing story. The top three concerns, vendor lock-in (15%), integration issues (14%), and limited customisation (13%), are fundamentally about control and flexibility. The next tier, hidden costs, lack of control over updates, and security concerns, reinforces the same theme. For architects, this is a familiar pattern. The promise of COTS is speed and reduced engineering burden. The reality is that you're trading one set of problems (building and maintaining software) for another (integration complexity, vendor dependency, and reduced agility). The question isn't which set of problems is smaller, it's which set your organisation is better equipped to manage.

When it comes to AI specifically, the decision framework gets richer. There are nine influential factors that CIOs must weigh when choosing between buy, blend, and build. The first three are strategic: external differentiation (will this AI capability set you apart from competitors?), compliance (can the solution meet regulatory requirements?), and security (what are the risk implications?). The next three are ecosystem-related: vendor ecosystem maturity, data origin and its influence on model accuracy, and available skills. The final three are economic and operational: short-term implementation costs, long-term maintenance costs, and impact on workers. What makes this framework powerful is the recognition that these factors carry different weight depending on your strategic intent, and this is where the defend/extend/upend categorisation becomes genuinely useful.

The most actionable insight from this framework is the three-way categorisation of AI use cases by strategic intent. "Defend" use cases aim to maintain competitive parity, think augmenting individual productivity with tools your competitors are also adopting. For these, the bias should be heavily toward buying from incumbent vendors with embedded AI, because the goal is commodity capability delivered quickly and reliably. The decision here is essentially a series of "yes/no" questions about whether your incumbent vendor can handle the use case, and in most cases, the answer should favour the incumbent. Minimise customisation, leverage existing vendor relationships, and move on to higher-value problems.

"Extend" use cases aim to differentiate, transforming processes or teams to create competitive advantage. Here, the buy/blend/build decision gets genuinely complex. Questions like "does this offer more than minor differentiation?" and "is the upfront cost of building justified by freedom from future vendor price hikes?" don't have easy answers. The presence of "yes, but..." responses throughout the framework is telling. It acknowledges that extend decisions are inherently contextual and require careful judgment rather than formulaic answers.

"Upend" use cases aim to disrupt, creating new propositions, products, or markets. For these, the framework tilts toward blend or build, but with important caveats. Speed to market may still justify buying as a temporary measure, vendor access to data you can't replicate may make blending essential, and compliance and security requirements in unfamiliar geographies may demand vendor partnerships. The key insight is that even for disruptive AI initiatives, pure build is rarely optimal.

One of the most compelling ways to visualise the enterprise AI stack is as a layered sandwich. The bottom layers, your centralised data and custom-built AI, are within your control. The top layers, external data and embedded AI from vendor products, are outside your control. Trust, risk, and security management and bring-your-own-AI sit in the middle, mediating between what you own and what you consume. For enterprise architects, this translates into a practical design principle: invest in strong foundations (data infrastructure and governance), build protective layers, and be intentional about which AI capabilities you build versus consume. The organisations that get the layering right will be the ones that can absorb the rapid pace of AI innovation without constantly re-architecting their stack.

There's an interesting maturity dimension to all of this. High-maturity organisations are three times more likely than low-maturity ones to adopt a hybrid vendor management strategy (33% vs 11%). Low-maturity organisations overwhelmingly default to centralised approaches (61%), while high-maturity ones are more evenly distributed across centralised (41%), decentralised (26%), and hybrid (33%) models. This suggests that as organisations mature in AI, they naturally evolve away from one-size-fits-all vendor strategies toward more nuanced, context-dependent approaches. This tracks with what I've observed in practice: early AI adoption benefits from central coordination, but scaling AI across the enterprise requires giving business units more autonomy while maintaining guardrails.

There's also a procurement reality that architects need to confront. Ninety percent of recent software purchases included GenAI capabilities, but only 25% of respondents felt they achieved high-quality deals on those purchases. This gap reveals the current market dynamic, GenAI is being bundled into almost everything, but buyers are struggling to assess value and negotiate effectively. For architects advising on procurement, this means applying extra scrutiny to the GenAI components of vendor pitches. Are the AI features genuinely useful for your use cases, or are they checkbox additions designed to justify premium pricing? Does the vendor's AI actually leverage your data to deliver differentiated outcomes, or is it generic capability that any competitor could also access?

If I were to distil all of this into practical guidance for fellow architects, it would be this. Start by classifying every AI initiative as defend, extend, or upend. This single step will dramatically simplify your sourcing discussions by establishing the right default bias for each initiative. For defend initiatives, fight the urge to over-engineer. Your incumbent vendors will almost certainly add the capability you need, and the integration cost of a new vendor rarely justifies the marginal improvement. For extend initiatives, invest in the "blend" capabilities, low-code customisation, API integration, and connector architecture, that let you combine vendor platforms with proprietary logic. This is where your architecture practice adds the most value. For upend initiatives, be ruthlessly honest about your organisation's readiness. The data, skills, compliance, and security requirements for disruptive AI are substantial, and underestimating them is the fastest path to an expensive failure. And for all three categories, model the full TCO, including decommissioning costs and data retention, before committing to a path. The most expensive decision is the one you have to reverse.

An Enterprise Architect's Guide to Sourcing AI for Real Value

Victor Leung — Mon, 09 Mar 2026 14:07:55 +0000

Most enterprises are spending more on AI than ever, and getting less clarity in return. The budgets are approved, the vendors are circling, and the executive decks are full of the word "transformation." But beneath the surface, the same uncomfortable reality keeps surfacing: the data isn't ready, the platforms aren't ready, and the use cases aren't mature enough to scale.

AI readiness depends on three pillars that most organisations are still building simultaneously, clean, governed, AI-ready data; platforms that can orchestrate agents across ecosystems; and use cases that have graduated beyond experimentation into repeatable business outcomes. If you're honest with yourself, your organisation is probably still working on all three. So what should an Enterprise Architect actually do about it?

