MBUS 853 — Session 1: Digital Strategy

Block 1 — Course Foundations: Strategic Mindset First

Why This Course, Why Now

Every industry is at risk of disruption by digital technologies, and most organizations are trying to digitally transform. The course promise: frameworks from research and examples across industries and regions to understand digital transformation's impact on business — taught as a problem-solving, discussion-based course, not a lecture on tools.

Business Problem/Opportunity → Business Solution, informed by (not dictated by) Technology Problem/Opportunity

"Developing a Strategic Mindset" — the direction of the arrows matters. Technology should never jump straight to a technology solution without first passing through the business lens.

The Trap Slide Warns Against

The diagram in the deck draws two loops: Business Problem/Opportunity → Business Solution, and a supporting loop where Technology Problem/Opportunity feeds up into the Business Problem box. The one path marked with an X is Technology Problem/Opportunity → Technology Solution directly — solving a tech problem with a tech fix without ever routing it through business logic. That's the instinct this course is designed to correct.

From Information System to Digital Transformation

The course's foundational diagram nests three layers: at the base, classic IT (hardware, software) and IM (data, procedures, people) combine into an Information System. That IS sits between two forces — customer preferences/expectations on one side and emerging digital technologies on the other. Where those two forces meet and get actively managed, you get Digital Transformation — which in turn enables entirely new business and operational models, not just faster old ones.

Three Classic Questions, Three Modern Questions

Classic (IT-era)

What information do I need to run my business? What technology do I need to manage the information? What policies do I need to organize the information and technology?

Modern (Digital-era)

What new processes can I design and execute? What new products and services can I offer? What new businesses can I create and build?

The shift is from managing information as an internal resource to using digital capability to invent new value. DBS's marketplaces (car, property, electricity) and Climate Impact X are the "new businesses I can create" question, answered literally.

Economics of Information and Technology (Evans & Wurster)

In the physical economy, richness (bandwidth, customization, interactivity) trades off against reach (connectivity/scale) — the further you reach, the less rich the experience. Digital removes that trade-off: reach and richness both climb together. This is the underlying economic reason AI-native firms can serve a billion customers (Ant Financial) with a personalized, "rich" experience that a traditional bank branch model cannot replicate at scale.

Block 2 — What Is Digital Strategy? Definitions and Types

Digital Strategy Is Not a Separate Strategy

Digital strategy is the approach of identifying and deploying digital technologies that enable the organization's business strategy. The slides flag a recurring executive mistake: asking "What's your IT Strategy?" then "What's your Internet Strategy?" then "What's your Social Media Strategy?" and now "What's your AI Strategy?" — each version implies digital is separate from business strategy. It isn't. Most organizations also default to digital experiments, independent innovation units, or point automation — which solve a symptom but often create new long-term problems (fragmentation, shadow IT, orphaned pilots).

Digitization vs. Digitalization vs. Digital Transformation

Term	Definition	Example
Digitization	Converting analog to digital	Scanning paper loan applications into PDFs
Digitalization	Replacing manual/older-digital processes with newer digital processes for efficiency and innovation	DBS moving core banking apps to virtualized, then cloud, infrastructure
Digital Transformation	Adopting digital technologies to transform business strategy, structure, and culture for new revenue and value	DBS reorganizing into an "AI-fuelled" tech company with new marketplaces and a carbon exchange

Shorthand from the slides: we digitize data, we digitalize products/processes, and we digitally transform the business (strategy, structure, culture).

Three Types of Digital Strategy (Ross et al., MIT Sloan)

Type 1

Customer Engagement Strategy

Build loyalty and trust through omnichannel options, rapid response, and personalized relationships. DBS's IVR redesign and hyper-personalized "nudges" (30M/month in Singapore alone) are this type.

Type 2

Digital Solutions Strategy

Integrate diversified products, services, and information into solutions that add value across the product/service lifecycle. DBS's API platform (150 APIs, 50+ partners) and marketplaces fit here.

Type 3

Operational Excellence Strategy

Streamline processes and workflow systems to maximize efficiency. DBS's RPA automation of 100+ processes and the AML "Cruise" case-management system are this type.

DBS is unusual precisely because it pursued all three simultaneously rather than picking one — that breadth is exactly what "All in on AI" means operationally, and it's the reason the case is used to open a Digital Strategy course.

