Dumra / MIT Sloan Management Review — How DBS Bank Makes Everyone an Innovator (2026-06-18)
DBS Bank believes innovation is critical to its survival, and it ties innovation benchmarks to every employee’s performance review. In this video interview, Group head of innovation and Future of Work Bidyut Dumra explains how DBS turned its ambition into an organizational system that scales innovation from the bottom up.
(Channel description, MIT Sloan Management Review YouTube — Leaders at All Levels series, Episode 9.)
A ~25-minute long-form interview published 18 June 2026 by MIT Sloan Management Review on its YouTube channel as part of the Leaders at All Levels series (hosts Kate Isaacs and Michael). Guest Bidyut Dumra — Group Head of Innovation and Future of Work at DBS Bank, the Singapore-headquartered bank with ~39,000 employees. Dumra joined DBS from outside banking (Cathay Pacific) and frames his mandate as “not the ability to think different but the ability to get other people to think different” across the whole organisation.
This is the wiki’s second Me, Myself, and AI-adjacent MIT SMR video-interview ingest after Philips (May 2026), but from a different MIT SMR show — Leaders at All Levels (distributed-leadership focus) rather than Me, Myself, and AI (AI-in-practice focus). It is the wiki’s richest single operator-altitude dynamic-capabilities case to date: a decade-long incumbent reinvention narrated first-person by the executive who owns the innovation system, with named mechanisms at every layer of the W&W process model.
TL;DR
Seven substantive contributions.
1. The GANDALF aspiration — sensing via competitive re-framing. In 2014, having completed its first transformation wave (2009 “Asian bank of choice”), DBS looked at the environment — “a huge flurry of fintech coming in … at the other spectrum, the large giants. You had Google come out with Google Play, Apple come out with Apple Pay. These guys had tech talent, deep pockets, distribution, reach, the customer’s attention” — and decided its real competition was big tech, not other banks. GANDALF = Google, Amazon, Netflix, DBS, Apple, LinkedIn, Facebook — “these organizations were not only our role models, but also possibly our greatest threats.” DBS deliberately inserted itself as the “D” in GANDALF. The wiki’s cleanest worked example of competitor re-framing as a sensing act — the aspiration set the yardstick (“best bank in the world by 2020”) that drove everything downstream.
2. The outsider’s edge — a “healthy dose of unreasonableness.” Dumra came from Cathay Pacific, not banking. The CEO’s stated hiring logic: “if I need bankers I’ve got 39,000 of them” — the intent was someone who could get others to think differently. Dumra’s self-description: a zero-to-one background and “a healthy dose of unreasonableness … if you do want to push for change, push for differentiation, sometimes you’ve got to look beyond reality.” The digital-mindset-crafting microfoundation, narrated as a deliberate hiring choice.
3. Managing Through Journeys (MTJ) — the horizontal operating-model redesign. Origin story: a lean PIE (process improvement event) on credit-card-loss replacement cut the cycle from 5 days to 24 hours — but when the first customer was interviewed she asked “when am I getting my passbook, my digital token, my debit card — everything I said I lost?” The light-bulb moment: “a customer is beyond a process — it’s an intent.” DBS rebuilt the operating model horizontally around customer intent instead of functional silos. A single call-center-script change (no tech change) — freezing the card, sending help-numbers, asking “do you need any help getting home?” — produced a ~90% uplift in positive feedback. Each MTJ team is “the N in a box” (MTJ-in-a-box) with a clear mandate and three accountability axes: customer happiness, business profitability, employee.
4. Mini CEOs + dynamic funding — strategic agility as a funding ritual. MTJ leaders are structured as mini CEOs — “if they’re servicing the customer, they’re the kings and queens.” DBS replaced the annual-lock-in budgeting trap with a QPR (Quarterly Performance Review): every MTJ can re-state where it’s tracking, re-prioritise, and have the tech teams + business adjust the book of work — “the funding follows suit.” Plus a slush fund for the unknown: “if the MTJs came up with something mid-year that would move the needle, they could dip into that slush fund for that extra push.” The strategic-agility microfoundation operationalised as standing rituals, not one-off reallocations.