The single most important shift I've seen in mature AI programmes is the refusal to treat "AI" as a single budget line item. Instead, every AI initiative gets classified against one of three strategic postures. Defend means embedding AI into existing applications to maintain competitive parity, think copilots bolted onto your CRM, or ML models running fraud detection on your existing transaction platform. The adoption model here is simple: consume an application, embed model APIs. Your sourcing rigour should be proportionate, medium diligence on talent and advisory, high on accelerators and IP, low on industry-specific expertise because the vendor is doing the heavy lifting. Extend means optimising specific workflows with custom agents, retrieval-augmented generation, or fine-tuned models. This is where most ambitious enterprises sit today: building domain-specific intelligence on top of foundation models. The evaluation profile shifts dramatically, you need strong partnership ecosystems and data engineering depth, and vendor copyright and compliance scrutiny moves to medium. Upend means building entirely new business models or products powered by AI, custom model development, novel data strategies, proprietary training pipelines. Everything is High on the evaluation scale: talent, ethics, industry experience, commercials. If you're truly trying to upend your market, you should be treating AI sourcing with the same rigour you'd apply to an M&A deal.

The point is that the level of governance, the vendor you hire, the contract you write, and how you measure success should all flow from which posture you're in. Conflating "Defend" with "Upend" is how you end up overspending on commoditised capabilities or under-governing a genuinely transformational bet.

We're watching three traditionally separate disciplines collide: Data & Analytics, Software Engineering & Infrastructure, and Business Process Design. At the intersection sits what's increasingly being called "AI Services", but I'd reframe it for architects as the new integration layer you need to own. The emerging competencies in this convergence zone, Insight Engineering, AI Integration, and Work Orchestration, don't belong neatly to any one team. Your data scientists won't build the agent orchestration. Your platform engineers won't design the business workflows. And your business analysts won't understand the infrastructure constraints of running inference at scale. As an Enterprise Architect, this convergence is your territory. You need to be the one defining the reference architecture that connects these domains, establishing the shared services layer, and ensuring the platform strategy doesn't fragment into shadow AI projects.

The two-tiered model for AI use-case management is worth studying here. They separate the "why" (a management team of senior marketing leaders meeting monthly to set strategic direction and approve priorities) from the "how" (use-case leaders from each subfunction meeting weekly to plan, adapt, and track execution). The critical insight: having a dedicated subfunctional head for each use case who assesses their team's capacity to take on AI projects is what prevents the all-too-common pattern of AI initiatives being layered on top of already-overloaded teams. Capacity assessment isn't glamorous, but it's the difference between a use case that ships and one that dies in a shared backlog.

One of the most underappreciated failure modes in AI programmes is misalignment across the C-suite. Each leader brings fundamentally different concerns to the table. Business leaders want to define strategic ambition, but worry about losing human control. IT and Data leaders need interoperability with existing systems, but struggle with integration complexity. Legal and Compliance care about IP protection and regulatory adherence, but can't always ensure fairness and accountability at model level. And Finance and Procurement want ROI and cost transparency, but often lack the frameworks to measure AI-specific value. The architect's role is to make these concerns legible to each other. A gated approach with human decision points at each stage addresses the business leader's concern. Validation checkpoints at every integration gate satisfy IT. Contractual IP coverage handles Legal. And FinOps rigour, which I'll get to, gives Finance what they need.

Here's the contract model shift that too few organisations have made: if your vendor uses AI to code 50% faster but you're still paying by the hour, you lose. The productivity gains from AI tooling flow entirely to the vendor's margin. The move is from buying hours to buying outcomes. The contract model spectrum runs from traditional time-and-materials and staffing deals (labour-intensive, "as-is" performance) through to outcome-based and shared-risk models that tie payment to actual business results. For AI-heavy engagements, you should be demanding outcome-based or value-based pricing instead of FTE blocks, shared-risk models for innovation and proof-of-concept work, AI-Augmented Capacity (AI PODs) rather than pure staff augmentation, and Agent Efficiency Metrics written directly into the contract. The new KPIs for measuring these "digital employees" include metrics like Agent Efficiency Index (how efficiently does the agent complete tasks versus the optimal workflow?), Autonomy Utilisation Ratio (what percentage of tasks complete without human intervention?), and Decision Accuracy. If you're not measuring these, you're flying blind on whether your AI investment is actually performing.

The risk landscape for AI is genuinely complex. There are at least 18 interconnected risks across four categories, behavioural (accuracy, bias, scope violations), security (sensitive data leakage, hacker abuse, vendor copyright issues), transparency (failure to disclose AI involvement, explainability gaps), and a catch-all of operational risks (energy waste, HR dependency, multi-agency complexity). This is why the framework of Trust, Risk & Security Management matters for architects. The five mandates are worth internalising: a Governance Framework that integrates AI into your existing enterprise risk taxonomy; Compliance and Accountability mechanisms with continuous monitoring and defined vendor responsibility; Human Oversight and Transparency requirements where vendors must disclose AI use on client data and high-risk outputs need human review; Cross-Functional Collaboration through fusion teams that include Legal, Risk, IT/Data, and Business; and Capability and Training Transfer, meaning your vendors should be contractually obligated to support change management and AI literacy, not just deliver code.

The case study is instructive here. Their GenAI Centre of Excellence delivers structured training that covers everything from foundational model concepts through to the cost implications of token size and prompting decisions. The CIO's point is sharp: anyone can download a GenAI training, but leaders need to understand the cost impact of their AI decisions. That's a FinOps discipline, not just a technology one.

The final architectural shift is perhaps the most fundamental. The traditional shared services model, centralised, factory-like, building everything in-house, doesn't scale for AI. You can't be the bottleneck. Instead, shared services should provide the AI platform and the guardrails, then let the business build safely within those boundaries. The accountability model flips: "You build it, you pay for it." Technical debt gets tied back to the owner, not absorbed centrally. This also means getting serious about shared decision rights and shared costs. Decisions about what to share versus what to keep sovereign need to be driven by compliance and data sovereignty requirements. Variable costs, which are inherent to consumption-based AI pricing, need FinOps discipline to manage, with a focus on recovering investment with demonstrable ROI.