Block 3 — Four Expert Perspectives on Digital Transformation

Four Lenses, Same Underlying Claim

The slides deliberately present four practitioner voices with different backgrounds — marketing, technology execution, strategy consulting, and academia — to show that despite different vocabularies, they converge on one idea: digital transformation only works when it's embedded in the whole business, not run as a side project.

1 · Marketing Practitioner

David Rogers — Five Domains

Customers, Competition, Data, Innovation, Value. "Digital transformation is not about technology — it is about strategy and new ways of thinking... be faster, be easier, be everywhere, be always on."

The Digital Transformation Playbook, Columbia Business School

2 · Technology Executive

Tony Saldana — 5-Stage Model

Foundation → Siloed → Partially Synchronized → Fully Synchronized → Living DNA. "70% of digital transformations fail... the surprising answer is a lack of discipline in defining and executing the right steps."

Why Digital Transformations Fail, Berrett-Koehler

3 · Strategy Consultant

Lamarre, Smaje, Zemmel (McKinsey) — Six Capabilities

Business-led roadmap, Talent, Operating Model, Technology, Data, Adoption & Scaling. "Rewire your business so hundreds or thousands of teams can harness technology to continuously create great customer experiences."

Rewired, McKinsey & Company

4 · Business Professor

Sunil Gupta — Digital Leadership Wheel

Reimagine your business, Reevaluate your value chain, Reconnect with customers, Rebuild your organization — all circling "Digital Leadership." "You can't just create a separate digital unit... you must embed it in the DNA of your organization."

Driving Digital Strategy, HBS Press

McKinsey's Simpler Working Definition (used across the course)

A digital and AI transformation is the process — led by the CEO and top team, never finished — of developing organizational and technology-based capabilities that let a company continuously improve its customer experience and lower its unit costs, and over time sustain a competitive advantage.

Source of advantage = organizational + tech capabilities. Finality = sustained competitive advantage. Note: "it's never done" and "both [experience and cost] matter."

Block 4 — Anchoring Digital in Business Strategy Fundamentals

Strategy 101, Applied to Digital

Eisenhardt: strategy answers "Where do you want to go?" and "How do you want to get there?" Porter: "Operational effectiveness is not strategy... strategy is about making choices, trade-offs; it's about deliberately choosing to be different." Sustainable competitive advantage still comes from differentiation — being unique from competitors and valuable to customers. Digital doesn't repeal these fundamentals; it changes the toolkit available to pursue them.

Levels of Organizational Strategy

Corporate Strategy (purpose, mission, values, vision) → Business Strategy (objectives, KPIs, initiatives) → Functional Strategy (operational, financial, HR, IT, etc.). Digital strategy is not a fourth, parallel track — it's an input that now shapes the business strategy layer directly.

Traditional Alignment vs. Modern Integration

	Traditional Business-IT Strategy	Modern Business-Digital Strategy
Sequence	Corporate → Business → IT Strategy (IT projects)	Corporate → Business Strategy & Digital Transformation (roadmap + digital/IT projects)
Role of Technology	A resource that enables an already-decided strategy	A capability that actively shapes what strategy is possible
Worked Example	Business sets "increase revenue 25% via customer engagement" → IT is handed the CRM implementation as an action item	DBS didn't hand IT a CRM ticket — it made "digital to the core" one of three transformation pillars at the CEO level

Revisiting Porter's Five Forces Through a Digital Lens

Ask the classic questions (suppliers, competitors, buyers, substitutes, new entrants) but push each one further: Does digital technology affect the value proposition to our target customer? How could digital improve the way we add value in this industry? How can digital enhance capabilities that further differentiate us? For DBS, this reframes "who are our competitors" from other Southeast Asian banks to PayPal, Grab, and Ant Financial — firms with fundamentally different cost structures (26% cost-to-income ratio for disruptor banks vs. 60% for traditional banks).

Block 5 — Digital Transformation Roadmap: WHY / WHAT / HOW

A Practical Sequence for Transformation

"Transforming an established business to thrive in a world of constant digital change." The roadmap runs Current State → Future State (WHY) → Plan → Design (WHAT) → Prototype & Test → Refine/Implement (HOW), with Strategy and Change running underneath the entire sequence as a connective thread.

Why

Why do we need to transform?

