The Ansoff Matrix in 2026 — Four Growth Strategies, With Real Examples

Table of contents

The four quadrants

The matrix’s whole insight: not all “growth” is equal. The strategies in the bottom-right (Diversification) require fundamentally different capabilities than the ones in the top-left (Market Penetration).

1 · Market Penetration — sell more to who you already serve

Risk: lowest. Strategy: maximize wallet share within your existing customer base. No new product. No new market. Just better extraction of the customers you’ve already won.

How to execute:

• Price optimization, bundling, upgrade tiers
• Loyalty and retention programs
• Increase usage frequency
• Reduce churn

2026 example: Costco’s price/membership levers. Each membership-fee increase (Executive Gold to $130/yr in 2024 was the most recent) is pure market penetration — same product (the warehouse) to the same market (US shoppers), just a higher revenue per customer.

2026 example: Netflix’s paid-sharing crackdown. Same product (streaming subs) to the same market (households already using Netflix), just now monetizing the previously-free extra-member relationships. Added ~22M net subs above prior trend.

2 · Market Development — same product, new market

Risk: medium. Strategy: take a working product into adjacent segments or geographies. You know the product works. You’re now testing whether it works for new buyers.

How to execute:

• Geographic expansion (new countries, regions)
• Adjacent customer segments (B2C → SMB, SMB → enterprise)
• New use cases for an existing product
• New distribution channels

2026 example: Notion entering schools and universities. Notion’s product is essentially unchanged from its mid-2010s release, but they spent 2023–2025 building education-specific go-to-market — free for students, dedicated education plans, integration with Google Classroom. New market for the same product.

2026 example: Stripe pushing into India and Latin America. Same payments infrastructure, materially new local-payment-method integrations, but architecturally the same Stripe.

3 · Product Development — new product, existing market

Risk: medium. Strategy: build new products for buyers who already trust you. You know the market and have distribution. You’re testing whether you can build something they’ll buy.

How to execute:

• New product lines for existing customer base
• Subscription tier additions
• Adjacent product categories
• Vertical integration into your supply chain

2026 example: Apple Vision Pro launching to existing Apple buyers. Same buyers (Apple customers with $3,500 to spend on a new toy), genuinely new product (mixed-reality headset). Vision Pro materially undersold projections — a useful reminder that “same market” doesn’t guarantee adoption of a new product.

2026 example: Cloudflare adding Workers AI to existing Cloudflare customers. The customer (developer/CTO using Cloudflare for CDN + Workers) is already there. The new product (LLM inference on edge) is novel but maps to existing infrastructure relationships.

4 · Diversification — new product, new market

Risk: highest. Strategy: build something new for buyers you don’t know. You’re betting on both market entry and product success simultaneously — independent probabilities that compound.

How to execute:

• M&A into unrelated categories
• Greenfield product builds into new verticals
• Platform plays that change what the company is

2026 example: Amazon’s Project Kuiper (satellite broadband). New product (satellites + ground stations + consumer modems), new market (rural broadband customers). Different from AWS, different from retail, different from Prime Video.

2026 example: Netflix entering games. New product (interactive games), new market (mobile gaming users who weren’t necessarily streaming subscribers). Still in the “engagement metric” phase rather than the “real revenue line” phase — the prototypical diversification trajectory.

Penetration vs Development vs Diversification — risk vs reward

Directional. The actual risk of each quadrant varies by company and execution; the rank-ordering rarely does.

How to actually use the matrix in 2026

The Ansoff Matrix isn’t a recommendation; it’s a clarifier. Use it like this:

1. 01
   Map every growth option you're considering into a quadrant. "Launch in Canada" is Market Development. "Sell to a different ICP" is Market Development. "Build a paid analytics product" is Product Development. "Acquire a logistics company" is Diversification.
2. 02
   Stack-rank them by your honest probability of success. Most teams discover their highest-confidence options are clustered in Penetration + the closer quadrants.
3. 03
   Pick the lowest-risk option that meaningfully moves the number. If Market Penetration moves you 20% but Diversification has a 10% chance of moving you 200%, the EV calculation matters — and so does your appetite for variance.
4. 04
   Pre-commit to what kills each option. Specifically: at what point do you cut a Diversification bet? Most companies don't decide this upfront and end up over-committing.

The 2026 update — does it still work?

The Ansoff Matrix predates the internet, social media, the cloud, and AI. It still works in 2026 because the underlying insight (risk increases as you change either axis) is platform-independent.

What’s changed:

• Distribution costs collapsed, making Market Development (new geos, new segments) cheaper to test than it was in 1957
• Product development costs partially collapsed (AI-assisted dev) but expectations got higher (users expect polished products from day one)
• Diversification risk got worse for capital-constrained startups; better for capital-rich incumbents (because failed bets are cheaper to write off)

Bottom line

The Ansoff Matrix is a 67-year-old framework that still works because the core insight — risk increases when you change either axis — is fundamental. The 2026 update is mostly about what specific moves cost (distribution is cheaper, expectations are higher) rather than what the moves are.

For operators: use it to clarify which growth options you’re considering and what risk profile you’re picking. Don’t use it to pick the option — that’s your job.

Related: BCG Matrix examples · Land and expand strategy · Top of mind awareness

This guide is part of alejandrorioja.com — written by Alejandro Rioja, who builds AI agent systems for founders. Including the agent that keeps this site current. How it works →