When you do go to market, the procurement process itself needs to evolve. The traditional RFP-then-negotiate cycle is too slow and too rigid for AI partnerships. A sprint-based competitive co-design approach works better. Start with a long list and an initial RFS to short-list and onboard candidates. Then run a co-creation sprint with 3-6 suppliers to shape the deal collaboratively. Narrow to 2 for detailed co-creation, due diligence, and SOW development. Final sprint: competitive negotiation and contract signature. This is agile, structured, outcome-driven, and, critically, competitive throughout. You're not just evaluating proposals on paper; you're seeing how vendors actually work with your team before you commit. The RFP questions themselves should be outcome-driven: How does the business case demonstrate ROI? How is pricing structured across renewals, scaling, and managed services? How will governance manage risk? What frameworks exist for prioritising use cases? Can you show industry-specific examples with measurable outcomes?

One final caution: AI services costs are cumulative and easy to underestimate. Model development, data management, licensing, infrastructure, integration, and ongoing support all compound. The critical contract terms to watch are pricing mechanics and data/IP terms (high risk), XLAs/SLAs/KPIs and liability allocation (high risk), and exit and continuity clauses (moderate risk). The AI services market is projected to reach $1.11 trillion by 2029, with indirect services growing at a 49% CAGR. Application implementation alone is forecast at $350 billion (combining $160B direct and $190B indirect). The money is flowing, the question is whether it's flowing towards outcomes or just towards activity.

If I had to distil this into a single actionable framework for fellow architects, it would be this. Source for value, categorise every AI initiative as Defend, Extend, or Upend, and right-size your sourcing rigour accordingly. Stop buying AI generically. Govern for safety, operationalise framework, mandate cross-functional fusion teams, and require vendors to support change management and AI literacy as part of every engagement. Capture the AI dividend, shift from T&M to outcome-based contracts, demand agent efficiency metrics, and establish FinOps discipline before the consumption-based costs spiral. The organisations that get this right won't just be adopting AI. They'll be architecting a fundamentally different relationship between technology, business outcomes, and vendor partnerships. And that's exactly the kind of convergence zone where Enterprise Architects should be leading the conversation.

How Selfless Thinking, Positive Mindset, and Relentless Effort Drive Enterprise Transformation

Victor Leung — Tue, 24 Feb 2026 15:35:07 +0000

As an Enterprise Architect, I operate in environments where strong opinions, competing priorities, and high-stakes decisions collide. Strategy reviews, vendor debates, transformation programs, AI adoption, platform modernization—these are not purely technical discussions. They are arenas where ego, identity, and status can easily overshadow truth.

The most powerful insight is this: the best form of reasoning begins in a state of selflessness.

When ego enters the room, exploration turns into combat. Dialogue becomes about winning. Positions harden. Facts are selectively presented. Weaknesses are concealed. Counterarguments are distorted. In the end, someone may “win” the meeting, but the organization loses clarity. Truth is sacrificed on the altar of self-preservation.

This dynamic is particularly dangerous in architecture and digital transformation. When roadmaps are defended because they are “mine,” when design decisions are protected because they were publicly endorsed, when failure becomes personal rather than systemic, learning stops. And once learning stops, evolution stops.

True leadership requires the ability to detach identity from ideas.

A person’s value is determined by the heart or, in modern organizational terms, by the quality of their intent and character. The architect who can say “I was wrong” without defensiveness is more valuable than the architect who is always right but never open. The executive who adjusts course based on evidence demonstrates strength, not weakness.

Selflessness does not mean passivity. It means intellectual courage. It means pursuing what is true and effective rather than what is personally validating.

In transformation programs, we often declare bold, multi-year ambitions: cloud-native by 2028, AI-enabled workflows across the enterprise, complete platform modernization. Ambition is necessary. But if the distance between effort and visible progress is too large, morale deteriorates. Teams disengage. “Close enough” replaces excellence.

From a systems perspective, this is a feedback-loop problem. Human motivation requires reinforcement. If intermediate milestones are invisible or unattainable, entropy wins. The solution is not to shrink ambition, but to architect progress into meaningful increments. Large vision. Short feedback cycles. Visible progress.

Another powerful theme is mindset. The story of the beggar who, even when given opportunity, thinks only about improving his ability to beg is a metaphor for organizational inertia. Many companies adopt new technology but keep old thinking. They migrate to the cloud but preserve on-prem governance models. They deploy AI but continue manual approval chains. They invest in platforms yet optimize for legacy KPIs.

Tools do not transform organizations. Mindsets do.

Positive thinking, in this context, is not naïve optimism. It is constructive agency. It is the belief that constraints can be redesigned, processes reimagined, and systems improved. It is the discipline of reframing problems as solvable architecture challenges rather than immovable realities.

In enterprise environments, pressure is constant. Regulatory scrutiny. Budget limitations. Market volatility. Like rowing upstream, if you do not advance, you drift backward. Seeking comfort—avoiding difficult conversations, postponing refactoring, deferring governance reform—is not neutrality. It is regression.

There is a reason leaders emphasized making an effort no less than anyone else’s. Sustained excellence is not episodic. It is daily. In architecture, this means documentation that is not postponed, principles that are enforced consistently, and standards that are improved continuously. Strategy without disciplined execution is theatre.

However, relentless effort without reflection becomes burnout. The deeper message is about character elevation through work. Work is not merely output production; it is a vehicle for refining judgment, humility, and resilience.

Arrogance inevitably leads to rejection. In technical leadership, arrogance manifests subtly: dismissing junior engineers, ignoring operational feedback, resisting alternative viewpoints. Humility, by contrast, compounds. It attracts talent. It accelerates learning. It builds trust capital across functions.

The idea that development depends on fighting for recognition is outdated. In modern enterprises, influence is not seized through confrontation but earned through clarity, consistency, and integrity. The most respected architects are not the loudest voices in the room; they are the ones whose frameworks consistently reduce complexity and improve outcomes.

