The Four Entrepreneurial Strategies

Foundations of Business Development - NIT Northern Institute of Technology / TUHH, Hamburg · part of my Technology Management MBA · study notes for revision.

---

The last chapter left us with four strategic choices - customers, technology, competition, identity. This chapter turns those choices into a map. Once you have an innovation worth commercialising, how do you actually take it to market? The framework here - the Entrepreneurial Strategy Compass from Gans, Scott and Stern - says there is no single right answer, but there are only four fundamentally different routes. Pick the one that fits your innovation, and you have a coherent strategy. Mix them up, and you fall between stools.

Tip

The whole chapter in one line: every start-up commercialising an innovation is really answering two questions - do I collaborate with the big incumbents or compete against them? and do I win by controlling a key asset or by out-executing everyone? Cross those two questions and you get four strategies. The compass is a thinking tool, not a formula.

1 · The two questions behind every go-to-market

An incumbent is an established firm that already owns the market - the customers, the distribution, the brand, the factories. Economists call those things complementary assets: the resources you need alongside a good invention to actually make money from it. A brilliant coating is worthless until it reaches a shelf; a clever algorithm needs users. The central tension of entrepreneurial strategy is that incumbents usually own the complementary assets, and you usually own the fresh idea.

Out of that tension come two big decisions.

Question 1 - Compete or Collaborate?

• Collaborate: work with incumbents - license your idea to them, partner, supply them, or get acquired. You borrow their complementary assets.
• Compete: go up against them - build your own value chain, win your own customers, and try to take share.
• The deep trade-off: collaboration is lower-risk but you share the value; competition keeps the upside but you must build everything yourself.

Question 2 - Control or Execution?

• Control: win by owning a defensible asset - a patent, a trade secret, a platform standard - so nobody can copy you.
• Execution: win by doing it better and faster - superior operations, speed, and hustle in the open market.
• The deep trade-off: control gives durable protection but is slow and legalistic; execution is fast but the moat is only as good as your last quarter.

Two axes, two trade-offs. Almost every strategic argument a founding team has is secretly an argument about one of these two questions.

2 · The Compass: crossing the two axes

Put “Collaborate versus Compete” along the top and “Control versus Execution” down the side, and the four quadrants each name a distinct strategy. This is the Entrepreneurial Strategy Compass.

← COLLABORATE (work with incumbents)　|　COMPETE (go against them) →

Intellectual Property collaborate + control

• Keep the invention, protect it, and license or partner it out.
• Win because nobody else can legally use your idea.

Architectural compete + control

• Control a whole new platform or standard, own the customer.
• Win by reshaping the value chain around yourself.

Value Chain collaborate + execution

• Slot into an existing chain as the best supplier of one step.
• Win because you execute your layer better than anyone.

Disruption compete + execution

• Build your own value chain from an overlooked foothold.
• Win by out-executing slow, complacent incumbents.

Top row = win on CONTROL　·　Bottom row = win on EXECUTION

The compass. Left column collaborates with incumbents; right column competes. Top row wins through control of an asset; bottom row wins through execution. Four quadrants, four strategies.

Notice how the compass connects to the four choices from the previous chapter. Your quadrant is not a fifth, separate decision bolted on top - it is what emerges when your choices about customers, technology, competition and identity all line up. Choosing under-served customers, a nascent technology and a “hustler” identity pushes you toward the Disruption corner; choosing to control a novel invention and a collaborative posture pushes you toward Intellectual Property. The compass is really a way of reading back the strategy your four choices imply - and of checking they agree with each other.

The rest of the chapter walks each quadrant in turn - its logic, what it needs, its risks, and a real example.

3 · Intellectual Property strategy - collaborate + control

Here you treat your invention as the crown jewel. You protect it with formal intellectual property - patents, copyrights, trademarks, trade secrets - and then you don’t try to build a whole company around it. Instead you license the protected idea to incumbents, or partner with them, letting them supply the complementary assets (factories, sales forces, customers) while you collect royalties and stay in the lab. You compete in the “market for ideas” rather than the market for products.

Note

Why patents make this work. A patent is a quid pro quo: the state grants you the right to exclude others for around 20 years, and in return you publicly disclose how the invention works. To be patentable an idea must be novel (genuinely new), non-obvious (not trivial to someone skilled in the field) and useful. That legal right to exclude is what lets you sit across the table from a giant and say “you cannot make this without me.”