5. The Innovation Pyramid — a four-tier portfolio that makes innovation a KPI. DBS’s framework for driving innovation through the org:
| Tier | What it is | The KPI |
|---|---|---|
| Big themes / big bets | Top-of-house bets (blockchain, metaverse, quantum) | 2–3 pursued at any time |
| Horizon 3 | Every business unit, every 2 years, must propose new products/propositions that don’t exist | new propositions every 2 yrs |
| Managing Through Journeys | Each journey must find one differentiated feature | ≥1 differentiated feature / yr |
| Entrepreneurs | Every employee has a mechanism to explore and pitch an idea | ~120–130 initiatives running |
The tipping point Dumra names is not the framework but making innovation not a choice: “all parts of the organization have a KPI.” Plus a governance flex: Horizon-3 / big-bets can launch without a business case — the business case is written retrospectively, one year post-launch — because “if I can write a business case and I know exactly what’s going to happen, I’m not really pushing the needle.”
6. The Playbook Model — top-down KPI meets bottom-up delivery, scaled from journeys to agents. DBS’s 20%-of-scorecard rule: “20% of the scorecard is dedicated to transformation, to change — right down to every team to every team member.” The cascade: MDs are trained (e.g. the 4D framework — discover, define, develop, deliver — taught over two days at the group leadership conference, with embedded startup founders and a 48-hour build-a-product Shark-Tank exercise). The KPI passes to teams; teams say “what is a journey?”; they’re told “go to the transformation team” — “when they come to us, we’re ready. We’ve got a playbook, training, a complete agenda.” The same machine was re-pointed at agentic AI: at the 5 March group leadership conference every MD built an agent right there and then, and now each has a KPI to build agents for their teams — “you end up with 250 journeys or you end up with 10,000 agents.” Capability-building (improving-digital-maturity) operationalised as a reusable top-down-meets-bottom-up engine.
7. Information gap → action gap — the headline thesis + actionable advice. Dumra’s framing for leaders whose orgs are “nowhere near as sophisticated as DBS”: “It used to be you needed an information gap before you had the action gap. The information gap is gone [with ChatGPT, perplexity, etc.]. Now you’re just left with the action gap.” The discipline DBS built around this: after every learning journey, a conscious self-reflection ritual — “within the next 3 months, have I done anything with that knowledge?” On change-fatigue and burnout: it’s “not a choice” (“if you are not changing to address the customer, someone else is going to do it”), and organisations “underestimate the receptiveness of people to change … People want to be part of something big. Don’t tone it down. Turn it up.” Closing line: “there is no right way, but there is a wrong way — and the wrong way is to do nothing.”
Hosts’ closing synthesis
Kate Isaacs and Michael close with three transferable practices: (a) the customer journey as a discipline tied to purpose — DBS’s “AI-enabled bank with a heart,” operationalised through concrete north stars like RED (ridiculously easy to deal with) and customer hours saved; (b) tying innovation performance to compensation and evaluation so accountability for the customer journey is a core part of how you’re assessed; (c) *training + tooling so teams that want to innovate can do it at scale. Their “simple first move” prescription: take one customer journey and map it end-to-end (DBS used the replacement-card journey), practise the muscle of saving customer hours, then figure out who on your side needs to be in the room to own that outcome together.