Ultimately, the purpose of professional life is not title accumulation or architectural dominance. It is the cultivation of judgment. The strengthening of character. The discipline to live and work fully in the present moment.

Everything happens now. Strategy is executed now. Culture is shaped now. Trust is built now.

If we approach conversations without ego, set ambitious but architected goals, cultivate positive and constructive thinking, and commit to sustained effort, then whether our careers are dramatic or quietly impactful, they will be meaningful.

In the end, the true architecture we are building is not only systems and platforms.

It is ourselves.

The Only Strategy That Matters

Victor Leung — Tue, 10 Feb 2026 14:02:47 +0000

In every era of business history, the companies that endure share one common trait: they consistently deliver products that people genuinely want. Those that fail to do so eventually disappear. This pattern is neither new nor surprising, yet it is often obscured by layers of strategy frameworks, operating models, and architectural abstractions. As Enterprise Architects, we frequently discuss strategy, business models, platforms, and capabilities. But strip everything back, and one uncomfortable truth remains: without products or services that resonate with real user needs, all strategic intent is theoretical. Profits, brand equity, market leadership, and even employee well-being are not primary achievements; they are downstream effects of sustained value creation.

The essence of business is therefore simple, though not easy: continuously provide what users actually need, not what the organization wishes they needed. Business, at its core, is an ecosystem of demand and supply. Hungry people seek food, cold people seek warmth, and bored people seek entertainment. Organizations that survive are those that sense these needs accurately and respond effectively. The medium may change—physical products, digital platforms, AI-driven services—but the underlying dynamic does not.

This ecosystem perspective challenges traditional enterprise thinking. Architectures designed purely for internal efficiency, compliance, or cost optimization often drift away from user value. When that happens, strategy becomes detached from reality, and business models turn into elegant diagrams with diminishing relevance. Enterprise Architecture should exist to shorten the distance between emerging user needs and organizational response, not lengthen it.

The true competitive advantage of an enterprise is not its size, brand, or market share, but its capability to understand evolving user needs, translate those needs into concrete offerings, and do so repeatedly faster than competitors. This requires more than technology. It demands organizational sensing, continuous feedback loops, empowered teams, and architectures that favor adaptability over rigidity. When needs change, as they inevitably do, the enterprise must detect the shift early and respond decisively.

Architecturally, this means favoring modularity over monoliths, learning loops over fixed roadmaps, decentralized decision-making supported by shared standards, and platforms that enable experimentation rather than simply enforce control. These are not technology choices alone; they are reflections of how seriously an organization takes its responsibility to remain relevant.

One of the most dangerous illusions in large enterprises is the pursuit of stability for its own sake. Working in a large organization, following authority unquestioningly, or optimizing purely for personal advancement may feel safe in the short term. Yet ecosystems do not reward comfort; they reward relevance. When individuals lose sight of user needs and focus instead on hierarchy, status, or internal politics, the damage compounds over time.

As this mindset spreads, corporate culture subtly shifts. Employees who genuinely strive to create value for users find it increasingly difficult to do their work. Internal competition replaces collaboration, and personal survival begins to outweigh collective success. Eventually, capable people sense the dysfunction and leave. What remains is an organization that appears busy and well-structured, yet is fundamentally disconnected from the ecosystem it depends on.

Ironically, success itself often marks the beginning of decline. Rapid growth and aggressive hiring can dilute purpose if not guided with care. When scale outpaces a shared understanding of value creation, entropy sets in. From an architectural perspective, this is the moment where discipline matters most. Architecture must reinforce alignment to user outcomes, protect the capabilities that create real value, and preserve the organization’s ability to adapt and learn.

Enterprise Architecture is not about producing perfect structures or exhaustive blueprints. It is about sustaining relevance in a constantly changing ecosystem. Strategy, operating models, and technology stacks are means, not ends. When they drift away from real human needs, they lose their legitimacy. Organizations exist to serve those needs, and those that do so continuously will endure. The responsibility of the Enterprise Architect is to ensure that the enterprise never forgets why it exists in the first place.

How Enterprise Architecture Shapes Strategy in a Volatile World

Victor Leung — Sun, 08 Feb 2026 15:08:58 +0000

Enterprise Architecture sits at a unique crossroads between strategy and execution. It is one of the few disciplines that must simultaneously understand the external forces shaping the market and the internal capabilities that determine what an organisation can realistically do next. In an era of global competition, digital disruption, and constant change, architects can no longer afford to align with only one strategic worldview. Instead, we must integrate multiple perspectives into a coherent, actionable lens.

Two classic schools of strategy — the positioning perspective and the resource-based perspective — continue to shape how organisations think about competitiveness. Each has strengths, limitations, and profound implications for how architecture should be designed and governed.

The positioning perspective assumes that the external environment defines strategic freedom. Customers, competitors, suppliers, substitutes, and new entrants collectively constrain what is possible. Michael Porter’s Five Forces framework remains influential because it forces leaders to confront uncomfortable truths about industry attractiveness and competitive pressure. From this view, strategy is about choosing where to play and how to defend that position. Success depends on continuously monitoring market signals, anticipating shifts in customer demand, and responding faster than rivals.

For Enterprise Architects, this perspective reinforces the importance of external awareness. Architecture decisions cannot be made in isolation from pricing pressure, regulatory change, ecosystem dynamics, or platform competition. A technically elegant architecture that ignores these forces risks optimising the wrong outcomes. However, the positioning perspective also has a blind spot: it implicitly assumes that firms within an industry are largely similar, differing mainly in how well they adapt.

This is where the resource-based perspective fundamentally changes the conversation. Rather than starting with the market, it starts with the firm. Organisations are not interchangeable. Each has a unique combination of tangible and intangible resources — skills, knowledge, processes, culture, data, brand, and technology — that competitors cannot easily replicate. From this angle, strategy is not about fitting into an existing industry structure, but about reshaping it.