What it requires. Genuinely strong, defensible IP on the technology frontier, and a team that is as good at managing and licensing patents as at inventing them - an “ideas factory”. The biotech firm Millennium Pharmaceuticals is the classic template: a superstar advisory board, tight coupling between IP and research, and a disciplined string of licensing deals (Roche, Eli Lilly, Bayer, Monsanto) where each deal built on the last.

Patents are only the headline tool. In practice the control toolkit has three layers, and most IP strategies stack several of them:

Formal IPpatents, copyrights, trademarks

+

Design & secrecyproduct design, trade secrets

+

Contractslicences, non-competes, NDAs

The three layers of control. A patent excludes copiers; a trademark stops customer confusion; a trade secret (Coca-Cola) protects what you never disclose; contracts pin down how a partner may use what you share.

One subtlety worth remembering: patents are probabilistic. An IP strategy is partly a bet on how strong your rights turn out to be when tested, because examiners have wide discretion and courts can narrow or overturn a grant. There is even a timing signal here - the pace of licensing deals jumps sharply after a patent is actually granted, because the grant resolves uncertainty over scope and hands you a stronger bargaining position.

The core risk - Arrow’s disclosure problem. Kenneth Arrow spotted the trap: to sell an idea you have to describe it, but once a buyer understands it, why pay? The very act of disclosure can destroy the idea’s value, and a big partner may simply “hold up” the small inventor. Founders resolve this with secrecy (Coca-Cola’s formula), by creating competition among several buyers so exclusivity is at stake, or by dealing only with buyers whose reputation makes cheating too costly (Cisco is famous for treating start-ups fairly, so entrepreneurs trust it).

Invent & patentnovel, non-obvious, useful - protect it

↓

License / partnersell rights in the “market for ideas”

↓

Incumbent goes to marketthey supply the complementary assets

The IP play: you keep and protect the invention; the incumbent does the manufacturing and selling; royalties flow back to you.

Example - LiquiGlide. Spun out of MIT in 2012, LiquiGlide’s patented “liquid-impregnated surface” makes viscous liquids slide out of any container. Rather than build ketchup bottles itself, it protected the technology tightly (issued patents plus a dozen pending) and signed exclusive licensing partnerships - for example with glue-maker Elmer’s - to reach markets it could never have served alone. Even J.K. Rowling fits the pattern: she keeps strict control of the Harry Potter rights but collaborates with publishers and studios who own the distribution.

Caution

When IP strategy fails - Bionym / HeartID. Two PhDs built ECG (electrocardiogram) authentication - identifying you by your heartbeat - and tried to license it. It flopped as an IP play: the tech was not plug-and-play, so partners faced huge integration costs; the founders had no track record, so cautious buyers wanted a proven firm; and they never landed a first customer to prove it worked. Lesson: an IP strategy needs IP that is genuinely easy to transfer and protection strong enough to bargain with. Bionym eventually gave up, rebranded as Nymi, and built its own product instead - a switch to a different quadrant entirely.

4 · Value Chain strategy - collaborate + execution

Same collaborative half of the compass - you work with incumbents rather than fight them - but now you win on execution, not legal control. You slot into an existing value chain as a component or supplier, do one step brilliantly, and let the incumbent keep owning the end customer. You are not trying to remake the industry; you are trying to become the irreplaceable link in one part of it.

What it requires. A vital, hard-to-replicate capability at your layer, and enough bargaining power that the incumbent cannot simply squeeze you or swap you out. There are three practical ways to manufacture that power:

• An interdependent team where you can ration access to key people and no single member is enough on their own - the capability lives in the combination. Peter Thiel’s line about the early PayPal team (“far more incredible than we realised”) is exactly this.
• Unique competencies that simultaneously make the chain more valuable and cannot be replicated by a rival supplier or copied by the partner itself.
• Deep embedding - wiring yourself into the chain so completely that you become the “irreplaceable” supplier for your layer.

Choosing partners with a reputation for treating start-ups fairly matters enormously, because you are betting your company on their goodwill.

Your layerexecute one step better than anyone

→

Incumbent’s chainthey own brand & distribution

→

End customerserved through the partner

The value-chain play: become the indispensable supplier of one link, and let the incumbent’s chain carry your work to market.

