Contents
Unbundling is the strategic act of extracting a single feature, function, or service from a bundled offering and building an entire company around it — targeting the customers who overpay for the bundle when they only want the slice.
Section 1
How It Works
Every mature industry eventually produces bundles. Newspapers bundle classifieds, weather, sports scores, opinion columns, and investigative journalism into a single product. Banks bundle checking accounts, mortgages, wealth management, and wire transfers under one roof. Universities bundle lectures, credentialing, social networks, career services, and research into a four-year degree. The bundle exists because, at some point, it was cheaper to distribute everything together than to distribute each piece separately. Unbundling happens when that distribution cost advantage disappears.
The cognitive shift is this: stop looking at a product and start looking at the jobs inside it. A newspaper isn't one product — it's fifteen products stapled together. Each of those jobs has a different customer, a different willingness to pay, and a different competitive bar. The unbundler identifies the single job where the bundle is weakest — where customers are either overpaying for something they don't fully use, or underserved because the bundle optimizes for the average rather than the specific — and builds a standalone product that does that one job dramatically better.
The underlying mechanism is the disaggregation of distribution economics. Bundles form when distribution is expensive and scarce (printing presses, branch networks, physical campuses). When the internet collapses distribution costs to near zero, the economic rationale for the bundle weakens. Suddenly, a solo writer on Substack can reach the same audience as a columnist at The New York Times, without subsidizing the foreign bureau or the crossword puzzle. A fintech app can offer commission-free stock trading without maintaining a network of financial advisors. The bundle's overhead becomes the unbundler's margin.
— Jim Barksdale, former CEO of Netscape"There are only two ways to make money in business: one is to bundle; the other is unbundle."
This is not a one-directional process. Unbundling and rebundling are cyclical. Craigslist unbundled newspaper classifieds. Then dozens of vertical startups — Indeed for jobs, Zillow for real estate, Tinder for personals — unbundled Craigslist. Eventually, some of those verticals will rebundle adjacent services. The strategic question is always: where in the cycle is the industry you're targeting, and which direction is the value migrating?
Section 2
When to Use This Framework
✓
Best Conditions for Unbundling
| Dimension | Ideal conditions |
|---|---|
| Founder profile | Product-obsessed operators who can identify the single highest-value job inside a bundle and build a 10x better experience around it. Domain expertise in the incumbent industry is a major advantage — you need to know which seams are load-bearing and which are cosmetic. |
| Stage | Ideation through Series A. Unbundling is primarily an opportunity-identification framework. It's most powerful when choosing what to build and how to position it. Less useful once you're already scaling and need to think about rebundling or platform expansion. |
| Market conditions | Best when an incumbent bundle is showing signs of strain: declining customer satisfaction, rising prices without corresponding value, regulatory pressure to open up, or a technology shift (mobile, cloud, AI) that changes distribution economics. The bundle should feel bloated to its own customers. |
| Competitive environment | Ideal when the incumbent is structurally unable to unbundle itself — because the bundle cross-subsidizes unprofitable but strategically important components, or because cannibalizing the bundle would destroy the incumbent's core business model. |
| Technology catalyst | A distribution shift has occurred or is occurring — the internet, mobile, APIs, AI — that makes it economically viable to deliver a single component of the bundle as a standalone product at scale. Without this catalyst, the bundle's distribution advantage holds. |
| Inputs needed | Deep customer research on which parts of the bundle users actually value, pricing analysis of the incumbent, competitive mapping of other unbundlers targeting the same bundle, and a clear thesis on why the standalone product can sustain a business (not just a feature). |
The framework is particularly potent right now because AI is creating a new distribution cost collapse comparable to what the internet did in the 2000s. Services that previously required human expertise bundled inside a firm — legal review, financial analysis, customer support, code review — can now be delivered as standalone AI-powered products. Every professional services firm is a bundle waiting to be picked apart.
