Strategy Decide

Ansoff Matrix

The Ansoff matrix is four routes to growth from existing or new products in existing or new markets, each carrying a different level of risk.

The questionShould we grow by selling more of what we already sell, taking it to new customers, building new products, or doing both at once?

Also called the product-market matrix or the product-market growth matrix. Ansoff's 1957 article uses "missions" for what later slides call markets.

Matrix · H. Igor Ansoff, 1957

New Markets Existing

Market development Same products, new customers or geographies.

Diversification New products and new markets. Highest risk.

Market penetration More of the same, to current buyers.

Product development New products for customers you already serve.

Existing Products New

Four growth routes from existing or new products and existing or new markets. Risk generally rises toward new/new.

When to use it

A management team wants growth and the ideas on the table mix share-gain, new channels, new SKUs, and a new business, with no way to compare risk.
A single business, not a portfolio of units, has to choose its next growth path.
The CEO has set a growth rate the core market cannot deliver, and you need to show which boxes would have to do the work.
Adjacent-looking options have to be sorted by how many of product, customer, and capability are actually new.

When to avoid it

The question is how to split cash across several existing businesses. The BCG matrix or the nine-box is the portfolio tool; Ansoff is a growth-path tool for one business.
Industry structure is unknown. A new-market box drawn without knowing buyers and rivals is a wish. Do the market work first.
The live issue is how to win where you already play, not whether to change product or market. Generic strategies or Playing to Win fit better.
Diversification is already a done deal and the work is integration. The matrix will only relabel a decision that has been taken.

Overview

The Ansoff matrix lays out four growth routes for a single business: sell more of the current products in the current market, take those products to a new market, build new products for current customers, or do both at once. Risk generally rises as product and market become new, and diversification is the cell that usually demands new capabilities on both fronts. The decision it supports is which path to fund, given a growth target the core may not be able to meet.

The grid is only as honest as the definition of “existing.” A new channel or a new SKU filed as penetration hides the risk, and funding all four boxes is a refusal to choose. Pair each cell with evidence of demand and with a view of who already competes there, then pick one primary path and a stop rule for any adjacent test.

The four routes

Market penetration

Existing products in existing markets. Growth comes from winning share, raising usage among current buyers, or finding non-users inside the served market. It is usually the lowest-risk box because the product and the customer are known. It is also the box that saturates first. Tactics include price, promotion, distribution intensity, and buying a rival in the same market.

Market development

Existing products in new markets. New can mean a new geography, a new channel, or a new customer group, such as selling a household product to industrial buyers. The product is proven; the unknown is demand, access, and the local competitive set. Risk sits in the market side, including channels and regulation you have not run before.

Product development

New products for existing markets. The firm uses a customer relationship, brand, or channel it already has, and asks those buyers to take something they do not buy today. Risk sits in development, launch, and whether the current channel can sell the new thing. A new product that also needs a new buyer is not this cell; it has already become diversification.

Diversification

New products in new markets. Ansoff singled this out as the move that demands new skills and new market knowledge together. Related diversification stays close to a technology, channel, or customer logic the firm has. Unrelated diversification does not. It is the highest-risk cell for the operating team, even when a conglomerate logic claims it lowers portfolio risk.

What it is not

A portfolio cash map. It does not tell you which existing units fund which. Use BCG or the nine-box for that.
A promise that penetration is safe. A price war in a known market can destroy more value than a careful adjacent move.
A complete strategy. It names a growth path, not how you will win on it, and not whether the industry is worth growing in.

The steps

Freeze what "existing" means

Write the current product line and the current served market in one paragraph each. If a product variant or a customer group is already 20 percent of sales, it is not new. Arguments about the grid are usually arguments about this definition.
Fill four boxes with real options

Put named initiatives in each cell, with a revenue range, a cost, and the capabilities each would add. An empty cell is fine. A cell stuffed with slogans is not a strategy.
Score risk as newness, not as optimism

Count how many of product, customer, channel, and capability are new in each option. Diversification should score highest unless the firm already owns those capabilities from another business.
Test demand before capex

For any box other than penetration, get evidence that a buyer will take the offer. A three-city test, a broker letter of intent, or a customer co-design beats a market-size slide.
Choose a path and a stop rule

Pick one primary box and at most one modest adjacent test. Write what would cause the adjacent test to be killed, and what share or revenue the primary box has to deliver before a second front opens.

In practice

A $65 million specialty coffee roaster sold bagged beans through US grocery and held about 8 percent of the specialty-grocery bean category. The CEO wanted 15 percent annual growth. Penetration to 12 percent share, via $4 million of extra trade spend and a second SKU at the same retailers, was modeled at roughly $18 million of added revenue in three years. Market development into hotel and office foodservice with the existing blends was modeled at $12 million, and needed a broker network the firm did not have. A ready-to-drink can for the same grocery buyers needed $8 million of capex and was modeled at $22 million in year three if it reached 2 percent of specialty RTD. Forty cafes were diversification: $28 million of capex and a labor model the firm had never run.

The consultant ranked the boxes by evidence of demand and by how many of product, customer, and capability were new. The board funded penetration and foodservice, put $1.5 million into a three-city RTD test, and killed the cafes. Risk rose as both product and customer became new, and the cafe plan asked the firm to learn both at once.

What you leave with

A written definition of current products and current markets, so "new" has a meaning.
Four boxes with named options, or an explicit empty, each with revenue, cost, and new capabilities.
A risk ranking based on how much is new, not on how much the sponsor likes the idea.
A chosen primary path, an optional test, and the evidence required before the test scales.
Stop rules and a date to look again, especially if penetration is already saturating.

Failure points

Everything called penetration

A new channel, a new country, or a new SKU is filed as "selling more to our market" so it looks safe. The definition of existing was never frozen, and the risk was hidden by the label.

Diversification to hit a growth number

The core cannot deliver 15 percent, so cafes, acquisitions, and unrelated products appear in the same breath. Ansoff's point was that this box needs new resources on two fronts. Treat it as a separate business case, or drop it.

New product, new customer, old cell

A product development project quietly needs a different buyer, or a market-development project needs a redesigned product. That is diversification by stealth, and it will need the budget and skills of that cell.

Ignoring the competitor's move

The grid is filled as if rivals will sit still. Penetration against a cost leader, or market development into a geography a rival already owns, is a fight, not a blank market. Pair the cell with a view of who is already there.

Four boxes, four funded programs

Each function gets a box and a budget, which is a way not to choose. The matrix is a decision aid. One primary path, and a small test if you must, is the use that earns the drawing.

The failures above are usually Framing: the wrong problem was chosen before this framework was applied.

Origin

H. Igor Ansoff published "Strategies for Diversification" in Harvard Business Review, September–October 1957. He defined a product-market strategy as a joint statement of a product line and the missions those products are meant to fulfill, and he set out four growth alternatives from present or new products in present or new missions. The 2x2 that later teaching calls the Ansoff matrix is that set of alternatives. He treated diversification as a different kind of move from the other three, because it usually requires new technical, financial, and merchandising resources at the same time. Corporate Strategy (McGraw-Hill, 1965) placed the same logic inside a broader method for top-management decisions on growth and expansion.

Later textbooks split diversification into related and unrelated, which Ansoff's 1957 article approached through vertical, horizontal, and lateral directions rather than through those labels. The matrix is not licensed. The 1957 article and the 1965 book are copyrighted. Some training businesses use "Ansoff Matrix" as a course title; that does not restrict the original.