Example - Sun Microsystems and Computervision. Early Sun was a 40-person firm with an incomplete workstation, and Computervision had already picked a rival (Apollo) as its standard. Co-founder Vinod Khosla made an offer they could not refuse: let Sun serve as the OEM (original equipment manufacturer - the firm that actually builds a product another company sells under its own name), sending half of Sun’s people over to put it into production, for almost nothing in return. Computervision took the deal to save margin - but could never match Sun’s day-to-day technical expertise, and ended up ceding control of design and even manufacturing to Sun. That one contract made Sun a “legitimate” company, and over time it inched closer and closer to the final customer. Infosys tells the same story at national scale: Narayana Murthy perfected the Indian IT-outsourcing model as a supplier layer for global firms, then gradually climbed to cover more of the value chain.

The main risk - capture. Your buyer is powerful. If you are replaceable, they will drive your margins to the bone, and you live and die by their goodwill. The whole strategy hinges on being the linchpin that cannot be swapped out - the moment you can be, the value flows to them, not you.

5 · Disruption strategy - compete + execution

Now cross to the competing half of the compass. In a disruption strategy you build your own value chain and go head-to-head with incumbents - but you rarely charge the front door. You start from a foothold they ignore: a low-end, niche or “unfashionable” segment they are happy to cede because their core customers do not want it. From that lonely position you improve, gather capabilities, and climb toward the mainstream. You win on execution and speed - being fast and cheap enough that by the time incumbents notice, you are already ahead.

Note

Why incumbents freeze - the Innovator’s Dilemma. Clayton Christensen’s insight: good managers ignore disruptors rationally. Market inertia - the newcomer first serves a segment the incumbent’s best customers don’t care about, so responding looks like a distraction (and cannibalising your own profitable product feels like self-harm - Arrow’s “replacement effect”). Organisational inertia - big firms are built around their existing architecture and struggle to reconfigure. The disruptor’s advantage is precisely this predictable slowness.

What it requires. A credible, well-grounded hypothesis for why incumbents will be slow to react - plus a lean, hustling team that can experiment its way to product-market fit cheaply. Paul Graham’s dictum applies: being cheap is almost the same as iterating fast, because most start-ups die by running out of money before they build something people want.

Underneath, a disruption strategy rests on two hypotheses you must both believe - and this pairing comes straight back in the experimentation chapter:

Value-creation hypothesis

• Can you create real value for under-served customers the incumbents neglect?
• If no one actually wants the foothold product, there is nothing to build on.

Value-capture hypothesis

• Will incumbent firms react slowly enough for you to keep the value you create?
• If a giant can copy you overnight, you created value but capture none of it.

Disruption needs both to be true: you must create value for a neglected segment AND have a real reason the incumbents stay asleep long enough for you to bank it.

Foothold the incumbent ignoreslow-end or niche segment, avoid detection

↓

Out-execute & iteratelean experimentation, hustle, keep costs low

↓

Climb toward the mainstreambuild capabilities before incumbents wake up

The disruption play: enter where nobody is watching, out-execute, and move upmarket before the giants respond.

Example - Netflix versus Blockbuster. Reed Hastings and Marc Randolph attacked Blockbuster’s late-fee-laden empire with DVD-by-mail, deliberately courting cinephiles and less-popular (cheaper) films, plus a recommendation engine and a USPS partnership. Having no stores made Netflix nearly invisible to Blockbuster, whose franchisees resisted moving away from bricks-and-mortar. Crucially, Netflix later disrupted itself by pushing into streaming. Other exemplars from the deck: Warby Parker attacking Luxottica’s 80%-share eyewear grip with 100-dollar online glasses; Vanguard, where John Bogle’s “average returns, low fees” pitch was mocked as “Bogle’s Folly” and is now the world’s second-largest asset manager; and Salesforce, where Marc Benioff’s Software-as-a-Service bet declared “the end of software.”

Tip

The disruptor’s do’s and don’ts. Do avoid detection, occupy a “lonely place on the strategic frontier”, obsess over the new customer, experiment in real time, and hustle. Don’t “moon the giant” with loud public comparisons that wake it up, don’t try to rush the market’s adoption curve, and never underestimate the cost of building a value chain - it is many times more expensive than developing the idea. And note: a disruption strategy doesn’t even need an existing firm to disrupt (Ministry of Supply invented a whole new menswear category “between the lines” of Brooks Brothers and Nike).

