Kai Roses

Context

In essay 11, I explored the foundational concept of business competency, drawing on C.K. Prahalad and Gary Hamel’s groundbreaking 1990 Harvard Business Review paper2. Their metaphor of the corporation as a tree serves as a vital framework for understanding how organizations maintain long-term stability. The roots, which represents core competencies, provides the essential sustenance that allows the trunk (core products) and the leaves (end products) to flourish. The defining challenge for business leaders in 2026 is avoiding the trap of surface-level AI. If a company simply bolsters its leaves, such as slapping a generative AI chatbot onto its consumer facing application, competitors will copy it within weeks, resulting in zero sustainable market differentiation. To build a true competitive moat in the AI economy (where value creation, capture, and distribution is powered by AI), organizations must also cultivate AI at the roots.

In Essay 2, my goal is twofold: first, to make identifying a firm’s core competency as intuitive as identifying its end products; and second, to use business core competency as a guide to a firm’s AI implementation strategy.

The empirical basis for this essay is my analysis spanning organizations with annual revenue from $80M to $300B+, covering diverse sectors from Energy and Healthcare to Financial Services and Logistics. I began with an initial pool of over 200 companies, distilling them into a final cohort of 90+ organizations across 28 industries.

From this research, three strategic imperatives emerged. These provide a recommended sequencing for an AI implementation strategy that generates a true, compounding competitive advantage.

Approach

Inclusion Logic: I included diversified software leaders like Adobe and Salesforce alongside non-tech enterprises to compare how varying business models anchor AI in their roots and leaves to remain competitive. To focus on how established businesses are evolving into the AI economy, I excluded AI-native startups (e.g., Cursor) and foundational AI infrastructure providers (e.g., Google, Nvidia, OpenAI).
Data Sourcing: I utilized Gemini Deep Research to aggregate and verify data from public sources, including annual earnings reports, investor presentations, and verified press releases.
Limitations: This study skews toward mid-to-large-cap firms with transparent public disclosures. Privately held firms or those without documented AI initiatives are not represented.

The remainder of this essay is broken into two parts:

I. Identifying Business Core Competency

II. Business Core Competency as a Guide to AI Implementation Strategy

I. Identifying Business Core Competency

Core competency is often invisible to the naked eye, yet it dictates consumer behavior. It is the underlying mechanism that explains why you are willing to pay a premium for a product you could get cheaper from a competitor. Core competency is the quiet advantage that registers in a consumer’s mind, creating a high switching cost even when alternatives exist. Consider these three examples:

FedEx (Routing Velocity): When a package must arrive overnight, you inherently trust FedEx. This is not because they have the best cardboard boxes (the end product), but because their core competency is unparalleled routing velocity. You are paying for the invisible, global logistical choreography that guarantees on-time delivery.
Starbucks (Habitual Loyalty): In an unfamiliar city, you choose Starbucks for a reliable place to meet or work. They don’t just sell customized beverages; their core competency is habitual loyalty. They have engineered a predictable third-place experience that makes a store in Tokyo feel as safe and familiar as one in New York.
Visa (Systemic Trust): When you swipe a card in a foreign country, you expect the transaction to clear in milliseconds without your identity being compromised. You aren’t paying for the plastic; you are paying for systemic trust. Visa’s competency is the invisible, high-speed risk engine that validates billions of transactions while simultaneously hunting for fraud.

To further build on these examples, here are the core competencies that emerged from my analysis. While firms may possess multiple competencies, these core competencies represent their primary roots driving their market position:

Core Competency -- Representative Businesses

Operational velocity & efficiency -- McDonald’s, FedEx, Walmart
Supply chain & logistics -- Walmart, Nike, PepsiCo, P&G
Autonomy & perception --John Deere, Caterpillar, Lockheed, Ford, GM, Rivian
Risk & compliance -- Bank of America, Visa, Amex, PayPal, State Farm, Met Life
Trust & safety -- Airbnb, Bumble
Personalization & loyalty -- Starbucks, Marriott, Sephora
Portfolio / brand curation -- LVMH, Berkshire
Fast data orchestration --JPMorgan, Salesforce, RELX, Thomson Reuters
Content & algorithm --ByteDance, Tencent, Spotify, Pinterest
Scientific discovery -- AbbVie, AstraZeneca, Novo Nordisk, Pfizer, Oura
Creative & design -- Adobe, Canva, Louis Vuitton (LVMH)
Human-first / artisan -- Hermès, Chanel

A core competency is not a static strength; it is a regenerative engine. Because these competencies are built on corporate-wide learning rather than siloed tasks, they allow companies to transcend their original niche and dominate entirely new categories.