The early days of Google illustrate this clearly. The founders did not begin by analysing entry barriers or industry profitability. They focused on what they could do uniquely well, and in doing so, they changed the rules of the search market entirely. For architects, this perspective validates a core architectural instinct: sustainable advantage comes from what is hard to copy, not what is easy to benchmark.

Yet resources alone are not enough. Servers, engineers, data, and brands do not create value by themselves. What matters is the organisation’s ability to combine them effectively. This is the realm of capabilities — the complex bundles of skills, learning, and coordination embedded in organisational processes. Capabilities bridge the gap between internal potential and external value.

Thinking in terms of capabilities shifts architectural conversations away from isolated systems and towards end-to-end outcomes. The ability to anticipate customer needs, to sense emerging technologies, to deliver at scale, or to orchestrate partners across borders are not properties of individual applications. They emerge from how systems, people, and processes work together. Enterprise Architecture, at its best, is a capability-design discipline.

Global competition raises the stakes further. Competing internationally introduces new demands: managing currency risk, scanning global technology trends, and transferring tacit knowledge across borders. Firms also face structural disadvantages — liabilities of foreignness, expansion, smallness, and newness — that compound complexity. Architecture must therefore enable learning, adaptability, and local responsiveness without fragmenting the enterprise.

This is where comparative analysis becomes critical. Many organisations believe they are above average, yet few systematically study their competitors. Without credible competitor intelligence, claims of “core capabilities” are often little more than internal myths. Architects should challenge this complacency. Understanding how rivals structure their platforms, manage costs, or scale operations provides essential context for architectural trade-offs.

Benchmarking, when done well, is not imitation for its own sake. Xerox’s response to Canon in the 1980s was not about copying a product, but about learning better processes. For modern enterprises, benchmarking might involve cloud cost structures, DevOps maturity, data platform scalability, or ecosystem integration patterns. The goal is not to be identical, but to close blind spots.

Strategic intent, however, means little without effective implementation. Organisational structure plays a decisive role here. Matrix structures promise synergy across products and markets, but often collapse under their own complexity. Dual reporting lines, overlapping accountability, and decision latency can undermine execution, especially across geographies. Some firms abandon the matrix after painful experience; others, like Disney, sustain hybrid forms through strong leadership and clarity of purpose. For architects, this reinforces a hard truth: structure and governance matter as much as technology.

Finally, strategy and architecture must evolve through change. Not all change is equal. Incremental change refines what already exists; transformational change redefines beliefs, identities, and priorities. The latter demands leadership, not just management. Global organisations must choose change styles deliberately, based on urgency, environmental fit, and internal support. Participative approaches work when time and alignment exist. Dictatorial transformation, while uncomfortable, may be necessary when survival is at stake.

Enterprise Architects are often positioned as neutral facilitators, but in transformational moments, neutrality is not enough. Architects must help leaders translate vision into coherent operating models, align capabilities with ambition, and ensure that change is structurally and technologically possible.

In a volatile world, sustainable advantage does not come from choosing between positioning or resources, structure or culture, incremental or transformational change. It comes from integrating them. Enterprise Architecture provides the connective tissue — linking market insight to internal capability, strategy to execution, and vision to reality. When done well, it does not merely support strategy. It helps shape it.

AI Governance for the Enterprise

Victor Leung — Thu, 05 Feb 2026 15:02:02 +0000

AI is no longer a peripheral capability in the enterprise. It is rapidly becoming embedded into core business platforms, decision-making processes, and customer interactions. From copilots that assist sales teams to autonomous agents capable of triggering actions across workflows, AI is reshaping how value is created. Yet as AI becomes foundational, governance has not kept pace. Recent analysis of Salesforce’s AI strategy highlights a growing concern across the industry: while vendors race to embed AI into platforms, customers are exposed to new risks around cost predictability, data governance, and operational control.

The current AI landscape is characterised by speed and fragmentation. Large SaaS providers are bundling generative and agentic AI into existing products, often with evolving licensing models that are difficult to forecast over multi-year horizons. At the same time, confidence in AI reliability remains mixed. Even vendors acknowledge that large language models can hallucinate, misinterpret context, or act unpredictably when granted autonomy. For enterprises, this creates a tension between the pressure to adopt AI for competitive advantage and the responsibility to protect customers, data, and financial sustainability.

One of the most immediate risks is data governance. AI systems are only as trustworthy as the data they consume, yet generative AI blurs traditional boundaries around data usage. Sensitive customer, commercial, or operational data can be unintentionally exposed through prompts, model training, or generated outputs if controls are insufficient. For organisations operating in regulated environments, this risk extends beyond reputational damage into regulatory and legal liability. Enterprise architects must therefore treat AI access to data as a privileged operation, governed by the same rigor as access to core transactional systems.

Cost and commercial risk is another emerging challenge. Consumption-based AI pricing, while flexible in theory, introduces significant uncertainty at scale. Analyst warnings about AI licensing structures converting from capped agreements to defined-quantity pricing underscore a broader issue: enterprises may only fully understand their AI cost exposure after adoption is widespread. Without architectural mechanisms to observe, limit, and forecast AI usage, organisations risk budget overruns or unfavourable contract renegotiations at renewal time. This shifts AI governance from a purely technical concern into a strategic financial discipline.

Autonomy introduces a different class of risk. As AI agents are granted the ability to act — not just recommend — the boundary between assistance and decision-making becomes blurred. Automated updates to customer records, workflow escalations, or financial adjustments can amplify errors at machine speed if not governed properly. The absence of human-in-the-loop controls in critical processes can turn isolated model inaccuracies into systemic business failures. For enterprise architects, designing where autonomy is acceptable — and where it is not — is a core governance responsibility.

Compounding these challenges is the rise of shadow AI. Business users increasingly experiment with AI tools outside sanctioned platforms, often with good intentions but little awareness of compliance or security implications. This creates blind spots that traditional IT governance models struggle to detect. AI governance, therefore, cannot rely solely on policy documents; it must be embedded into architecture, tooling, and operational oversight.

