Ansoff Matrix Examples: 4 Growth Strategies (2026)

The Ansoff Matrix sorts every growth idea into one of four strategies by asking two questions — is the product new, and is the market new — and then ranks those four boxes by rising risk. It's the fastest way to see whether a company's growth bets are concentrated in safe, familiar territory or stacked in the corner where most expensive failures live. This guide works the matrix on real, current companies — Nvidia's June 2026 Computex portfolio, Netflix, Apple, Amazon — rather than abstractions, because the framework only earns its keep when you can see a named company's actual choice inside it.

Economist Igor Ansoff introduced the model in a 1957 Harvard Business Review article, "Strategies for Diversification." Nearly 70 years later it survives because the two-by-two is honest about one thing every growth deck tries to hide: risk rises as you move away from what you already do.

Want to run the matrix on your own company? Framework for iPhone & iPad fills in each quadrant with AI-assisted prompts. Free to start.

Prefer to watch or listen?

The four growth strategies

The two axes are product (existing vs. new) and market (existing vs. new). The four combinations are the four strategies — listed here from lowest to highest risk:

Strategy	Product	Market	What it means	Relative risk
Market Penetration	Existing	Existing	Sell more of what you have to who you already serve	Lowest
Product Development	New	Existing	Build something new for your current customers	Moderate
Market Development	Existing	New	Take your current product to new customers or geographies	Moderate
Diversification	New	New	New product and new market at once	Highest

The ordering is the whole point. Market Penetration leans on two known quantities; Diversification leans on neither. Everything Ansoff teaches flows from that single gradient.

How to actually use it

1. Define the unit of analysis. One product line, one business unit, or the whole company — pick the altitude before you start, or moves will land in the wrong box.
2. List the growth initiatives in play. Real ones with budget behind them, not aspirations. Aim for 5–12.
3. Place each one with the two-question test. Is the customer new to us? Is the product new to us? The answers assign the quadrant deterministically (see the Placement Test below).
4. Apply the adjacency discount. A "new" product that reuses an existing moat — a brand, a distribution channel, a software platform — is less risky than a cold-start move into the same quadrant. Note it; don't ignore it.
5. Map the portfolio, not the project. The insight is rarely about one bet. It's about the distribution — how much of your capital and management attention sits in each quadrant.
6. Read the imbalance. If most of the spend is in Diversification while the core business is already under pressure, the matrix has done its job: it made the risk concentration visible.

Worked example: Nvidia at Computex 2026

The clearest current example is Nvidia's June 1, 2026 Computex keynote, where Jensen Huang announced three things the press treated as separate product stories. Ansoff treats them as one portfolio decision:

Initiative	Product new?	Market new?	Quadrant	Note
More Vera Rubin GPUs to hyperscalers	No	No	Market Penetration	CUDA lock-in; the cash engine
Vera Rubin platform, Spectrum-X to existing AI-cluster buyers	Yes	No	Product Development	Same customers, next-gen silicon
Sovereign-AI and enterprise build-outs	No	Yes	Market Development	Same accelerators, new buyers
RTX Spark — a superchip for mainstream Windows PCs	Yes	Yes	Diversification	New product and new buyer
Isaac GR00T humanoid-robot reference	Yes	Yes	Diversification	Physical-AI: a wholly new market

Read this way, the announcement is a statement about where Nvidia is putting its riskiest chips. The penetration and product-development quadrants do the cash-generating work; RTX Spark and GR00T sit in Diversification — the highest-risk box — even as the core business is already concentrated in AI compute. RTX Spark looks like Product Development because Nvidia already sells GPUs to PC makers, but the addressable buyer (the general Windows-PC owner) and the product category (an Arm+Blackwell PC SoC) are both new enough to land in Diversification — with an adjacency discount from CUDA and existing OEM relationships.

For the full quadrant-by-quadrant breakdown, see Nvidia Ansoff Matrix 2026: the RTX Spark PC bet. For a different company with the same center-of-gravity problem — its valuation thesis sitting in the riskiest quadrant — see Tesla Ansoff Matrix Analysis 2026.