Section 3
When It Misleads
⚠
Failure Modes & Blind Spots
| Blind spot | What goes wrong |
|---|---|
| The slice is a feature, not a business | You unbundle a component that customers value but won't pay for independently. The feature only had perceived value because it was part of the bundle. Standalone, it can't sustain pricing power, retention, or a viable business model. Many "unbundled" products are really just features the incumbent could ship in a sprint. |
| Cross-subsidy blindness | The bundle's pricing obscures the true cost of the component you're unbundling. Banks offer free checking because they make money on mortgages. If you unbundle checking, you inherit the cost without the cross-subsidy. You need an independent monetization path the bundle never needed. |
| Incumbent retaliation | The incumbent notices the unbundler, improves the targeted component, and uses the rest of the bundle as a moat. Google unbundled many standalone tools — then bundled them into Workspace. Microsoft responded to Slack by bundling Teams into Office 365 for free. The bundle fights back. |
| Winner-take-all dynamics in the slice | Unbundling a popular component attracts dozens of competitors to the same slice. Job boards unbundled from newspapers — and then Indeed, LinkedIn, ZipRecruiter, Glassdoor, and fifty others all competed for the same job-listing slice. The unbundling thesis was correct, but the resulting market was brutally competitive. |
| Premature unbundling | The distribution cost shift hasn't actually happened yet. You're unbundling a component that still benefits from being inside the bundle because customers prefer the convenience of one-stop shopping, or because the infrastructure to deliver the standalone product at scale doesn't exist. Timing matters as much as the thesis. |
| Ignoring the rebundling clock | You successfully unbundle one component — then a new entrant rebundles your product with adjacent services and offers a more compelling package. The unbundler becomes the unbundled. If you don't have a plan to expand or defend, you're building a transitional business. |
The single most common mistake is confusing customer complaints about a bundle with willingness to pay for a standalone alternative. Users complain about cable TV bundles constantly — but when given the option to buy individual channels, many discovered that à la carte pricing was actually more expensive than the bundle. The unbundling thesis requires not just dissatisfaction with the bundle, but a viable standalone economic model for the slice.
Section 4
Step-by-Step Process
Step 1 — MapTools: Jobs-to-be-Done interviews, value chain analysis, customer journey mapping, NPS breakdowns by feature
Decompose the bundle into its constituent jobs
Take the incumbent product or service and list every distinct job it performs for the customer. A traditional bank performs at least twelve: store money, transfer money, earn interest, borrow for a home, borrow short-term, invest for retirement, get financial advice, process payments, manage payroll, issue cards, provide insurance, and verify identity. Each of these is a potential unbundling target. Map them all before choosing.
Step 2 — ScoreTools: Willingness-to-pay surveys, competitive analysis, incumbent pricing teardowns, customer satisfaction data
Identify the highest-value, most underserved slice
For each job, evaluate three dimensions: How much do customers value this specific job? How poorly does the bundle serve it? And can it sustain an independent business model? The sweet spot is a job that customers value highly, that the bundle delivers poorly (because it optimizes for the average), and that has a clear monetization path as a standalone product. Score each job on these three axes and rank.
Step 3 — ValidateTools: Landing page tests, pre-sale campaigns, concierge MVPs, pricing experiments, competitive benchmarking
Confirm standalone demand and willingness to pay
Before building, confirm that customers will actually pay for the unbundled component on its own. Run a landing page with pricing. Offer a concierge version to 20–50 users. The critical question is not "do people want this?" (they probably do — it was part of a successful bundle) but "will people pay enough for this alone to sustain a business?" If the answer requires advertising revenue or cross-subsidies of your own, you may be building another bundle, not an unbundled product.
Step 4 — BuildTools: Rapid prototyping, user testing, competitive teardowns, design sprints
Create a 10x better experience for the single job
The unbundled product must be dramatically better at its one job than the bundle ever was. Not incrementally better — categorically better. Robinhood didn't offer slightly cheaper trades than Charles Schwab; it offered zero-commission trades with a mobile-native interface that made Schwab's app feel like a relic. The performance gap must be wide enough that customers are willing to leave the convenience of the bundle.
Step 5 — DefendTools: Network effects analysis, switching cost design, data flywheel architecture, expansion roadmap
Build a moat before the rebundling wave arrives
The moment you succeed, two threats emerge: the incumbent improves its version of your slice, and new entrants try to rebundle your product with adjacent services. Your defense is to build switching costs, network effects, or data advantages that make your standalone product increasingly hard to replicate. Alternatively, begin your own controlled rebundling — expanding into adjacent jobs from a position of strength in your core slice.