W&W cells touched
This is one of the wiki’s most complete single-source W&W operationalisations — 11 cells, each with a body twin above:
digital-sensing/digital-scouting— the 2014 environmental scan (fintech flurry + big-tech entrants) that produced the GANDALF re-framing (TL;DR #1).digital-sensing/digital-mindset-crafting— the “best bank in the world” long-term vision + the deliberate hire of an outsider for “unreasonableness” and a zero-to-one mindset (TL;DR #1, #2).digital-seizing/rapid-prototyping— the 48-hour Shark-Tank product builds with embedded startup founders; the build-an-agent-right-there sprints; design-thinking / customer-centricity tooling (TL;DR #6).digital-seizing/balancing-digital-portfolios— the Innovation Pyramid as a four-tier innovation portfolio across big bets / Horizon 3 / journeys / entrepreneurs, plus the slush fund (TL;DR #5).digital-seizing/strategic-agility— the QPR quarterly re-prioritisation + funding-follows-suit + mid-year slush-fund draws (TL;DR #4).digital-transforming/redesigning-internal-structures— Managing Through Journeys: reorienting the operating model horizontally, mini-CEO leadership, changed incentive structures and reviews (TL;DR #3, #4).digital-transforming/improving-digital-maturity— the 20%-of-scorecard transformation KPI down to every employee; the playbook + training engine; org-wide agent-building skilling (TL;DR #6).strategic-renewal/business-model— the customer-is-an-intent-not-a-process shift and the “AI-enabled bank with a heart” value-proposition renewal (TL;DR #3; hosts’ synthesis).strategic-renewal/organizational-culture— innovation is not a choice, don’t tone it down, turn it up, reframing change with excitement not fear, “the wrong way is to do nothing” (TL;DR #5, #7).contextual/external-triggers— the disruptive fintech entrants + big-tech platforms (Google Play, Apple Pay) that triggered the 2014 digital wave (TL;DR #1).contextual/internal-enablers— CEO sponsorship of the outsider hire; the dedicated transformation team as the standing enabler that holds the playbook (TL;DR #2, #6).
Roles override (roles: frontmatter explicit, overriding W&W defaults): ceo, chro, cdo, transformation-lead, innovation-lab-lead. The source’s centre of gravity is scaling innovation org-wide via KPIs, culture, and a transformation team — Dumra’s own role (Group Head of Innovation and Future of Work) maps to innovation-lab-lead + transformation-lead + cdo; the scorecard/performance-review mechanism is chro; the CEO sponsorship is ceo. The override drops the broad finance/strategy/product cluster the 11 cells would otherwise inherit, focusing on the who-owns-innovation-at-scale roles.
Linked entities and concepts
Promoted entities referenced: MIT Sloan Management Review (channel/publisher; source-count bump +1), Bidyut Dumra (interview subject; entity created on this ingest), DBS Bank (the firm; entity created on this ingest).
Concept pages this source informs (Process step 6 targets): dynamic-capabilities (the richest operator-altitude case in the corpus — sensing/seizing/transforming all narrated first-person), warner-wager-process-model (an end-to-end operationalisation of the process model from inside a banking incumbent), enterprise-ai-adoption (KPI-embedded org-wide adoption; the information-gap→action-gap framing; agent-building at scale), micro-productivity-trap (a target-firm operator escaping the trap via measured innovation obligation, radical targets, and workflow-not-task redesign).
Dangling (single-source mentions, deferred per the second-source promotion rule): Kate Isaacs (co-host of Leaders at All Levels; first appearance), Michael (co-host; surname not cleanly recoverable from the ASR — first appearance), GANDALF (the Google/Amazon/Netflix/DBS/Apple/LinkedIn/Facebook aspiration acronym; concept-mention, not an entity), Cathay Pacific (Dumra’s pre-DBS employer; one-mention).
Source-quality note
Auto-generated (ASR) transcript fetched via the youtube-transcript-skill (Playwright/DOM route). The fetch returned the transcript doubled (594 segments repeated); the duplicate tail was removed at ingest. Light ASR cleanup applied in the body to proper nouns (the ASR rendered “Bidyut Dumra” variously as “be Dumra” / “Bjude” / “Bijuz”; “Cathay Pacific” as “Cafe A Pacific”; “fintech” as “fintex”). The raw file at raw/videos/laal-ep9-how-dbs-bank-makes-everyone-an-innovator.md preserves the near-verbatim ASR with a notes: block documenting the corrections. Transcript provenance (ASR) does not feed confidence per [§Lifecycle].