The main risk. Incumbents can wake up - and building an entire value chain yourself is capital-intensive and slow. Your whole thesis rests on the assumption that the giants stay asleep long enough; if that assumption is wrong, a well-funded incumbent can crush you.

6 · Architectural strategy - compete + control

The boldest quadrant: you compete and win through control - but the thing you control is not a single patent, it is an entire new architecture. You do not slot into a value chain or attack one; you build a new one and sit at its centre, usually as a platform that brings multiple parties together and intermediates between them. You own the customer relationship end-to-end and set the standard everyone else must plug into.

Note

Product versus platform. A product business sells one thing down a linear chain - Apple selling you an MP3 player. A platform business matches two or more sides and owns the chokepoint between them - Apple’s iTunes and App Store connecting users and content creators, taking a cut, and controlling the rules. The magic ingredient is network effects: each new user makes the platform more valuable to everyone else, so demand feeds demand. Examples: eBay, Airbnb, PayPal, the PlayStation, App stores.

What it requires - three moves. Gans, Scott and Stern describe building a platform in three steps:

1 · Identify valuewhere do network effects live?

Find the sources of platform valuedirect network effects, multi-sided matching, market design

2 · Core the platformbuild the centre you keep

Create the core & keep controldecide the sides, the functionality, the rules, who pays

3 · Tip the platformreach critical mass

Win the standard warincreasing returns and lock-in decide the winner

Coring, then tipping. “Tipping” means getting enough adopters that the market locks in around your architecture - and the winner is decided by the dynamics of adoption, not always by who has the best underlying product.

Coring is designing the centre of the platform: who comes on board, what the core functionality is, what rules govern transactions, and who pays how much. A neat pricing rule falls out of it - subsidise the side that drives adoption, and charge the side that follows. Tipping is the standard war: once a critical mass adopts, increasing returns and lock-in make every future customer pick you, even if a rival is intrinsically better. The QWERTY keyboard won exactly this way - not because it was best, but because adoption tipped its way.

Note

When you are designing a market, not just a product. The deepest architectural plays are acts of market design. Economists Roth and others argue a marketplace works only when it has thickness (enough participants ready to transact), lack of congestion (people can weigh enough options to settle on good matches), and safety (transacting on the platform is safe and simple, so nobody games it or defects). And some exchanges hit repugnance - social norms that forbid using prices at all (kidneys, for example). An architect is not only building software; they are writing the rules of a small economy.

A practical warning the deck stresses: you almost never architect the whole value chain at once. You stage it - establish one side first with a compelling value proposition (Facebook starting at Harvard, then the Ivy League, then every US campus), then fold in the other side. Basing your strategy on the value that would exist if everyone adopted at once is how platforms fail.

Example - Care.com. Sheila Lirio Marcelo built a two-sided marketplace matching families with caregivers. Every architectural question was a design choice: which sides to onboard, whether to require background checks, whether to allow negative ratings, and - critically - a freemium model on both sides to reduce search costs and prime the pump. That is coring in action. Deeper in history, the medieval Champagne Fairs were the same idea: entrepreneurs who created value by architecting an entirely new marketplace that fused a physical venue with the rules of exchange.

The main risk - highest of all four. Platforms demand large upfront sunk investments under huge uncertainty, and tipping requires you to subsidise switching, accept low margins, and even cede profit to other players to get them to invest. Worse, the strategic sting in the tail: even a successful platform architect may be limited in how much value they can ultimately capture. Highest reward, highest risk - and it lives or dies on scale and timing.

7 · The comparisons the cases keep forcing

Two pairs of quadrants sit right next to each other and are the ones teams most often confuse. Getting the distinction crisp is the fastest way to check you have actually chosen.

Value Chain vs Disruption - both are execution plays, so the difference is purely the first axis. In a value-chain strategy you collaborate: you hand the customer to an incumbent and win by being the best link in their chain. In a disruption strategy you compete: you keep the customer and build your own chain. Sun-through-Computervision versus Netflix-against-Blockbuster - same “win by executing”, opposite answer on whether you plug into the giant or go around it.

Architectural vs Disruption - both compete, so here the difference is the second axis. A disruptor wins on raw execution against slow incumbents; an architect wins by controlling a new standard and locking customers in. A disruptor could, in principle, be out-executed by a faster rival; an architect who tips the market is protected by network effects and lock-in. Roughly: disruption is “out-hustle them”, architecture is “own the rails everyone rides on”.

