#007 Feb 11, 2026

What Is Disruptive Innovation?

Notes from Christensen, Raynor & McDonald — HBR, Dec 2015

Most people use "disruption" to mean a startup beat a big company. That's wrong. Christensen's theory has a specific mechanism, and the precision is the only thing that makes it useful. Strip the precision away and you're left with a buzzword that explains everything and predicts nothing.

Disruption in One Paragraph

A smaller company with fewer resources successfully challenges an established incumbent. Incumbents focus on improving products for their most demanding (and profitable) customers, overshooting the needs of everyone else. Entrants find a foothold with overlooked segments, then move upmarket until they deliver what mainstream customers need.

The key trajectory: Start at the bottom or the edges. Improve. Move up. By the time incumbents notice, it's too late.

Two Footholds

Low-End Foothold Incumbents over-serve their most profitable customers, overshooting what less-demanding ones need. Opens the door for a "good enough" product at a lower price.

Examples: Steel minimills, discount retailers

New-Market Foothold Create a market where none existed. Turn non-consumers into consumers. Serve people who were previously priced out entirely.

Examples: Personal copiers (vs Xerox), transistor radios, PCs

My play is both combined. Start new-market—create customers where none exist (like Tata Nano for two-wheeler families). Then move upmarket once the model is proven. That's the classic disruptive trajectory, and it connects directly to the access inequality problem I care about.

Sustaining vs Disruptive

Sustaining Innovation Makes good products better for existing customers. Can be incremental or breakthrough. Incumbents usually win here.

Only 6% of sustaining entrants succeeded in the disk drive industry.

Disruptive Innovation Initially considered inferior by mainstream customers. Appeals to overlooked or non-consumers. Quality improves until mainstream adopts it. Entrants win here.

Critical: A technology or product is rarely inherently sustaining or disruptive. It's the business model and trajectory that determine which path you're on.

Is Uber a Disruptive Innovation?

Why Uber Fails the Test Launched in San Francisco—a well-served taxi market
Targeted existing ride-hailers, not non-consumers
Product was better than taxis from day one (not inferior)
Started mainstream and expanded—opposite of disruptive trajectory

Where Uber IS Disruptive UberSELECT vs limousines—cheaper, fewer features, appeals to low end of limo market
This part follows the classic disruptive path upward

Christensen is technically right, but the theory is incomplete. Innovation isn't just about the trajectory (low-end up vs mainstream down). It's also about whether you're addressing an unmet need or opportunity gap. Uber addressed a massive gap in convenience, transparency, and reliability—even if it didn't follow the textbook path.

Why the Label Matters

Different types of innovation require different strategic responses:

If It's Sustaining Incumbents should respond aggressively. They usually win these fights. Ignore at your peril.

If It's Disruptive Incumbents should NOT panic-dismantle their core. Create a separate division. Let it operate independently.

The danger of calling everything "disruption": You end up applying the wrong playbook. Mixing sustaining and disruptive strategies yields inconsistent, failing results.

Four Things People Get Wrong

1. Disruption is a process, not an event
It can take decades. Steel minimills took 40 years to match integrated steel. Netflix took 10+ years to threaten Blockbuster. The first minicomputers weren't disruptive because they appeared—they were disruptive because of the path they followed from fringe to mainstream.

2. Disrupters build different business models
Healthcare: GPs use a "solution shop" model (diagnose, prescribe). Convenient care clinics use a "process" model (standardized protocols). Apple's iPhone disrupted laptops by building a platform ecosystem connecting developers to users.

3. Some disruptive innovations succeed, some don't
Success is not proof of disruption. Failure is not proof the theory is wrong. Not every disruptive path leads to triumph, and not every triumph follows a disruptive path. The theory predicts the journey, not the destination.

4. "Disrupt or be disrupted" is dangerous advice
Incumbents should NOT panic-dismantle a still-profitable business. They should strengthen core customer relationships AND create a separate division to explore the disruptive opportunity. The two must coexist—for a while.

Why Incumbents Can't Respond

The Resource Allocation Trap

In interviews with disk drive company managers, Christensen found that internal resource allocation processes prioritized sustaining innovations (high margins, known customers) while inadvertently starving disruptive ones (small markets, unknown customers).

It's not that executives are stupid. Their own systems—built to serve existing customers efficiently—are structurally incapable of funding small bets on uncertain markets. The system does exactly what it's designed to do. That's the trap.

