Dynamic Capabilities
Confidence 0.95 · 13 sources · last confirmed 2026-07-09
A firm-level capability for sensing opportunities and threats, seizing them, and transforming the firm’s resource base in response to changing environments. Distinguished from ordinary capabilities (doing things right; replicable; outsourceable) by their role in governing the rate of change of ordinary capabilities. Origin: David Teece (1997, 2007).
Working definition
Per Teece (2007, as quoted in Warner & Wäger 2019):
Dynamic capabilities = “a company’s capacity to (a) sense and shape opportunities and threats, (b) seize opportunities, and (c) maintain competitiveness through enhancing, combining, protecting, and, when necessary, reconfiguring the business enterprise’s intangible and tangible assets.”
“Dynamic capabilities are about doing the right things, whereas ordinary capabilities are about doing things right.” (Teece & Leih 2016)
Key claims
The three-cluster framework (Teece 2007)
| Cluster | Function |
|---|---|
| Sensing | Scanning the external environment for trends/threats; opportunity identification |
| Seizing | Mobilizing resources to capture opportunities; new business model design |
| Transforming | Reconfiguring the firm’s asset base; renewal of structures and culture |
Microfoundations for digital transformation (Warner & Wäger 2019)
Empirically identified across 7 incumbent German MNCs and 18 strategy-consultant interviews — nine subcapabilities organized under the three clusters:
| Cluster | Microfoundation | What it does |
|---|---|---|
| Digital Sensing | Digital scouting | Tech trends; competitor screening; customer-centric trend sensing |
| Digital scenario planning | Signal analysis; future-scenario interpretation; digital-strategy formulation | |
| Digital mindset crafting | Long-term vision; entrepreneurial mindset; cultural promotion | |
| Digital Seizing | Rapid prototyping | MVPs; lean startup; digital innovation lab |
| Balancing digital portfolios | Internal/external option balance; scaling new BMs | |
| Strategic agility | Rapid resource reallocation; redirection acceptance; strategic pacing | |
| Digital Transforming | Navigating innovation ecosystems | Partner interaction; co-creation; ecosystem capabilities |
| Redesigning internal structures | CDO appointment; team-based structures; BM digitalization | |
| Improving digital maturity | Workforce maturity; digital natives; internal knowledge leverage |
Three forms of strategic renewal that result
- Business model renewal — replacing transactional product logics with relational/multi-sided value propositions.
- Collaborative-approach renewal — replacing siloed, internal-only collaboration with cross-functional and external-ecosystem collaboration.
- Cultural renewal — refreshing or replacing legacy cultures with digital-mindset / entrepreneurial cultures.
Contextual factors
| External triggers | Internal enablers | Internal barriers |
|---|---|---|
| Disruptive digital competitors | Cross-functional teams | Rigid strategic planning |
| Changing consumer behaviors | Fast decision making | Change resistances |
| Disruptive digital technologies | Executive support | High level of hierarchy |
Why digital transformation requires new dynamic capabilities
- New digital technologies (AI, cloud, IoT, blockchain) change the nature and purpose of dynamic capabilities — not merely their content.
- Organizations can now scale up/down at speed, ease, and cost not previously possible.
- The convergence and generativity of digital technologies forces incumbents to behave entrepreneurially even when entering competitively established markets.