Analyze environment/industry, define the business goal and strategic problem, justify via differentiation/customer-centricity/efficiency, revise business strategy to integrate the transformation.

What

What do we need to transform?

Assemble a cross-functional transformation team led by a senior business executive, define measurable success factors, run a gap analysis current-vs-future, get executive buy-in and select technology partners.

How

How do we need to transform?

Prepare phases/milestones/timeline, establish governance and a metrics dashboard, build a change-management and communication plan, map processes and assign roles.

Mapped onto DBS: WHY = declining customer-service ranking + institutional growth constraints + stagnant revenue. WHAT = the "Every Unit" mandate across all business units and even support functions (Audit, HR), led from the CEO down. HOW = the multi-year sequence — overhaul core tech (2009–2015) → GANDALF framework and early AI wins (2013–2017) → Data First and "everything goes cloud" (2018+) → All-in-on-AI scaling (200+ use cases, ADA platform).

Approach to case analysis (from the slides): Place yourself in the case's timeline — use only the facts and data given, not outside knowledge of what happened next. The recommendation must be a going-forward solution, not a critique of what the company should have done differently in the past.

Block 6 — Article: Competing in the Age of AI (Iansiti & Lakhani, HBR 2020)

The Core Claim: A New Kind of Firm

Ant Financial reached one billion users five years after launch, serving 10x more customers than the largest U.S. banks with a tenth of the employees, at a 2018 valuation of $150B — nearly half of JPMorgan Chase's. The reason: AI runs the operating decisions. No manager approves the loan, no employee gives the financial advice. Microsoft's Satya Nadella calls AI the new "runtime" of the firm.

The AI Factory — Four Components

Component	Role
Data Pipeline	Semiautomated process that gathers, cleans, integrates, and safeguards data systematically and at scale
Algorithms	Generate predictions about future states or actions of the business
Experimentation Platform	Tests hypotheses about new algorithms to confirm they have the intended effect
Infrastructure	Embeds the process in software and connects it to internal and external users

Crucial nuance: the AI driving this growth is often "weak AI" — not science fiction, not simulated human reasoning. It just needs to reliably perform tasks traditionally done by people. DBS's rule-based-plus-machine-learning anti-money-laundering system is exactly this: unglamorous, narrow, and transformative in aggregate.

Removing Limits to Scale, Scope, and Learning

Traditional operating models hit diminishing returns on scale. AI-driven models don't — network effects and learning curves reinforce each other, so value can keep climbing as users grow. When an AI-driven firm "collides" with a traditional firm serving the same customers (Amazon vs. retailers, Ant Financial vs. banks, Uber/Didi vs. taxis), the traditional firm can be overwhelmed — not through a single disruptive innovation, but through the emergence of an entirely different kind of competitor.

Cold-start caveat: AI-driven models take time to catch up to traditional-model value — Alipay took years to reach its current volume. This is why incumbent executives often dismiss the threat early, then get overtaken once critical mass hits.

Putting AI at the Firm's Core — Five Principles

One Strategy

Rearchitect every business unit on one integrated foundation of data, analytics, and software — top-down mandate, not a skunkworks.

A Clear Architecture

Centralize and standardize data assets; if not fully centralized, at least maintain an accurate catalog and explicit protection guidelines.

The Right Capabilities

Systematically hire a different kind of talent with dedicated career paths and incentives — not just a handful of data scientists bolted on.

Agile "Product" Focus

IT teams need a product-management orientation — deep understanding of use cases — not the traditional "keep the lights on" IT mandate.

Multidisciplinary Governance

Integrate legal and corporate affairs into product/technology decisions — data privacy, algorithmic bias, and cybersecurity now carry regulatory weight.

"You don't have to be a software start-up to digitize critical elements of your business — but you do have to confront silos and fragmented legacy systems, add capabilities, and retool your culture."

— Iansiti & Lakhani, on Nordstrom, Vodafone, Comcast, Visa, and other traditional-firm transformations

The Leadership Challenge: Frictionless Systems Are Dangerous

Removing operating constraints isn't purely upside. Frictionless digital systems are "prone to instability and hard to stop once in motion" — like a car without brakes. A viral signal can spread to billions before anyone controlling the network can intervene. For banks specifically, the article flags that digital banks aggregating consumer savings at unprecedented scale is a real systemic risk — directly relevant to a case about a bank going "all in" on exactly this model.