Section 5
Questions to Ask Yourself
Discovery
Which large, established products or services in my target industry are actually bundles of distinct jobs?
Which component of the bundle generates the most customer complaints or has the lowest satisfaction scores?
Has a distribution cost shift (internet, mobile, AI, APIs) recently made it viable to deliver this component independently?
Is the incumbent structurally unable to unbundle this component without cannibalizing its core business?
Validation
Would customers pay for this component as a standalone product — and at what price point?
Can the unbundled product sustain a business on its own, or does it depend on cross-subsidies that only a bundle can provide?
How many other startups are targeting the same slice of the same bundle right now?
Is the TAM of the standalone slice large enough to build a venture-scale business, or am I carving off a niche that caps at $10M ARR?
Execution
Can I deliver a 10x better experience for this single job compared to how the bundle delivers it today?
What happens when the incumbent notices me and improves their version of this component?
What is my moat after the initial unbundling — network effects, data, brand, switching costs, or regulatory advantage?
Do I have a rebundling strategy — a plan to expand into adjacent jobs once I own the core slice?
Timing
Am I early enough that the bundle still dominates, or late enough that the unbundling opportunity is already crowded?
Is the customer behavior shift (e.g., mobile-first, remote work, AI adoption) that enables this unbundling still accelerating or plateauing?
How long before a rebundler emerges to aggregate the unbundled pieces into a new, superior bundle?
Section 6
Company Examples
S
Substack
Unbundled individual writers from the newspaper and magazine bundle
Traditional media bundled dozens of writers, columnists, and reporters under a single masthead. Readers paid for the bundle even if they only cared about one or two voices. Substack unbundled the individual writer — giving them direct distribution via email, direct monetization via subscriptions, and full ownership of their audience. By 2023, Substack reportedly had over 35 million active subscriptions, with top writers like Emily Atkin and Heather Cox Richardson earning seven figures annually. The key insight was that the internet eliminated the distribution advantage of the masthead, making the individual writer — not the publication — the unit of value. The bundle's overhead (editors, offices, printing) became a tax on the writer's earnings rather than a necessary cost of reaching readers.
R
Robinhood
Unbundled stock trading from the full-service brokerage bundle
Full-service brokerages like Merrill Lynch and Charles Schwab bundled trading, research, financial advice, retirement planning, and wealth management into a single relationship — and charged $7–$10 per trade as part of the package. Robinhood identified that a large segment of customers — younger, mobile-native, self-directed investors — only wanted the trading component. By offering commission-free trades through a mobile-first interface, Robinhood attracted over 23 million funded accounts by 2023. The company monetized through payment for order flow and premium subscriptions (Robinhood Gold) rather than per-trade commissions. The unbundling forced the entire brokerage industry to drop commissions — Schwab went to zero-commission trades in October 2019, triggering its merger with TD Ameritrade.
C
Coursera
Unbundled individual courses from the four-year university degree bundle
A university degree bundles lectures, credentialing, social networking, career services, campus experience, and research access into a single four-year, six-figure package. Coursera unbundled the lecture component — partnering with universities like Stanford, Yale, and the University of Michigan to offer individual courses and certificates online. By 2023, Coursera had over 136 million registered learners. The critical adaptation was recognizing that many learners wanted specific skills (data science, machine learning, project management) without the four-year commitment or the $200,000 price tag. The tension: Coursera's certificates still lack the signaling power of a full degree, which limits the unbundling's reach in industries where credentialing — not learning — is the primary job the bundle performs.
Craigslist
Unbundled classified advertising from the newspaper bundle — then got unbundled itself
Craigslist is the canonical two-stage unbundling story. First, it unbundled classified ads from newspapers — extracting the single most profitable component of the newspaper bundle and offering it for free online. This devastated newspaper economics: U.S. newspaper classified ad revenue fell from $19.6 billion in 2000 to under $5 billion by 2012. Then Craigslist itself became a bundle — jobs, apartments, dating, furniture, services, all in one undifferentiated interface. A second wave of unbundlers picked it apart: Indeed and LinkedIn for jobs, Zillow and Trulia for real estate, Tinder for dating, Airbnb for short-term rentals, Etsy for handmade goods. The lesson: every successful unbundler eventually becomes a bundle that someone else will unbundle.