Operator-narrated cases at mid-tier regional incumbent scale (DFI 2026)
[[2026-05-14-price-dfi-retail-asia-reinventing-how-it-sells|Scott Price’s CNBC Managing Asia interview]] (May 2026) supplies a compact case set of seizing- and transforming-cluster microfoundations operating at multi-brand multi-country retail-incumbent scale (DFI Retail Group: ~thousands of supermarket / 7-Eleven / Guardian / Mannings / IKEA-Asia / Maxim’s outlets across HK / SE Asia / mainland China). All cases are first-person CEO-narrated and have explicit named-numbers anchors:
| Microfoundation | DFI case |
|---|---|
| Balancing digital portfolios (seizing) | Sold the Singapore supermarket business for S$125M (~US$93M); closed all ~100 Mannings stores in mainland China after concluding 2,000 stores would be needed to win at scale, retaining only online presence via Chinese e-commerce platforms; redirected capital to Southeast Asia health & beauty (now the strongest segment, 1,500+ stores). |
| Strategic agility (seizing) | “We source from more than 50 countries around the world. We always have to have the ability to pivot very quickly to protect that pricing to customers.” — supply-chain pivot capability named explicitly. |
| Navigating innovation ecosystems (transforming) | The Yuu loyalty platform as a data-monetization flywheel linking millions of shoppers across DFI’s brand portfolio — “the way you protect the bottom line for shareholders is you create your own digital revenue that has a higher margin. Data is the core to that.” Vendor-insight sales + cross-segment promotion-permissioning already monetised. |
| Business model renewal (strategic renewal) | The Chinese Wellness Hub at Mannings HK — TCM-practitioner consultation + in-store health pod with basic-vitals measurement; pivot from commodity shampoo retail to functional-wellness platform. Plus the low-water rice programme in Thailand as a scope-3-emissions / value-pricing renewal (sold in stores at the same price — “our customers won’t pay a penny more”). |
| External-triggers sensing (contextual) | Named the agentic-AI personal-assistant disintermediation thesis as “what keeps me up at night” — the seller-side mirror of Ognibeni’s buyer-side warning that search-driven e-commerce will be the first format agents kill. |
The 2,000-store competitive-scale floor is particularly reusable as a Western/regional-incumbent-anchored data point for the balancing-digital-portfolios microfoundation — a public CEO articulation of what scale is required to win against established Chinese platform incumbents in their home market, a quantification the W&W literature treats only abstractly.
End-to-end practitioner operationalisation of the W&W process model at AWS-vendor altitude ( AWS London Exec Forum 2026)
Where the DFI case set anchors specific microfoundations with named-numbers worked cases at mid-tier retail-incumbent scale, Allen traces the entire W&W process arc end-to-end at AWS-Executive-in-Residence advisory altitude:
| W&W bucket | Allen’s operationalisation |
|---|---|
| Digital sensing | Anthropic labour-market report + MIT NANDA 95%-of-AI-pilots-fail framing + Nvidia SLM paper + Jevons-paradox / Schumpeterian-disruption macro-frame as the digital-scenario-planning discipline. |
| Digital seizing | The USE / COMPOSE / BUILD economic-decision framework as balancing-digital-portfolios; Brooklyn Solutions’ 4-phase iterative progression (basic → conversational → agentic → multi-agent) as rapid-prototyping; the embedded-pod model as strategic-agility. |
| Digital transforming | The hourglass-organization shape + builders-to-orchestrators role-shift as redesigning-internal-structures; data-engineers-as-context-architects as improving-digital-maturity. |
| Strategic renewal | The moats-erosion thesis as the load-bearing business-model strategic-renewal claim: the old moats (workflow embeddedness, software scale, integration lock-in, engineering complexity, IP) erode under agentic AI; replacement moats — compounding proprietary data, network effects, regulatory permission, capital at scale, physical infrastructure, time that can’t be parallelised — re-anchor sustainable competitive advantage. The bank-branch-network expansion worked example operationalises physical infrastructure as moat under the new conditions. |
| Contextual | The junior-hiring crisis (Ravio 73% European-tech entry-collapse) as external-trigger; AWS’s CFO-office partnership for opportunity-cost measurement as internal-enabler; toll-gate / ticket-culture legacy enterprise discipline as internal-barrier. |
This is the wiki’s first vendor-altitude end-to-end W&W operationalisation source — distinct from the theoretical anchor itself, from the DFI single-incumbent case set, and from advisory-firm operationalisations like McKinsey. Allen’s keynote is best read as the AWS-advisory-channel translation of the W&W process model into agentic-AI-era enterprise prescription.
Non-AI control case — industrial transformation at Rolls-Royce (Erginbilgiç 2026)
Erginbilgiç’s Rolls-Royce turnaround supplies the wiki’s first pure non-AI industrial-transformation anchor for the dynamic-capabilities lens — every prior source tagged with strategic-renewal/* cells has been AI-adoption-flavoured. The non-AI case is load-bearing because it allows the wiki to separate what’s specific to AI-era dynamic capabilities from what’s dynamic-capabilities primitives full stop.