10x

Ant Financial's customers per employee vs. the largest U.S. banks

$150B

Ant Financial's 2018 valuation — ~half of JPMorgan Chase

~50%

Iansiti & Lakhani's estimate of current work activities potentially replaceable by AI systems

Block 7 — Case: DBS Bank: A Tech Company Going All In on AI

From "Damn Bloody Slow" to AI-Fuelled Tech Company

DBS (Development Bank of Singapore) began in 1968 financing Singapore's industrialization. It grew into Southeast Asia's largest bank through acquisitions — but customer service lagged, and the bank was ranked lowest in customer service among Singapore's largest banks. When Piyush Gupta joined as CEO in 2009, he launched an organization-wide digital and AI transformation, with board-level budget support.

The Motivation (Why)

Ranked lowest in customer service among Singapore's largest banks
Institutional constraints on acquisition-led growth and organic expansion across emerging Southeast/South Asia
Dwindling revenues and stagnant growth, with digital solutions seen as the new growth avenue
22,000 employees across markets, heavy reliance on outsourced technology talent (85% outsourced), making change slow and cumbersome

Transformation Timeline

Late 2009 — Every Unit Approach

Three pillars: be digital to the core, embed in the customer journey, think and act like a start-up. Applied to every unit, including support functions like Audit (fraud analysis, risk profiling) and HR (attrition prediction).

2009–2015 — Overhauling Technology

Technology and operations divisions merged into a single reporting line. Lean principles applied to core processes. A new core banking system deployed across 12 markets in 28 months. Customer Experience Council and Innovation Council established.

2013–2014 — Early AI Failures and Wins

A*STAR AI lab partnership yielded mixed results; an IBM Watson wealth-management pilot failed outright. Lesson learned: agile, minimum-viable-product execution beats big-bang delivery. Meanwhile, a predictive-maintenance model for ATMs cut downtime and out-of-cash incidents by over 90%.

2015–2017 — GANDALF and Insourcing

GANDALF acronym (Google, Amazon, Netflix, Apple, LinkedIn, Facebook — DBS as the "D") rallies the org to think like a tech titan. Talent strategy flips from 85% outsourced to a target of 85% insourced; 200 technologists hired in 2017 alone. Digibank launches in India as the country's first mobile-only, Aadhaar-onboarded bank — 6x customer growth, 12x transaction growth, same headcount over four years.

2017 — API Platform and RPA

API platform launches with 150 APIs across 20+ categories and 50+ partners (AIG, McDonald's, Foodpanda). RPA Center of Excellence formed with IBM; by 2020, 100+ complex processes automated.

2018 — Data First

Four pillars: analytics capability, culture & curriculum, enablement of data usage, technology platforms. "Everything goes cloud" policy retires reliance on physical data servers. PURE framework (Purposeful, Unsurprising, Respectful, Explainable) governs responsible data use.

2020–2023 — Scaling All-in-on-AI

90%+ of the organization on the ADA platform ("Advancing DBS with AI"); 99% of applications migrated to virtual private cloud. 200+ enterprise data use cases identified, 150+ using advanced analytics/AI, 700+ dedicated data and analytics experts. 100+ AI/ML algorithms analyzing 15,000 customer data points generate ~30 million hyper-personalized "nudges" a month in Singapore alone, lifting customer response rates 10–12%.

Results — With a Data Wrinkle Worth Flagging

The case states net profit grew 20% to US$6.02B in FY2022 (a record). Separately, in the "Rewards" section, it states net income grew from US$4,853M in 2009 to US$0.96B in 2018 to US$1.65B by 2022 — a noticeably different 2022 figure from the same case. Cite both as given rather than resolving the discrepancy — it's a useful reminder to read exhibits carefully rather than taking narrative prose figures at face value (Exhibit 5 separately shows Total Income vs. Net Profit bars by year, which is likely the source of the mismatch).

What's Next? — The Case's Closing Tension

By early 2023, the emergence of GPT-3/ChatGPT/GPT-4 and generative AI represented a "dramatic step change" that the case explicitly frames as "uncharted territory" — layered on top of 5G, IoT, and blockchain. Gupta is confident ("we're relatively well positioned... we should be able to hold our own") but the case ends on open questions: How can DBS compete with technology companies in a fast-changing marketplace? How can it keep operating like a "digital native"? What new opportunities should it bet on, and how does AI help?