This is a huge founder advantage. Big companies literally can't fund small bets—their own systems prevent it. That's my opening. Their resource allocation is optimized for today's profitable customers. My resource allocation is optimized for tomorrow's underserved ones.

Connecting the Dots

How disruption theory maps to what I've been studying:

Three Boxes (#003)

Disruption = Box 3. Incumbents are trapped in Box 1. Their resource allocation starves Box 3.

Emerging Markets (#003)

Tata Nano = new-market foothold. Ford's mistake = sustaining innovation applied to a disruptive context.

AI Playbook (#006)

"Don't start with moon shots" = start with a foothold. RPA first, then move up to cognitive insight and engagement.

The Sweet Spot for Founders

When you're attacking an incumbent, you don't want them fully asleep (no market validation) and you don't want them fully awake (they'll crush you).

Incumbent Asleep No validation for your market. Hard to raise money. Hard to recruit. "If it were a real opportunity, someone big would be doing it."

Sweet Spot Incumbent is aware but structurally unable to respond fast. Their systems prevent small bets. Their best people are on the profitable stuff. You have the foothold.

You want them aware but paralyzed. Their awareness validates your market. Their paralysis gives you runway.

Speed vs Trajectory

It depends on the size of the company. Apple sleeps most of the time but innovates when the noise settles down—that's a sustaining innovator's luxury. As a startup, speed still matters, especially in the AI age. The foothold can be claimed fast or it can be claimed by someone else. Direction matters more than velocity—but velocity matters more than most people think.

The Timeline Spectrum

Steel minimills: 40 years to match integrated steelmakers. Compaq vs DEC: 12 years. Netflix vs Blockbuster: ~15 years. AI disruptions: likely much faster because the enabling technology is improving exponentially.

The AI age compresses everything. The same disruptive process that took decades in steel could take years in software. Be on the right trajectory AND move fast.

Education: Wave One is Done

Wave One (Happened) Online learning: Coursera, YouTube, Khan Academy. Made education accessible. Tuition falling. Quality improving. Top 20 universities still the top 20.

Wave Two (Coming) AI-personalized learning. Credentials that aren't degrees. Skills-based hiring. The disruption isn't to classrooms—it's to the credential itself.

The first wave of education disruption already happened. Coursera, YouTube, and their peers have made content accessible. But the relative standing of top institutions hasn't changed—exactly as Christensen noted. The next wave isn't about content access. It's about something deeper: the entire credentialing and hiring system.

Bottom Line

Disruption has a specific meaning. Small entrant starts at the bottom or edges, improves, moves up. Not every industry shake-up qualifies.
Two footholds: low-end and new-market. Best play is both combined—create new consumers, then move upmarket once the model is proven.
Sustaining vs disruptive requires different strategies. Calling everything "disruption" leads to applying the wrong playbook.
Incumbent paralysis is structural, not stupid. Their resource allocation is optimized for today's profitable customers. That's the founder's opening.
Disruption is a process, not an event. It can take years or decades. In the AI age, the timeline compresses dramatically.
Don't panic-dismantle what works. Incumbents should create a separate division. Founders should want incumbents aware but paralyzed.
Business model matters more than product. Apple's iPhone disrupted laptops not through features but through a platform ecosystem.
The theory is right but incomplete. Innovation also means addressing opportunity and need gaps—not just following a trajectory.
Education wave one is done. Content is accessible. The next wave disrupts the credential itself, not the classroom.
Speed + direction = the founder formula. Be on the right trajectory AND move fast. In AI, the foothold can be claimed quickly—or lost.

How This Applies to My Journey

My target: New-market foothold first. Create customers where none exist. Solve access inequality with AI. Then move up.
My advantage: Incumbents can't fund small bets on uncertain markets. Their systems prevent it. My entire company IS the small bet.
My timeline: AI compresses the disruption process. What took steel 40 years could take an AI startup 4. Speed matters.
My playbook: Don't be Uber (sustaining from mainstream). Be Netflix (start at the fringe, improve, move up).

The connecting thread: Disruption theory validates the whole thesis—use tech to give more people access, start where incumbents can't or won't go, and move up from there.

Building

Nyantrace.ai

Observability and governance for AI agent systems. If you're building with agents, I'd like to talk.

nyantrace.ai →

← back to diary