Mapped to the Teece sense / seize / transform clusters:
| Teece cluster | Erginbilgiç’s operationalisation (Rolls-Royce 2023–2026) |
|---|---|
| Sensing | External benchmarking commissioned Sept ‘22 before the Jan ‘23 start date — “put the mirror up for the organisation. You cannot say the things you just said without data” (~2:01–2:32). Resilience-as-scenario-rehearsal (~21:14–21:32): “It’s not about actually predicting the world, it is about how your company now thinks about dealing with external shocks.” Sensing here is organisational habit of dealing with shocks, not forecasting accuracy. |
| Seizing | The Jan ‘23 burning-platform speech as the speed-of-commitment moment + the four-pillars framework (people + granular strategy + commercial discipline + performance culture, ~11:03–18:42, with two pillars explicitly named) as the resource-reallocation logic. CEO-to-CEO contract renegotiation (~13:46–15:44) as direct seizing of margin restructuring. |
| Transforming | Layer elimination without operational-people cuts (~4:24–5:22); “we eliminated layers in the organisation… no operational people left” — transforming the org structure without losing institutional capability. The new-normal-as-eased-cadence observation (~7:36–8:23) is the transformation-as-completed signal: leader demand-intensity drops because team behaviour shifts to new norms. |
The convergence with the AI-era anchors is more informative than any single case:
- Strategic-renewal/organizational-culture as the load-bearing W&W cell holds outside the digital lens. Erginbilgiç’s culture-refresh (“non-compromising mediocrity at that level kills the organisations”, ~17:42–17:57) is structurally identical to what Allen 2026’s AWS Executive Forum names at the demanding-leader-as-cadence-primitive layer — different domains, same pattern.
- Strategic-agility as a process capability (the company’s habit) rather than a forecasting capability — Erginbilgiç articulates this explicitly; Krakowski 2025’s tailored-augmentation effects are the AI-era expression of the same primitive.
- The McKinsey-named “case study in the art of corporate transformation” validator anchors the case at consulting-firm altitude without being a McKinsey publication itself — third-party validation of the underlying transformation mechanics.
The implication for the dynamic-capabilities concept: the AI-era literature (Warner & Wäger 2019, the W&W-process-model operationalisations above) is best read as digital-flavoured variations on transformation primitives that the non-AI literature has been articulating for decades. The non-AI control case is a useful corrective against over-attributing the mechanics to AI-specific causes.
The richest operator-altitude case — DBS Bank’s decade-long innovation system ( DBS 2026)
[[2026-06-18-dumra-mit-smr-dbs-everyone-an-innovator|Bidyut Dumra’s MIT SMR Leaders at All Levels interview]] supplies the wiki’s most complete single-source operationalisation of the Teece sense → seize → transform arc — all three clusters narrated first-person by the executive (Group Head of Innovation and Future of Work) who owns the system, at 39,000-employee banking-incumbent scale across a decade (2009 → 2026). Where DFI anchors specific microfoundations and AWS traces the arc at vendor-advisory altitude, DBS supplies the lived end-to-end case:
| Teece cluster | DBS operationalisation |
|---|---|
| Sensing | The 2014 environmental scan (fintech flurry + Google Play / Apple Pay) → the GANDALF re-framing (competition is big tech, not banks; “be the D in GANDALF”). Competitor re-framing as a sensing act, anchoring the “best bank in the world by 2020” yardstick. |
| Seizing | The Innovation Pyramid (big bets / Horizon 3 / journeys / entrepreneurs) as portfolio-balancing; the QPR + slush fund as strategic-agility rituals (mid-cycle reprioritisation, “the funding follows suit”); 48-hour build sprints + agent-building as rapid-prototyping. |
| Transforming | Managing Through Journeys — reorienting the operating model horizontally around customer intent (“a customer is beyond a process — it’s an intent”), mini-CEO leadership, changed incentive/review structures; the 20%-of-scorecard transformation KPI + central-transformation-team playbook as improving-digital-maturity. |
| Strategic renewal | The “AI-enabled bank with a heart” value-proposition renewal + innovation-is-not-a-choice culture (“don’t tone it down, turn it up”). |
Two reusable primitives the DBS case sharpens: (a) innovation-as-KPI — “all parts of the organization have a KPI” — the mechanism that converts a transformation aspiration into a measured org-wide obligation, structurally identical to Erginbilgiç’s performance-culture pillar but in a digital/AI-flavoured incumbent; (b) governance flex for genuine novelty — Horizon-3 bets launch without a business case (written retrospectively a year later) because “if I can write a business case and I know exactly what’s going to happen, I’m not really pushing the needle” — a concrete operationalisation of the balancing-internal-and-external-options microfoundation under uncertainty. The DBS case completes the wiki’s operator/vendor/CEO-non-AI triangulation of the concept with a fourth corner: operator-altitude, AI-flavoured, decade-long, banking incumbent.