Charles Schwab
Originally unbundled discount trading from full-service brokerages in the 1970s
Before Robinhood, there was Schwab. In 1975, the SEC deregulated brokerage commissions, and Charles Schwab seized the moment by unbundling discount stock trading from the full-service brokerage model. Schwab stripped out the research analysts, the personal advisors, and the mahogany offices — and offered trades at a fraction of the price. The company grew to manage over $7 trillion in client assets by 2023. But the irony is instructive: over decades, Schwab gradually rebundled — adding wealth management, banking, financial advice, and retirement services — until it looked remarkably like the full-service brokerages it had originally disrupted. Then Robinhood unbundled Schwab. The cycle repeats.
Section 7
Adjacent Frameworks
Unbundling rarely operates in isolation. It connects to — and sometimes conflicts with — several adjacent strategic lenses:
Pairs well with
Niche down
Unbundling identifies the slice; Niche Down tells you which customer segment to serve first. The combination is powerful: unbundle the component, then focus obsessively on the narrowest viable audience that values it most.
Pairs well with
Clayton Christenson model of disruptive innovation
Christensen's disruption theory explains why incumbents can't respond to unbundlers. The bundle cross-subsidizes components, making it structurally impossible for the incumbent to compete on the unbundled slice without destroying its own economics. Unbundling is often disruption's delivery mechanism.
In tension with
Build feature requests on top of existing platforms
Building features on top of platforms assumes the platform (bundle) will persist and you'll add value to it. Unbundling assumes the platform is ripe for decomposition. These are opposing bets on the bundle's durability.
In tension with
Category creation
Category Creation builds something entirely new. Unbundling extracts something that already exists. The tension is real: unbundlers inherit proven demand but limited differentiation, while category creators face demand risk but own the narrative.
Apply next
Sell an Identity
Once you've unbundled a component and built a standalone product, the next challenge is differentiation against other unbundlers targeting the same slice. Selling an identity — making the product a statement about who the customer is — creates emotional switching costs that pure functionality cannot.
Apply next
Find processes for people and companies with a lot of steps and pain (friction) in going through and make fast and simple
After unbundling, look at the friction within your own slice. The standalone product should not just replicate the bundle's version — it should radically simplify the experience. Friction reduction is how you justify the switch from bundle to standalone.
Section 8
Analyst's Take
Faster Than Normal — Editorial View
Unbundling is the most reliably generative framework in the startup playbook — and also the most frequently misapplied. The reason it keeps producing billion-dollar companies is structural: bundles are inherently unstable equilibria. They exist because of distribution economics, and distribution economics keep changing. Every major technology shift — the internet, mobile, cloud, and now AI — triggers a new wave of unbundling opportunities. The supply of bundles to decompose is essentially infinite.
But here's what most people get wrong: they treat unbundling as a product strategy when it's actually a market entry strategy with an expiration date. Unbundling gets you in the door. It does not keep you in the building. The standalone product that wins by being better at one job will eventually face pressure from two directions: the incumbent improving its version of that job, and a new bundler aggregating your product with adjacent services. If you don't have a plan for what happens after the initial unbundling, you're building a transitional business.
The founders I see succeed with this framework share a specific trait: they understand the bundle's internal economics better than the bundle's own operators do. They know which components are profit centers, which are loss leaders, and which exist purely for strategic lock-in. This knowledge tells them not just what to unbundle, but what the incumbent's response will be — and how long they have before that response arrives. The best unbundlers give themselves 18–24 months of clear air before the incumbent can react.
My honest assessment: unbundling is the right framework for about 40% of startup opportunities, but founders apply it to about 80%. The result is a graveyard of "unbundled" products that were really just features — things that worked inside the bundle but couldn't sustain an independent business. The litmus test is simple: can you name three distinct monetization paths for the standalone product that don't require you to eventually rebundle? If you can't, you're not unbundling — you're building a feature that the incumbent will copy or a VC-subsidized product that dies when the funding stops.