Block 8 — Case Diagnostic: Issues, Decision, Position

Applying the Case Prep Protocol

Step 1 — Who and What

Decision maker: Piyush Gupta, CEO of DBS Bank (with Sameer Gupta, Chief Analytics Officer, as the internal voice of the AI platform). Core challenge: whether the "All in on AI" strategy that took DBS from "Damn Bloody Slow" to record profitability can still be a source of sustainable advantage now that (a) generative AI/LLMs are available to every competitor overnight, and (b) fintechs and digital-only banks operate at structurally lower cost (26% cost-to-income ratio vs. 60% for incumbents).

Step 2 — Five Issues Grounded in Case Facts

Commoditizing moat. DBS built its edge on proprietary AI use cases (200+) and insourced talent — but GPT-3/ChatGPT/GPT-4 give any competitor access to comparable generative-AI capability without a decade of internal build-out.
Structural cost disadvantage. Digital-only disruptor banks run at a ~26% cost-to-income ratio vs. ~60% for traditional banks like DBS, even after a 15-year transformation — branch and legacy-adjacent costs remain a drag.
Diversification risk. DBS pushed into non-bank marketplaces (Car, Property, Electricity, Carousel) and Climate Impact X; several partnerships underperformed expectations, partly due to COVID-19, raising the question of how much further to diversify versus double down on core banking AI.
Systemic/regulatory exposure. As DBS aggregates more customer data and automates more decisioning (credit, AML, personalization), the article's warning about digital banks aggregating consumer savings "in unprecedented fashion" becomes a direct risk to a bank entrusted with retail deposits.
Talent and governance scaling. The insourcing strategy (85% target) and PURE governance framework worked at the current AI scale — but generative AI raises new categories of governance risk (hallucination, IP, explainability) the PURE framework wasn't originally built for.

Step 3 — A Position

Recommendation direction: Extend the AI factory model to generative AI as a layer on top of DBS's existing proprietary data and use-case library — not a replacement for it. The durable advantage was never "having AI," it was 15 years of clean, centralized data (ADA platform, Data First) and 700+ in-house experts who know which of 250+ documented use cases a new model can plug into fastest. Off-the-shelf LLMs commoditize the interface layer; DBS's moat should shift toward being the fastest, most-governed integrator of new AI capability into a bank-grade, regulator-trusted stack — not a race to build the biggest model.

Counterargument to prepare for: One could argue DBS should narrow, not extend — pull back from marketplaces and carbon exchanges (several of which underperformed) and concentrate AI investment purely on core banking risk/cost reduction, since trust and regulatory scrutiny are existential for a deposit-taking institution in a way they aren't for Ant Financial's more loosely regulated adjacent businesses. The strongest response: DBS's marketplace ventures are optionality bets funded from a highly profitable core, not a distraction from it — the real discipline question is governance rigor (PURE, responsible data use committee), not diversification breadth.

Block 9 — In-Class Mini-Case: Monolith or SOA? (Peachtree Healthcare)

Max Berndt's Architecture Choice

Peachtree Healthcare (11 hospitals, 4,000 physicians, 1M patients, grown through M&A) has costly incompatible legacy systems — one recently crashed a major hospital entirely. Board chairman Paul Lefler wants a monolith (~$1B, proven technology, radical standardization, efficiency). CEO/surgeon Max Berndt worries brute-force standardization kills the innovative, unique culture of individual hospitals, and favors SOA (service-oriented architecture, ~$750M+, unproven in hospitals, modular, allows selective/incremental standardization).

Monolith — Case For

Proven technology, one consistent system, lower integration risk, decisive break from the legacy sprawl that just caused a shutdown.

Monolith — Case Against

$1B cost, minimal flexibility, risks alienating specialty units that legitimately need customization, single point of failure risk persists just at a bigger scale.

SOA — Case For

Modular, incremental, selective standardization, preserves innovative culture, lower upfront cost.

SOA — Case Against

Unproven in hospitals specifically, integration complexity across modules, risk of re-creating today's fragmentation with better labels.

Likely class answer: favor a hybrid — standardize the shared, high-risk core (the systems whose failure just caused a shutdown: patient records, billing, safety-critical infrastructure) as a near-monolith, and use SOA principles at the edges where hospital specialties genuinely need differentiation. This mirrors DBS's own approach: centralize core banking and data (ADA, cloud migration) while allowing business units to run their own "quasi-data-scientist" experiments on top.