Historically, Honda’s mastery of powertrain development enabled a seamless move from motorcycles to cars, lawnmowers, and generators. Similarly, Sony’s expertise in miniaturization allowed it to lead in disparate markets, from portable radios to the Walkman and digital cameras.

Today, this regeneration superpower continues to define market leaders like John Deere. While synonymous with tractors, Deere’s true core competency is machine perception. Decades of studying how equipment sees and interacts with soil allowed them to pivot from traditional manufacturing to high-tech robotics. Their See & Spray™ technology, treating every plant as a distinct data point, is the result of a 180-year-old firm evolving into a world-class computer vision powerhouse.

Adobe’s edge isn’t just PhotoShop; it is the orchestration of creative workflows. By mastering how creators think, layer, and iterate, they successfully navigated the leap from desktop publishing (PostScript) to a dominant cloud-based subscription ecosystem. This deep understanding of the design language allowed them to expand into document management (Acrobat) and marketing analytics, remaining central to the creator economy across print, web, and mobile.

II. Business Core Competency as a Guide to AI Implementation Strategy

With a foundational understanding of business core competency, we can now examine the structural evolution of the modern corporation as it adapts to the AI economy. This evolution is defined by the depth at which AI is integrated into an organization, distinguishing between the AI at the roots that power its core competencies and the AI at the leaves that define its end products.

AI at the roots involves developing fundamental internal infrastructure, proprietary data models, and specialized hardware and software investment that provide long-term stability.
AI at the leaves involves developing user facing products, incremental features, or UI enhancements. While essential for user experience, AI at leaves rarely constitutes a sustainable moat when implemented in isolation.

In the rush to join the AI revolution, many organizations have deployed AI at the leaves, such as basic generative chatbots or automated email summaries, without nurturing the roots of internal data velocity, proprietary modeling and infrastructure required to sustain them. This tactical adoption often fails; leaf-level applications are easily replicated by any competitor with access to the same off-the-shelf software and foundational models.

Sustainable differentiation is found in root-level integration, where technology is woven into the very mechanics of how the organization functions. For instance, an enterprise that develops proprietary neural networks to manage global logistics in real-time possesses a root competency embedded in its operational history and human capital. These core competencies are enhanced as they are shared across the organization via many different AI implemented at the leaves, creating a compounding advantage that is nearly impossible for outsiders to dismantle.

There are four root patterns, expressed through four leaf patterns that offer a blueprint for AI implementation. Any firm can apply this strategy by leading with their core competency, then selecting the specific AI at the roots and at the leaves that reinforce their foundations.

AI at the Roots

The Brain (Predictive Decision Engines): These AI implementations act as the company’s intelligence, moving a business from guessing to knowing. When Visa processes a transaction or Spotify suggests a song, they aren’t using static rules; they are leveraging deep-seated roots in risk scoring and preference matching. These live, probabilistic engines serve as high-velocity systems capable of making millions of accurate decisions per millisecond.
The Senses (Perception and Physical Autonomy): For companies operating in the world of dirt, steel, and atoms, AI must do more than predict, it must see. Consider John Deere or Caterpillar. Their AI at the roots involves computer vision and edge-AI allow a tractor to distinguish a weed from a crop in real-time. By building roots that see, these 180-year-old manufacturers have developed a level of physical autonomy that software-only competitors cannot replicate.
The Central Nervous System (Platforms and Governance): This is the connective tissue. In global enterprises like LVMH or Lockheed Martin, the challenge isn’t building one tool; it’s ensuring a thousand tools can communicate without breaking. These AI at the roots implementations provide the MLOps, governance, and data orchestration that allow AI to scale from a laboratory to a workforce of 100,000. Without this connected intelligence, the tree becomes a collection of disconnected, withering branches.
The Laboratory (Discovery and Generative Design): These are the smart roots of creation, used to reach into the unknown. In the laboratories of AstraZeneca or Pfizer, AI at the roots in genomic modeling design molecules in-silico before they ever reach a petri dish. Similarly, Adobe’s Firefly isn’t just a fun tool; it is a proprietary intelligent root in creative orchestration that keeps the company at the center of modern design.