In response to this landscape, enterprise-grade AI adoption demands clear architectural principles. First, AI must be mediated through trust and control layers that enforce data classification, anonymisation, encryption, and auditability before any interaction with models occurs. AI should not be treated as a direct consumer of enterprise data, but as a service operating behind controlled gateways that make governance enforceable by design.

Second, automation must remain human-centred. AI should augment human decision-making, not silently replace it in high-impact scenarios. Architectures should explicitly define approval thresholds, escalation paths, and explainability requirements so that responsibility remains clear and defensible. Human oversight is not a limitation of AI maturity; it is a safeguard for organisational resilience.

Third, cost predictability must be engineered, not hoped for. AI usage patterns should be observable in real time, tied to business outcomes, and constrained by access controls that reflect actual value creation. Enterprise architects should collaborate closely with procurement and finance teams to model AI consumption scenarios and ensure contractual terms align with architectural realities.

Finally, AI governance must be treated as a lifecycle capability rather than a one-off initiative. Models evolve, vendors change pricing structures, regulations tighten, and business expectations shift. Governance mechanisms must continuously monitor risk, accuracy, bias, and drift, with clear processes for review, rollback, and remediation. This requires embedding AI governance into existing enterprise disciplines such as architecture review boards, security operations, and compliance assurance.

For Salesforce customers, these principles are particularly critical. As AI becomes more deeply woven into CRM and customer engagement platforms, enterprises must ensure that convenience does not come at the expense of control. AI governance should protect the organisation from unintended data exposure, financial volatility, and operational risk while still enabling innovation and productivity gains.

Ultimately, AI governance is not about slowing adoption. It is about ensuring that AI scales safely, predictably, and sustainably. For enterprise architects, the challenge — and opportunity — is to elevate AI governance to the same level of importance as security, data management, and identity. Done well, it becomes a strategic enabler that allows organisations to embrace AI with confidence, clarity, and trust rather than hesitation and regret.

What Enterprise Architects Can Learn from IKEA and Global Strategic Management

Victor Leung — Mon, 02 Feb 2026 15:45:02 +0000

Strategy is often described as a plan, but in practice it is a long-term commitment to a direction, and a discipline for decision-making. As Enterprise Architects, we sit at the intersection of intent and execution. Our role is not merely to document the strategy, but to translate it into structures, capabilities, platforms, and governance that allow the organisation to succeed over time.

IKEA provides one of the clearest examples of what a strong strategy looks like in practice. Its core idea is deceptively simple: standardized, Swedish-designed, self-assembly furniture at low cost. That simplicity is not accidental, it is the source of its power. By standardising products and processes, IKEA benefits from economies of scale and scope, driving costs down while appealing to a broad customer base. The result is a sustainable competitive advantage that competitors struggle to replicate.

What is striking, however, is not just IKEA’s strategy, but its restraint. As the company expanded internationally, it did not fundamentally alter its core strategy. Instead, it resisted the temptation to over-localise, preserving global coordination wherever possible. This is a lesson many global organisations forget: strategy dilution often happens in the name of flexibility.

Yet IKEA’s early failures in the United States and Japan are equally instructive. Success in one context does not guarantee universality. Cultural norms, living spaces, consumer expectations, and even basic assumptions, such as willingness to self-assemble furniture, varied more than IKEA initially anticipated. The company learned that global strategy is not about rigid uniformity, but about intelligent adaptation. The challenge was not to abandon the core strategy, but to adjust activities at the edges without eroding the centre.

This tension, between global efficiency and local responsiveness, is at the heart of global strategic management. For Enterprise Architects, it translates directly into architectural choices. Which capabilities should be globally standardised? Which should be locally configurable? Where do we draw the line between shared platforms and market-specific extensions? These are not purely technical questions; they are strategic ones.

Strategic management, at its core, is about achieving a sustainable competitive advantage. “Advantage” implies a superior position, “competitive” implies relevance to rivals, and “sustainable” implies durability over time. From an architectural perspective, sustainability is achieved when the organisation’s capabilities, processes, and technologies reinforce each other in ways that are difficult to imitate. Architecture becomes a strategic asset when it encodes these advantages into the operating model.

The strategy-making process is often described as analysis, development, and implementation. In reality, these activities happen simultaneously, especially in a volatile global environment. A perfectly executed analysis can become obsolete overnight, as seen during the 2008 financial crisis or the Greek debt crisis. However, this does not diminish the value of analysis. Without it, decision-making becomes reactive and fragmented.

Environmental analysis remains essential, particularly for global firms. At the macro level, political, economic, social, and technological factors shape the boundaries within which organisations operate. At the industry level, buyers, suppliers, competitors, and intermediaries determine competitive dynamics. Internally, firm resources and capabilities define what is actually possible. Enterprise Architects must be fluent across all three levels, because architecture decisions are constrained, and enabled, by each of them.

Frameworks such as PEST analysis or Porter’s Diamond Model are not ends in themselves. Their real value lies in helping leaders ask better questions. Why do certain countries consistently lead in specific industries? How do demand conditions or supporting industries amplify innovation? And critically, how do these external forces interact with our internal capabilities?

One area where global strategy and architecture intersect most sharply is in sensing and responding to change, especially technological change. Weak signals rarely appear in headquarters reports. They emerge locally, in startups, research labs, customer behaviour, and regulatory shifts. Leading multinational firms recognise this and deliberately distribute their sensing capabilities. Bayer’s use of global research centres and technology scouts is a powerful example. Local intelligence is gathered close to the source, but synthesis and decision-making remain coordinated centrally.

For Enterprise Architects, this highlights an often-overlooked responsibility: designing feedback loops. It is not enough to deploy systems, platforms, or innovation hubs across regions. Information must flow back to the centre in a form that leadership can act upon. Otherwise, decentralised sensing becomes organisational noise rather than strategic insight.

Ultimately, global strategy succeeds when an organisation strikes the right balance between coherence and adaptability. Too much central control leads to rigidity; too much local autonomy leads to fragmentation. Architecture is where this balance becomes tangible. Through shared platforms, clear capability boundaries, and explicit governance, Enterprise Architects help organisations preserve their strategic core while remaining responsive to local realities.