The most exciting unbundling opportunities right now are in professional services — law, accounting, consulting, healthcare — where AI is collapsing the cost of delivering expertise that was previously bundled inside expensive human-staffed firms. A Big Four accounting firm bundles audit, tax, advisory, and consulting. Each of those is a bundle itself. AI-powered tools are beginning to unbundle specific tasks within each — and the firms are structurally unable to respond because their partnership economics depend on the bundle staying intact.
Section 9
Opportunity Checklist
Use this scorecard to evaluate whether a specific unbundling opportunity has the structural conditions to succeed. Score each item as yes (1 point) or no (0 points).
Unbundling Opportunity Scorecard
The incumbent product is clearly a bundle of 5+ distinct jobs, not a tightly integrated single-purpose tool.
A specific technology or distribution shift has recently reduced the cost of delivering the target component as a standalone product.
Customer research confirms that a meaningful segment uses only 1–2 components of the bundle but pays for the whole thing.
The target component can sustain an independent business model without cross-subsidies from other components.
The incumbent cannot easily unbundle the component itself without cannibalizing its core revenue or business model.
The standalone product delivers a 10x better experience for the target job compared to how the bundle delivers it today.
The TAM of the standalone slice is large enough to support a venture-scale business (or a profitable bootstrapped one).
Fewer than 3 well-funded competitors are currently targeting the same slice of the same bundle.
There is a clear path to building switching costs, network effects, or data advantages within 18 months of launch.
You have a thesis on how the product evolves beyond the initial unbundled slice — either through controlled rebundling or deepening the core.
The incumbent's likely response timeline gives you at least 18–24 months of clear air to build and scale.
Section 10
Top Resources
01
Book
The foundational text for understanding why incumbents can't respond to unbundlers. Christensen's core insight — that serving your best customers makes you vulnerable to attackers who serve overlooked segments — explains the structural logic behind every successful unbundling. Essential reading before you target any bundle.
02
Essay
Thompson's framework explains the internet-era mechanism that enables unbundling: when distribution costs go to zero, value shifts from controlling supply to controlling demand. This essay is the single best explanation of why bundles that depended on distribution scarcity are systematically vulnerable. Read this alongside the Christensen to understand both the why and the how.
03
Book
Written by two Berkeley economists (Varian later became Google's chief economist), this book lays out the economics of bundling and unbundling with analytical rigor. The chapter on bundling strategy explains when bundles are economically efficient and when they're vulnerable — giving you the quantitative framework to evaluate unbundling opportunities.
04
Book
Slywotzky's concept of value migration — the flow of economic value from outdated business designs to new ones — is the macro-level pattern that unbundling exploits. The book maps how value moves between companies and industries, giving you a lens for identifying which bundles are in value outflow and which slices are in value inflow.
05
Book
Anderson's thesis that the internet enables profitable serving of niche demand is the demand-side complement to unbundling. Bundles serve the average customer; unbundled products serve specific segments in the long tail. This book explains why the addressable market for a well-targeted unbundled product is often larger than incumbents assume.
Why this matters next
mental modelsNetwork Effects
Substack applied the Network Effects mental model
mental modelsNarrative
Substack applied the Narrative mental model
mental modelsPerceived Value
Substack applied the Perceived Value mental model
mental modelsScale
Substack applied the Scale mental model
mental modelsEnvironment
Substack applied the Environment mental model
mental modelsValue Chain Analysis
Substack applied the Value Chain Analysis mental model
Continue exploring
RRIAACTRNDCI
Framework
Recent funding rounds
Analyze companies that have recently secured significant investment, identifying
Framework
Industry timing arbitrage
Apply newly developed technology from one industry to another that hasn't yet ad
Framework
Acqui-Deaths
Identify opportunities created when large companies acquire startups, potentiall
Framework
Three-Star reviews
Find business opportunities by analyzing moderately satisfied customers' feedbac
Framework
Niche down
Focus on a highly specific market segment or customer base, becoming a specialis
Framework
Clayton Christenson model of disruptive innovation
Start with a big market that's ripe for disruption. Look for industries where cu
More like this, in your inbox
I send a newsletter every week — free, no spam, unsubscribe anytime.