Advisory-altitude AI-era read ( LangChain Interrupt 2026)
Andrew Ng supplies a fifth altitude — the advisor-to-the-G2000 vantage (via AI Aspire) — and frames the AI-era version of the sense/seize/transform loop crisply: sensing as continuous scanning of the coding-agent and vendor frontier; seizing as strategic-agility through optionality (≤1-year contracts, open-weight hedging, vendor-neutral observability) and portfolio-balancing (narrowing 300-idea spreadsheets to a handful of high-conviction bets, swing-for-the-fences over incremental); transforming as redesigning the whole workflow (the 10-minute-loan example) via small high-context generalist teams plus the data-architecture rework needed to feed agents. His central claim — bottom-up “thousand flowers” innovation generates point solutions; the transformation needs a complementary top-down motion to redesign the workflow — is a clean restatement of why dynamic capabilities are a system (sensing + seizing + transforming together), not a pile of point solutions. See enterprise-ai-adoption for the full treatment.
The clearest quantified case for portfolio-balancing under uncertainty — corporate venture building at McKinsey scale ( McKinsey Podcast 2026)
Where every prior case in this section supplies a single-firm or single-advisor narration of digital-seizing/balancing-digital-portfolios, Jason Bello’s McKinsey research is the wiki’s first cross-firm, quantified claim about the microfoundation itself: companies building three or more ventures simultaneously dramatically outperform those that try once, and the average cost to reach break-even on a new venture fell from ~$125M (2024) to ~$77M (2025) — a measured trend, not a single narrated case.
Two mechanisms sharpen the microfoundation further:
- Milestone-tranche funding as the seizing-cluster funding primitive. Splitting a venture’s total investment horizon into milestone-gated chunks (e.g. three 3-month tranches within a 9-month goal), releasing the next tranche only after checking whether the milestone was met, is a portable, generalized version of the mechanism DBS’s QPR + slush fund system operationalises at one specific bank.
- Fact-based, blame-free culture as the condition that makes fast pivots possible. “If the facts tell us our product stinks, so be it… there’s no fingerpointing” is Bello’s articulation of
strategic-renewal/organizational-culture— structurally the same claim Carroll’s academic culture-as-social-control-system theory predicts, and the same pattern Erginbilgiç’s non-digital performance culture exhibits.
Bello also names the corporate-vs-startup asymmetry directly: corporate venture builders spend far less time fundraising (the board and leadership already know them) and start with incumbent assets (existing customer base, untapped proprietary data/IP) that independent founders must build from zero — a structural explanation for why the portfolio-balancing microfoundation is more affordable for incumbents than for standalone startups attempting the same “multiple shots on goal” strategy.
Small-team, real-time redesigning-internal-structures — a practitioner’s own lived case ( AI Native DevCon 2026)
Where DBS supplies digital-transforming/redesigning-internal-structures at 39,000-employee banking-incumbent scale over a decade, Hannah Foxwell’s talk supplies the small-team, real-time instance: her own 2-person startup ran out of planned work by lunch on day one, forcing an immediate structural response (relearning ruthless prioritization, thinking further ahead, and — echoing DBS’s mini-CEO pattern at radically smaller scale — most subsequent time going to platform engineering and reliability rather than feature work).