"Digital Is the Strategy" — Agree or Disagree?

Position: Disagree with the literal statement, agree with the spirit. Digital is not the strategy — business strategy (differentiation, where-to-play, how-to-win) remains the strategy. But in an increasingly digital-native competitive landscape, digital capability has become inseparable from the strategy's execution, the way GANDALF-era DBS treated "digital to the core" as one of three co-equal transformation pillars rather than a bolt-on initiative. The nuance matters: mistaking "digital is the strategy" for literal truth is exactly the trap that produces disconnected "AI Strategy" documents divorced from business strategy — the mistake the slides open the session by warning against.

Block 10 — Discussion Questions & Sharp Answers

Likely Professor Questions

From the slides directly: (1) How did DBS Bank transform from being a traditional bank to an AI-fuelled company? (2) As the case concludes, what are the key issues and challenges facing DBS Bank? (3) What is your recommendation to DBS Bank going forward?

Q1: DBS insourced tech talent to build an internal "AI Factory." Was that the right capital allocation choice versus just buying best-in-class AI vendors?

Insourcing was right for DBS's specific starting point: 85% outsourced technology meant the bank couldn't move at the speed the transformation required, and vendor-built solutions (the failed IBM Watson wealth pilot) underperformed relative to internally-built, iteratively-tested tools. The Iansiti & Lakhani article's "Right Capabilities" principle backs this directly — systematic hiring and career-pathing of a different kind of talent, not a handful of consultants. The trade-off is cost and time (a decade-plus build) — which only pays off if the firm intends to keep building proprietary use cases indefinitely, which DBS clearly does (200+ and counting).

At Redamo Labs, the enterprise IAM platform faced the same insource-vs-vendor question for identity verification models — the decision hinged on whether verification accuracy was a core differentiator (it was) or a commodity function (it wasn't). DBS made the same call: personalization and risk-scoring were core to the value proposition, so they built rather than bought.

Q2: The case shows DBS entering non-bank businesses (car marketplace, electricity marketplace, carbon exchange). Is this smart diversification or mission drift?

The article's framing of "collisions" is useful here: firms built on a digital core can move across traditional industry boundaries because the underlying capability — data, algorithms, digital networks — doesn't respect those boundaries (the piece explicitly cites Alibaba and Amazon competing across retail, financial services, and health care on the same technological foundation). DBS's marketplaces are a bet that the same AI factory can be redeployed across adjacent industries profitably. The case notes these specific ventures underperformed expectations (partly pandemic-driven) — so the near-term evidence is mixed, not damning. The going-forward question isn't "stop diversifying" but "which ventures get the next round of investment given actual performance data."

Owo parallel — building a stock-valuation tool for NGX retail investors is a bet that the same data/analytics capability that works for one asset class transfers to how retail investors make decisions generally. The DBS marketplaces test the same transferability thesis at much larger scale.

Q3: Given generative AI didn't exist when DBS built its "All in on AI" advantage, does that advantage evaporate now that any bank can plug in GPT-4?

No — and this is the sharpest diagnostic question to raise. GPT-4 commoditizes the model layer, not the data layer or the governance layer. DBS's actual moat, per the case, was never "we have AI" — it was 15 years of clean, centralized, cataloged customer data (ADA platform, Data First's four pillars) and 700+ people who know exactly which of 250+ documented use cases to point a new capability at. A competitor bank can license GPT-4 tomorrow; it cannot instantly replicate a decade of insourced tech DNA, a PURE governance framework already tested against regulators, or a data lake with 15,000 customer data points per account. The advantage shifts from "who has AI" to "who can integrate new AI fastest, safest, and most profitably" — DBS is still ahead on that dimension.

This is the strongest position to open with in class — it reframes the discussion away from "is DBS's AI still special" (defensive) toward "what's the actual durable asset" (analytical), which is exactly the kind of diagnostic reframe this course rewards.

Block 11 — Participation Hooks & Taju's Edge

How to Contribute Distinctively

Team Class Contribution is 20% and graded on quality and diversity of participation, not volume from one or two voices. Session 1 has no memo due — the payoff is entirely in discussion quality.