Ansoff Matrix examples by quadrant (real companies)

The four-quadrant set most readers are searching for — one well-known company per strategy:

Strategy	Company	The move
Market Penetration	Coca-Cola	More volume to existing buyers via promotion, pricing, and distribution reach — no new product, no new market
Product Development	Apple	New devices (Apple Watch, Vision Pro) sold into the existing iPhone customer base
Market Development	Netflix	The same streaming service taken country by country into new geographies
Diversification	Amazon	AWS — a new product (cloud compute) for a new market (enterprise IT), distinct from its retail origin

Notice that the three "famous" cases (Coca-Cola, Apple, Netflix) lean on something the company already owns — a brand, a customer base, a service. Only Amazon's AWS was a genuine both-axes-new bet, and even then it reused Amazon's own data-center muscle. That reuse is the adjacency discount in action, and it's the part most Ansoff write-ups skip.

The Ansoff Risk Ladder and Placement Test

Most explainers stop at "diversification is risky." That's directionally true and operationally useless. The Ansoff Risk Ladder turns the gradient into a usable diagnostic by pairing each quadrant with its specific failure mode and the adjacency discount that can move a bet down a rung:

Rung	Quadrant	Why the risk	Dominant failure mode	What earns an adjacency discount
1 (safest)	Market Penetration	Both known	Saturation — you run out of room	A still-underpenetrated segment
2	Product Development	Customer known	Build the wrong thing for a known customer	Strong existing customer-research loop
3	Market Development	Product known	Misread the new market's needs	A transferable distribution channel
4 (riskiest)	Diversification	Neither known	Spread thin across two unknowns at once	A moat that travels — brand, platform, or channel

The Placement Test assigns any initiative to a rung with two questions:

1. Is the customer new to us? (yes / no)
2. Is the product new to us? (yes / no)

No + No → Penetration. No + Yes → Product Development. Yes + No → Market Development. Yes + Yes → Diversification. Then apply the discount: if the "new" axis reuses an existing moat, the bet behaves like the rung below it. Nvidia's RTX Spark scores rung 4 on the test but earns a partial discount toward rung 2 because CUDA and the OEM relationships travel with it — which is exactly why it's a defensible diversification rather than a reckless one.

This is the lens to carry into a deeper portfolio tool. Where Ansoff tells you which growth bets you're making, the BCG Matrix tells you which existing units should fund them — the two pair naturally, with Ansoff feeding the growth thesis and BCG allocating the cash.

When NOT to use the Ansoff Matrix

• Pure operational or pricing decisions. Ansoff is about direction of growth, not execution. Use RICE to rank features within a chosen direction.
• Industry-structure questions. "Is this market even attractive to grow into?" is a Porter's Five Forces question; Ansoff assumes you've already decided the market is worth entering.
• Macro-environment scans. Before placing a Market Development bet abroad, run a PESTEL analysis on the target geography — Ansoff won't surface the political or regulatory landmines.

Common mistakes

Mistake	Why it happens	Fix
Calling everything "Product Development"	New products feel safer than they are when sold to existing customers	Run the Placement Test on the market, not the org chart — a new buyer segment is Market Development even inside the same company
Ignoring the adjacency discount	Quadrants are treated as binary, so every diversification looks equally reckless	Ask what moat travels with the bet; a brand or platform that transfers lowers the real risk
Mapping projects, not the portfolio	Teams analyze one launch in isolation	Place all funded initiatives at once — the imbalance only shows in aggregate
Treating the matrix as a recommendation	The grid names risk; it doesn't tell you to avoid it	High-risk quadrants can be correct — the point is to fund them knowingly, with the core business healthy enough to absorb a miss

Related frameworks

• Nvidia Ansoff Matrix 2026 — the full worked deep-dive behind the example above
• Tesla Ansoff Matrix Analysis 2026 — a second company with its valuation thesis in the riskiest quadrant
• Ansoff Matrix (catalog entry) — the template to fill in for your own company
• BCG Matrix worked on Nvidia — the portfolio-funding companion to Ansoff's growth lens
• Porter's Five Forces: a beginner's guide — for the "is this market attractive" question Ansoff skips
• PESTEL — the macro scan to run before any Market Development bet abroad

Want to run this on your phone? Framework for iPhone & iPad — fill in any framework with AI assistance.

Sources

• Igor Ansoff — "Strategies for Diversification," Harvard Business Review (1957) — the original framework
• NVIDIA Newsroom — "Rubin platform in full production" (Computex, June 2026)
• Tom's Hardware — "Nvidia unveils RTX Spark superchip at Computex 2026" (June 1, 2026)
• Netflix — Investor Relations, quarterly earnings — international market-development history
• Amazon Web Services — the diversification reference case