Two mechanisms sharpen the microfoundation:
- Team-ratio experimentation as the redesign lever. Some teams trying 2 developers : 1 PM (vs. the traditional 6-8 : 1); Andrew Ng reportedly proposing the inverse at Davos (2 PMs : 1 developer), on the logic that decision-making speed, not coding speed, now bounds a single developer’s usable backlog. New role patterns — the vibe-coding product manager, the forward-deployed engineer, the product engineer — all reduce the same structural distance between the person who understands the problem and the artifact that solves it.
- “Minimum viable human” as a structural floor. An agent can’t hold an on-call pager — sustainable rotation (no one on-call more than 50% of the time, always primary + secondary) sets a hard lower bound on team size independent of how much coding velocity agents supply. A concrete, quantifiable counter-weight to unbounded headcount-reduction narratives.
Foxwell’s talk also touches strategic-renewal/organizational-culture directly — questioning mandatory code review as unsustainable at AI-authored-code volume, and citing Sophie Weston’s “broken comb” (not T-shaped) framing for what career depth should look like when a single generalist now owns more surface area.
Debates and supersession
- Sensing-as-prediction vs sensing-as-shock-readiness. The Teece (2007) framing of sensing emphasises opportunity and threat detection — close to forecasting language. Erginbilgiç 2026 argues against the prediction-framing: “It’s not about actually predicting the world, it is about how your company now thinks about dealing with external shocks” (~21:14–21:32). Warner & Wäger 2019’s
digital-scenario-planningmicrofoundation is closer to Erginbilgiç’s shock-readiness framing than to pure forecasting. No supersession; the productive tension is between sensing as accuracy and sensing as response capability. The wiki currently treats them as compatible (sensing must produce both signal-detection and the organisational habit of responding to signals). - Digital vs non-digital scope of W&W cells. Warner & Wäger 2019 derives its cell vocabulary from digital transformation case studies; Erginbilgiç 2026 is a pure non-digital case that nonetheless maps gracefully onto
strategic-renewal/organizational-culture,digital-seizing/strategic-agility, andcontextual/internal-barriers. The cells stretch outside the digital lens with the digital-mindset clause optional. No supersession; the wiki’s working hypothesis is that W&W cells are transformation primitives whose digital-flavour reflects the empirical setting of the original 27-interview sample rather than a load-bearing scope restriction. - Vendor-altitude vs operator-altitude vs case-study altitude. AWS supplies vendor-altitude operationalisation; DFI supplies operator-altitude case material; Erginbilgiç 2026 supplies CEO-altitude case material at a non-AI incumbent. No contradiction across the three; the wiki triangulates the dynamic-capabilities concept across all three altitudes deliberately.
Related concepts
- enterprise-ai-adoption — AI adoption is a contemporary instantiation of digital transformation; the same sensing/seizing/transforming logic applies, with AI-specific subcapabilities.
- generative-ai — extends the digital-transformation context; GenAI is a current sensing/seizing target for incumbents.
- automation-vs-augmentation — a strategic-deployment choice that lives within seizing capabilities.
- strategic-foresight — sensing-cluster microfoundation; FTSG-style methods are sensing tools.
- systems-thinking — adjacent lens for transforming-cluster decisions about flows and ecosystem boundaries.
- MIT CISR Four Stages — staged-maturity view of digital/AI transformation; complementary to the dynamic-capabilities lens.
- Tin Man — adjacent framing of org design under environmental change.
Open questions
- The Warner & Wäger study is from 2019, pre-GenAI. How do the nine microfoundations need to be updated for the 2026 GenAI context? (Open question; possible synthesis topic.)
- Cross-source mapping: MIT CISR’s Four S (Strategy/Systems/Synchronization/Stewardship) and the dynamic-capabilities framework appear to overlap substantially in scope but use different vocabularies. Would benefit from a future synthesis page.