Open Strong

If asked how DBS transformed, don't start with GANDALF or ADA — start with the sequencing: tech/ops consolidation first (2009–2015), then AI experimentation (2013–2017), then data centralization (2018+), then scaled AI (2020+). Order matters — DBS didn't scale AI until the underlying data and cloud foundation existed.

Push the Consensus

Class will likely say "DBS should keep investing in AI." Push further: the real strategic question is whether the next dollar of AI investment goes to defending core banking margins or to funding new marketplace bets — and the case gives evidence (underperforming marketplace revenue) that argues for near-term core-defense over further diversification.

Raise the Data Tension

Flag the net-income inconsistency in the case (US$6.02B vs. US$1.65B for 2022) as a live discussion point — it's a useful reminder that even well-regarded cases can have exhibit inconsistencies, and disciplined analysis means noticing them rather than reciting whichever number appears first.

Taju's Edge — Redamo Labs

Directing enterprise IAM strategy for 50,000+ users is a live version of DBS's insourcing bet: building analytics and identity-verification capability in-house because it's core to trust and differentiation, not a commodity to outsource.

Taju's Edge — Prodigy Education

Predictive and behavioral models at Prodigy (150M+ users) drove a 200% engagement increase through the same discipline DBS used: centralize data first, then layer algorithms, then run structured experimentation — not the reverse.

Taju's Edge — Owo

Owo is a live, small-scale test of the article's core claim — that AI-driven products can serve underserved markets (Nigerian retail investors) with a richness/reach combination that traditional brokerage research never offered them.

Block 12 — Reflections Journal Candidate (Concept + Example)

Session 1 Journal Draft (One of Five Required Entries)

Per the syllabus, the journal must draw its concept only from class slides (not the article or case) and its example only from professional/personal experience (not the article or case). Draft below for later refinement into the 150–200 word format.

Candidate Concept

Digital Strategy Is Not a Separate Strategy — The "Strategic Mindset" Diagram

From the slides: the correct sequence is Business Problem/Opportunity → Business Solution, with Technology Problem/Opportunity feeding into the Business Problem box — never routing straight from a technology problem to a technology solution. This captures why so many "digital strategies" fail before they start: they're actually IT strategies wearing a digital label, disconnected from an actual business objective.

Candidate Example

Redamo Labs — Resisting the Technology-First Instinct

When onboarding verification time needed to drop, the instinct inside the team was to jump straight to a technology fix — a faster document-scanning API. Reframing it as a business problem first ("why does verification take 8 minutes, and what does the customer actually need from us in that moment?") led to a different, better answer: a redesigned verification flow with analytics-driven triage, which is what actually cut time from 8 to 2 minutes. The technology fix alone would have shaved seconds, not minutes.

Block 13 — Key Takeaways (Case Debrief, from the Slides)

What to Walk Away Knowing

Successful companies treat digital transformation as enterprise-wide, not isolated technology initiatives. DBS's "Every Unit" mandate — including Audit and HR — is the clearest illustration in the case.

They invest in core capabilities — data, technology, talent — and align strategy and execution, creating momentum through innovation while maintaining operational discipline. DBS's sequencing (fix infrastructure → experiment → centralize data → scale AI) is discipline, not improvisation.

In an increasingly digital and AI-driven world, winners separate on both strategy and execution — the ability to differentiate through organizational and operational capability, and to integrate technologies so those advantages become self-reinforcing (more data → better models → more customers → more data).

Digital strategy is business strategy, informed by digital capability — not a parallel "AI Strategy" document. That's the thread connecting the strategic-mindset diagram, the four practitioner frameworks, and the DBS case itself.

Looking Ahead — Session 2: Digital Leadership

→ Session 2 (Leadership)

Session 1 asks "what is the strategy?" Session 2 (Digital Doesn't Have to Be Disruptive; GE case) asks "what leadership behaviors make or break the execution?" DBS's Gupta is the positive counterfactual to GE's leadership failures.

↔ Recurring Thread: AI Factory

The AI Factory concept (data pipeline, algorithms, experimentation, infrastructure) recurs through Sessions 3–5 (Landscape, Organization, Architecture) as the underlying technical substrate every later session builds on.

No Memo, High Discussion Weight

Memos start Session 3. Sessions 1–2 exist to build shared vocabulary and a strategic mindset before teams start writing graded case memos — participation quality here sets the tone for how the team is perceived for the rest of the course.