The Ansoff Matrix for Startups: Growth Directions

The Ansoff matrix is a 2x2 growth-planning tool, introduced by Igor Ansoff in 1957, that sorts your options by two questions: are you selling an existing or a new product, and to an existing or a new market? The four combinations range from low-risk market penetration to high-risk diversification.

Quick Answer: The Ansoff matrix maps four growth strategies by novelty. Market penetration (existing product, existing market) is safest; market development (existing product, new market) and product development (new product, existing market) each add one unknown; diversification (new product, new market) is riskiest. Most startups should win at penetration first, then expand one square at a time.

Growth is not a single decision. "Get bigger" can mean squeezing more from the customers you already have, carrying your product to a new audience, building something new for the audience you own, or leaping into a market and a product you have never touched. Those moves carry wildly different risk, and founders routinely treat them as interchangeable.

The Ansoff matrix exists to keep them separate. Introduced by Igor Ansoff in a 1957 Harvard Business Review article and later expanded in his book Corporate Strategy, it is sometimes called the product-market expansion grid. It is one of the classic startup strategy frameworks — but where SWOT or Porter's Five Forces diagnose your situation, Ansoff is specifically about the direction of your next growth bet. This guide shows how to place your options, read their risk, and sequence them.

The four Ansoff quadrants, ranked by risk

The matrix crosses two axes — product (existing or new) and market (existing or new) — producing four growth strategies whose risk rises with every step away from what you already know. Here is each quadrant, the product-and-market combination that defines it, and where it sits on the risk gradient.

Takeaway: Risk is a function of novelty. Penetration asks you to learn nothing new; market and product development each introduce one set of unknowns; diversification stacks two unknowns on top of each other, which is why Ansoff treated it as a category of its own.

Market penetration means growing without changing what or who you sell. You already have fit in a defined market, so the work is to capture more of it — win customers from rivals, raise usage and frequency, improve retention, or convert non-users. It is the cheapest growth because every unknown has already been paid for.

Market development takes a working product to a new market. Same offering, new buyers: a different industry vertical, a new geography, or a new use case for a different segment. Product risk stays low because the thing already works. The risk is that the new market may not want it, price it, or adopt it the way your first one did.

Product development builds a new product for a market you already own. You keep your hard-won customer understanding and distribution and add a new offering — a second product, a major capability, an adjacent line. Market risk is low. The risk is execution, and whether the new product earns its place with people who trust you for something else.

Diversification is a new product and a new market at once. Both axes are unfamiliar, so both sets of unknowns compound. Ansoff singled it out as the riskiest strategy precisely because you cannot lean on existing customers or an existing offering to cushion the bet. It can pay off, but it is the square a young startup should approach with the most suspicion.

How to place your growth options in the matrix

To use the matrix, list every growth idea you are weighing and drop each one into a quadrant by answering two questions honestly: is the product new to us, and is the market new to us? The classification only works if the honesty holds.

Define "existing market" by who you actually understand, not who you could imagine serving. Your existing market is the specific customer you already reach, know, and have evidence about. A new market is any buyer whose needs, channels, and willingness to pay you would have to relearn. Founders often mislabel a new segment as "existing" because it sounds adjacent, which quietly understates the risk.

Treat "new" as a spectrum, not a switch. Selling your tool to a neighboring vertical is less new than launching in another country, which is less new than inventing a product for strangers. Two ideas can share a quadrant and still differ in risk. Use the quadrant to sort, then rank within it by how far the "new" really stretches.

For each growth idea on your list, ask in order:

• Is the product one we already sell, or something new to build?
• Is the buyer one we already serve, or a market we would have to learn?
• How many "new" answers did I give? Zero is penetration, one is development, two is diversification — and the count is your best quick proxy for how much validation the idea needs.

The matrix will not tell you whether growth direction is even your most pressing question; an early team with no retention has a penetration problem to fix before it debates new markets. If you are unsure this is the right lens at all, our guide to which strategy framework a startup should use helps you match the tool to the decision in front of you. Whatever you track it in — a whiteboard, a spreadsheet, or a workspace like Edmired — the goal is to keep every bet and its risk level visible in one place, so the tempting leap and the sober next step sit side by side.