AI at the Leaves

While the roots provide stability, the leaves are the visible, often magical features that translate internal intelligence into consumer value.

The Voice (Conversational & Agentic Assistants): Whether it is Erica at Bank of America, Rufus on Amazon, or a voice-activated cockpit in a Rivian, you are experiencing AI at the leaf. These assistants turn complex, fragmented backend data into simple, intuitive human dialogue.
The Oracle (Predictive Insights & Diagnostics): These tools provide the ability to foresee. Whether it is GM OnStar alerts, Lockheed fleet health notifications, or Oura Ring wellness summaries, they tell a customer something is wrong before they even feel it. They distill massive computing power into a single, actionable insight: Your car needs a tune-up or your recovery score is low.
The Concierge (Hyper-Personalization & Discovery): Connecting a customer to exactly what they want in a sea of infinite choice is a superpower. We see this in Spotify’s AI DJ, Pinterest’s visual search, and Sephora’s perfect shade finders. The AI at leaf takes the intelligence generated using vast data available at the roots and narrows it down to a single, perfect recommendation, saving the customer’s most valuable resource: time.
The Apprentice (Creative Co-pilots & Design Tools): This is where AI acts as a high-speed assistant that manifests human intent instantly. When a designer uses Adobe’s Magic Eraser or Canva’s Generative Fill, the AI is the apprentice executing the vision. Similarly, Dior and Chanel use digital store experiences to help customers visualize products in their own lives, turning a simple transaction into an act of creative discovery.

Three Strategic Imperatives

Beyond the distinction between AI implemented at the roots and AI implemented at the leaves, my analysis of these 90+ organizations reveals three strategic imperatives for building a true competitive advantage:

1. The Root-Leaf Symbiosis

Competitive advantage is not found in the mere presence of AI, but in the reinforcement between AI at roots and leaves. A leaf without a root is a gimmick; a root without a leaf is an R&D project. Sustainable value requires both: the hidden engine and the visible interface working in lockstep.

2. Compressing the Feedback Loop

The most successful firms use AI to bridge the gap between a raw data signal and a decisive action in milliseconds.

Logistics: Demand signal → Routing optimization → Consumer delivery
Healthcare: Biomarker data → Risk insight → Proactive care
Finance: Transaction attempt → Fraud signal → Authorization decision

3. Data Architecture Designed for AI

Market leaders invest in data architecture, orchestration, governance, and multi-use portfolios where a single, powerful root, such as a proprietary demand model, feeds a dozen different leaves, from dynamic pricing to inventory management and targeted marketing.

The Recommended Sequencing

Building an AI-driven moat doesn’t happen overnight, here is the recommended sequence for implementation:

Anchor AI at the Root: Select one primary AI at the root implementation that aligns directly with your organization’s core competency (e.g., if you are a logistics firm, start with a routing engine, not an HR chatbot).
Fund the Pair: Never fund a root in a vacuum. Always fund one root platform block and one leaf delivery block together to ensure the technology has a visible, value-generating surface.
Define the KPIs: Measure success across four dimensions: Value (ROI), Velocity (Speed to action), Quality/Risk (Accuracy/Safety), and Adoption (User engagement).
Run a 90-Day Sprint: Focus on daily iterations, rapid feedback capture and measurable outcomes for a well-defined pilot group. If there are no outcomes in 90 days, pivot.
Codify for Scale: Before starting the next wave, codify your reusable assets, data pipelines, model evaluations and monitoring, workflows, and governance guardrails. This ensures the next AI implementation at the leaf will be cheaper and faster.

Closing

AI is no longer a peripheral technology but a structural necessity. Successful implementation requires aligning AI at roots with the company’s core competency, whether that is operational velocity (Walmart), safety and assurance (Lockheed Martin), or risk management (Visa), and prioritizing the appropriate AI at the leaves that best represents a company’s strategies and consumer needs.

As the landscape shifts toward agentic systems and autonomous hardware, the divide will widen: firms with proprietary AI at roots will build compounding advantages, while those relying on generic vendor tools with shallow AI at leaves risk falling into a commodity trap.

Next Month: I will go deeper into the internal frictions, cultural hurdles, and technical challenges that exist when trying to implement AI at a legacy organization, and the strategies to overcome them.

Thank you for reading my essay. If you found it insightful, please comment, share, like.

Explore everything Kai Roses offers through natural conversation in 40+ languages.

AI Competency | Essay #2

KAIROS