IKEA’s story reminds us that great strategy is not about constant reinvention. It is about clarity of intent, discipline in execution, and humility in adaptation. In a world of increasing complexity, the Enterprise Architect’s role is to ensure that strategy does not remain an abstract ambition, but becomes a living system, scalable, resilient, and unmistakably aligned with the organisation’s long-term advantage.

Why Enterprise Architecture Must Create Urgency, Clarity, and Trust in a Disruptive World

Victor Leung — Tue, 27 Jan 2026 15:26:41 +0000

When I joined the organisation as an Enterprise Architect, the expectation was clear and familiar. We would follow a disciplined, strategy-led approach aligned with TOGAF: start from vision and mission, understand business architecture, derive technology architecture, guide solutioning, and govern change through Architecture Development Method cycles. On paper, this approach creates coherence, traceability, and alignment between strategy and execution. In reality, what I observed was a set of deeply embedded organisational challenges that are far more cultural, structural, and behavioural than methodological.

One of the most fundamental issues is the absence of a shared sense of urgency. We are operating in a period of disruptive transformation—tokenisation of assets, real-time settlement, ecosystem-based platforms, and AI-driven operating models are no longer theoretical concepts but active forces reshaping industries. History has shown that organisations do not fail because they lack talent or resources, but because they fail to adapt in time. The stories of Nokia and Kodak are reminders that past success can become a liability when it breeds complacency. Urgency does not mean panic; it means a collective understanding that standing still is itself a strategic decision, and often the most dangerous one.

Instead of initiatives flowing from strategy to architecture and then to execution, many initiatives today originate from project or portfolio management channels and are passed directly to technical teams with a request to “find a solution.” Business objectives are often unclear, implicit, or reduced to a single dimension such as cost savings. Under tight timelines, teams default immediately to convergent thinking, searching for pragmatic, locally workable solutions rather than exploring the problem space. There is resistance to blue-sky thinking, scenario planning, or asking uncomfortable “what if” questions. Energy is spent refining final presentation slides rather than debating options, assumptions, or alternative futures. Design thinking may be referenced, but it is rarely practised meaningfully.

This way of working produces predictable outcomes. Solutions are optimised locally rather than globally. Short-term delivery is prioritised over long-term coherence. Architectural intent becomes reactive instead of intentional. When deadlines are driven by system end-of-life, regulatory pressure, or business launches, tactical fixes crowd out strategic solutions. Even when long-term visions are acknowledged as important, they are perpetually deferred because projects are bounded by fixed timelines and fixed resource allocations. Over time, technical and organisational debt accumulates—not because teams are careless, but because the system incentivises speed over alignment.

Another challenge is that strategy itself is often implicit. It lives in people’s heads rather than in clearly articulated, written, and shared artefacts. Long-tenured staff carry context about why decisions were made, what trade-offs were accepted, and which constraints were temporary. When those people move roles or leave the organisation, that knowledge leaves with them. What remains is a current state shaped by historical decisions and undocumented assumptions. Architects and teams are then forced to reverse-engineer intent from outcomes, a process that is inefficient, fragile, and prone to error.

Innovation is also too often assumed to be an internal activity. Large organisations tend to look inward and backward, reusing existing systems and defending sunk costs. Yet some of the most powerful signals for change come from customers. Their behaviours, frustrations, and workarounds provide direct insight into where operating models and platforms are no longer fit for purpose. When architecture and strategy are disconnected from real customer journeys, innovation becomes abstract rather than actionable.

The difficulty of first-principles thinking compounds these issues. Working-level teams rarely have a holistic view of the enterprise. They operate within functional, system, or domain boundaries and are incentivised to optimise locally. Constraints inherited from past decisions are treated as immutable truths rather than assumptions to be challenged. As a result, organisations repeatedly arrive at local optima while missing better global solutions. Enterprise Architecture exists to elevate thinking beyond these boundaries, but only if it is engaged early and given legitimacy.

Organisational hierarchy further reinforces this problem. Communication is often one-way, flowing downward, while feedback upward is filtered. People at the working level spend a disproportionate amount of time guessing the preferences of senior management rather than searching for the right answer. Risk aversion grows. Decisions are shaped by what feels politically safe rather than what is strategically sound. The organisation gets alignment on appearances, not on outcomes.

Change fatigue is another invisible but powerful force. After repeated rounds of cost cutting, restructures, and transformation programmes, staff become weary. Trust erodes when change feels constant but direction feels unclear. Without co-creating a change vision at the team level, people disengage or comply superficially. Buy-in becomes transactional, and transformation loses momentum before it delivers value.

Even the tools organisations rely on shape behaviour in unhelpful ways. PowerPoint decks and Excel spreadsheets dominate collaboration, fragmenting information across emails and versions. Context is lost, decisions are revisited, and alignment is shallow. Contrast this with the narrative-driven approach popularised by Jeff Bezos, where structured written documents force clear thinking, shared understanding, and meaningful discussion. Good tools do not merely store information; they shape how people think together.

Decision-making itself is often unclear. When choices span multiple teams and stakeholder groups, accountability blurs. Tough decisions stall or are endlessly escalated. Without clear decision rights, governance becomes slow and political. Empowerment and decentralised decision-making, supported by clear architectural principles and guardrails, enable speed without chaos.

Financial opacity further weakens strategic decision-making. Understanding total cost of ownership is surprisingly difficult. Procurement focuses on licence costs, while operational and support costs are fragmented across teams. In matrix organisations, FTE allocation is opaque, and different teams define run costs and change costs differently. Without a shared financial language, application rationalisation and investment decisions become debates about numbers rather than discussions about value.

Many organisations aspire to data-driven decision-making, yet data is often inaccessible, inconsistent, or mistrusted. Before decisions can be data-driven, data must be discoverable, well-defined, and usable by teams. Otherwise, “data-driven” remains a slogan rather than a capability.