Note

Where do Tesla, Uber and Amazon sit? The cases deliberately leave these open, because real firms drift and blend. Amazon disrupted bookselling on execution, then became an architect with a marketplace and AWS. Uber looks disruptive but is really architectural - a two-sided platform for riders and drivers. The point of the exercise is not to file each firm in one box, but to see that the dominant logic of value capture tells you which quadrant is really running the show.

8 · The four strategies side by side

Strategy	Relationship to incumbents	Source of advantage	What it requires	Main risk	Example
Intellectual Property collaborate + control	Collaborate - license or partner	Legal control of a protected invention	Strong, transferable IP; deal-making skill	Disclosure / hold-up; weak or copyable IP	LiquiGlide, Millennium Pharma
Value Chain collaborate + execution	Collaborate - supply the chain	Execution of one irreplaceable link	Unique capability; bargaining power as linchpin	Capture by a powerful buyer; thin margins	Sun Microsystems, Infosys
Disruption compete + execution	Compete - build your own chain	Execution, speed and hustle	A reason incumbents stay slow; lean, cheap team	Incumbents wake up; capital-intensive	Netflix, Warby Parker, Vanguard
Architectural compete + control	Compete - build a new chain	Control of a platform / standard	Scale, timing, network effects, deep pockets	Huge sunk cost; may struggle to capture value	Care.com, App Store, eBay

The four routes at a glance. Read the badges as compass coordinates: collaborate/compete on one axis, control/execution on the other.

9 · There is no single best strategy

The temptation with any 2×2 is to ask “which quadrant is best?” The honest answer is none of them, always - and that is the real lesson of the compass. The same underlying invention can, and often should, be commercialised through different quadrants depending on context: recall that Bionym’s heartbeat-authentication tech failed as an IP play but the founders could have run it as a value-chain or disruption strategy instead. The technology did not change; the fit between technology, market and choices did. So the right strategy depends on three things working together:

Your innovation - how defensible, how transferable?The incumbents’ complementary assets - essential, or can you go around them?Your four choices - customers, technology, competition, identity

If your IP is strong and incumbents own assets you cannot replicate, collaborate and control - an IP strategy. If those assets are essential but your edge is executing one step, collaborate and execute - a value-chain strategy. If incumbents are slow and their assets don’t help (or actively hurt) in the new game, compete and execute - a disruption strategy. And if you can set a brand-new standard and own the customer, compete and control - an architectural strategy.

A helpful way to sanity-check a choice is to run the two axes as questions in sequence:

Do incumbents own complementary assets I truly need?yes → collaborate　·　no → compete

↓

Is my edge a defensible asset, or superior execution?asset → control　·　execution → execution

↓

The two answers together name your quadrantand, ideally, match your four choices

Two questions, one quadrant. If the answer disagrees with the identity and technology choices you already made, something in the plan is incoherent - fix it before you go to market.

Caution

Coherence beats cleverness. The danger is not picking the “wrong” quadrant - it is being incoherent, half-collaborating while half-competing, so your customers, technology, competition and identity choices pull against each other. Remember Bionym: the honest move was to abandon a failing IP strategy and re-pick a coherent one, not to muddle on. The compass earns its keep by forcing that clarity: choose a quadrant on purpose, and align everything behind it.

Revision summary

Tip

The one-minute recap

• The Entrepreneurial Strategy Compass (Gans, Scott, Stern) crosses two questions: Compete vs Collaborate with incumbents, and win on Control vs Execution. That gives four strategies.
• Intellectual Property (collaborate + control): protect the invention, license or partner it out; needs strong, transferable IP; risk is Arrow’s disclosure problem and hold-up. LiquiGlide.
• Value Chain (collaborate + execution): become the irreplaceable supplier of one link; needs bargaining power as the linchpin; risk is capture by a powerful buyer. Sun Microsystems.
• Disruption (compete + execution): build your own chain from a foothold incumbents ignore, then move up; needs a reason they stay slow (the Innovator’s Dilemma); risk is they wake up. Netflix.
• Architectural (compete + control): control a new platform/standard, own the customer; needs scale, timing and network effects (core then tip); highest risk and reward. Care.com, App Store.
• No quadrant is universally best. The right one depends on your innovation, the incumbents’ complementary assets, and your four strategic choices. Coherence is what wins.

Next: Strategic Learning & Experimentation → - testing your riskiest assumptions before you bet the company.