A worked example: sequencing growth bets one square at a time

The matrix earns its keep when you use it to sequence, not just classify — starting in penetration and moving one square at a time. Here is the logic on a deliberately hypothetical startup; treat every detail as illustrative.

Imagine ClinicCal, a hypothetical scheduling app built for independent physiotherapy clinics. Its four Ansoff paths might look like this:

• Market penetration (start here): win more physio clinics in the target market, deepen usage inside each one, and push retention up. Same product, same market, lowest risk — and the foundation everything else rests on.
• Market development (one axis out): take the identical scheduling app to adjacent clinic types — dental, chiropractic, veterinary. The product is proven; the open question is whether a new kind of clinic adopts it the same way.
• Product development (the other axis out): build a new module — billing, patient reminders, insurance claims — for the physio clinics it already serves. The customers are known; the question is execution and whether they want a second product from ClinicCal.
• Diversification (last, if ever): launch a consumer-facing appointment-booking app for a wholly different industry. New product, new market, both unknown at once — the bet to postpone until the core is strong.

The sequence matters more than any single square. ClinicCal should exhaust penetration before spending on the others, because penetration gains are cheapest and they build the customer base and cash that fund riskier moves later. This is why most startups should live in market penetration first, dominating a narrow beachhead market completely before expanding along any axis. When it does expand, it should pick one adjacent square — development, not diversification — so it is only ever learning one new thing at a time. Jump straight to a new product for a new market and you are debugging two failures at once, with no stable base to fall back on.

When each Ansoff quadrant is the right move

Each quadrant becomes the right move at a different moment, usually signaled by where your growth has stalled and how strong your core is. Use these signals to decide which square deserves attention now.

Takeaway: There is no universally correct quadrant, only the right one for your stage. Early on, the honest answer is almost always penetration; the discipline is resisting the more exciting squares until your core can fund and survive them.

For a pre-revenue or early-revenue startup, the default is penetration, and the burden of proof sits on anything else. Diversification in particular is where young companies most often overreach — chasing a shiny new market with a new product while the original one is still unproven. Ansoff's gradient is a reminder that novelty is not free: every step off the penetration square should buy enough upside to justify the unknown it adds.

Key Takeaways

• The Ansoff matrix sorts growth by novelty, crossing existing-versus-new products with existing-versus-new markets to produce four strategies with four different risk levels.
• Market penetration is the lowest-risk quadrant — more of the same product to the same market — because every major unknown has already been resolved.
• Market development and product development each add exactly one unknown, a new market or a new product, making them the natural next steps after penetration.
• Diversification is the riskiest strategy, because a new product and a new market stack two sets of unknowns at once with no existing base to cushion the bet.
• Read risk by counting new axes: zero for penetration, one for either development path, two for diversification.
• Most startups should win at penetration first, then expand one square at a time rather than leaping across the matrix into diversification.
• The matrix classifies and sequences; it does not validate. Every square off penetration is a hypothesis about a new market or product that still needs real customer evidence.

Frequently Asked Questions

Who created the Ansoff matrix and when?

The Ansoff matrix was created by H. Igor Ansoff, a mathematician and business strategist often called the father of strategic management. It first appeared in his 1957 Harvard Business Review article "Strategies for Diversification" and was expanded in his 1965 book Corporate Strategy. It is also known as the product-market expansion grid.

Which Ansoff matrix strategy is the riskiest?

Diversification is the riskiest Ansoff strategy, because it combines a new product with a new market — so both the offering and the buyer are unknown at the same time. Market penetration is the safest, since neither changes. Market development and product development sit in between, each introducing a single new unknown rather than two.

What is the difference between market penetration and market development?

Market penetration grows an existing product within its existing market, winning more share, usage, or retention from the customers you already serve. Market development takes that same existing product to a new market, such as a different segment, geography, or use case. Penetration changes neither product nor buyer; development changes only the buyer.