Psychological safety is another casualty of prolonged cost pressure. After repeated reductions, people protect their own areas. Silos form as defensive mechanisms. Collaboration declines, and knowledge sharing slows. Architecture, which depends on surfacing risks and challenging assumptions, struggles to function in an environment where people fear being wrong.

Addressing these challenges requires more than refining processes or adopting new frameworks. It requires a re-anchoring of how organisations think about purpose, urgency, and alignment. Strategy must start with “why” and be explicitly articulated, not inferred. Urgency must be created without fear, linking disruptive forces such as tokenisation to real strategic choices and making the cost of inaction visible. Customer journeys should become a central organising construct, providing a shared reference point for prioritisation and investment. A frequently cited example is the digital transformation journey of DBS, highlighted by Harvard Business Review, where success was driven by strong platform foundations, partnership ecosystems, and clear ownership of end-to-end customer journeys.

Organisations must create space for divergent thinking early, before converging on solutions. Scenario planning, “what if” analysis, and option exploration should be treated as risk reduction, not overhead. Roles, decision rights, and engagement models across Enterprise Architects, Solution Architects, Data Architects, and Domain Architects need to be explicit. Financial transparency must be treated as a first-class architectural concern, with standardised definitions of run cost, change cost, and total cost of ownership. Governance should evolve from control to enablement, accelerating good decisions rather than merely preventing bad ones. Collaboration tools should privilege shared narratives and written thinking over slide decks. Data must be made available and trusted before demanding data-driven outcomes. Stakeholder engagement must be intentional, recognising that architecture succeeds through alignment, not mandate.

Enterprise Architecture is not about enforcing frameworks or producing artefacts. It is about creating clarity in complexity, continuity in change, and coherence over time. When urgency is missing, strategy fades. When trust is weak, architecture is sidelined. When purpose is shared and thinking is explicit, architecture becomes indispensable. In a world that will not wait, clarity is no longer optional—and neither is architecture.

Corporate Governance Across Emerging Markets

Victor Leung — Wed, 07 Jan 2026 14:21:05 +0000

For global investors and owner-managers alike, corporate governance in emerging markets is no longer a peripheral concern. It sits at the heart of capital allocation, risk management, and long-term value creation. China, Brazil, and South Korea offer three distinct but instructive governance stories—each shaped by history, ownership structures, and regulatory choices. Examined together, they reveal how governance frameworks evolve, where tensions persist, and what practical lessons can be drawn for investors and controlling families.

China’s corporate governance system has developed rapidly over the past three decades, largely in tandem with the country’s gradual transition from a planned economy to a market-oriented one. Early reforms in the 1990s focused on corporatizing state-owned enterprises (SOEs) and introducing stock exchanges in Shanghai and Shenzhen. Governance in this phase was heavily state-centric: boards existed, but real authority often rested with government bodies and Party committees. Over time, China introduced a modern company law, independent director requirements, audit committees, and disclosure rules broadly aligned with international practice. Today, governance is shaped by a dual structure. On the one hand, the China Securities Regulatory Commission enforces listing rules, disclosure standards, and corporate governance codes similar in form to those in developed markets. On the other hand, the State-owned Assets Supervision and Administration Commission represents the state as controlling shareholder in major SOEs, influencing board appointments, executive incentives, and strategic direction. For a major investor, the opportunity lies in China’s scale, liquidity, and improving transparency; the risk lies in understanding that control rights, political priorities, and shareholder value do not always align in the same way as in Anglo-American markets. Effective due diligence therefore requires not only financial analysis, but also a clear view of ownership, state influence, and regulatory signaling.

Brazil offers a contrasting governance journey, one driven less by the state and more by capital market innovation. Historically, Brazilian companies were characterized by concentrated ownership, extensive use of non-voting shares, and weak minority protection. In response, Brazil’s stock exchange introduced differentiated listing segments, most notably Novo Mercado, which raised governance standards beyond minimum legal requirements. These reforms demonstrated that better governance can be market-led and value-enhancing. One particularly relevant lesson from Brazil for family-owned enterprises is the role of structured family governance mechanisms, especially the family council. For a large family company now owned by two generations, a family council can serve as a formal forum to separate family matters from business management. The benefits include clearer communication across generations, agreed principles on dividends, succession, and employment of family members, and reduced risk of conflict spilling into the boardroom. It also helps professionalize decision-making without diluting family control. The costs are real but manageable: time commitment, the need for facilitation or external advisors, and the risk that poorly designed councils become symbolic rather than effective. The Brazilian experience shows that when family councils are clearly mandated, linked to but distinct from the board, and focused on long-term stewardship, they can significantly enhance both family harmony and corporate resilience.

South Korea illustrates yet another governance model, dominated by large business groups known as chaebol. Many of the country’s most prominent listed companies—such as Samsung Electronics, Hyundai Motor, and SK Hynix—are globally competitive firms with sophisticated operations, yet their governance has long been shaped by founding-family control. Samsung Electronics provides a useful example. It has strengthened formal governance practices over the past decade by increasing board independence, separating the roles of chair and CEO in practice, enhancing disclosure, and engaging more actively with international investors. At the same time, ultimate control remains closely linked to the founding Lee family through ownership structures and influence within the wider Samsung Group. This creates a hybrid governance model: outwardly aligned with global best practices, but internally anchored in family and group control. For investors, the key insight is not to assume convergence automatically means convergence in substance. Instead, governance must be assessed in terms of how effectively boards can challenge controlling shareholders, manage succession, and balance group interests with those of minority investors.

Taken together, these three cases underline a central theme in emerging-market governance: form is converging faster than substance. Codes, committees, and disclosure frameworks increasingly resemble those of developed markets, yet underlying power structures—state ownership in China, family capitalism in Brazil, and chaebol control in South Korea—continue to shape outcomes. For investors, this means governance analysis must go beyond box-ticking and focus on who really controls strategy and capital. For owner-managers, especially in family firms, the lesson is equally clear: well-designed governance mechanisms such as family councils and independent boards are not constraints, but enablers of continuity, credibility, and long-term value creation.