Ansoff matrix: four growth paths and why diversification costs most

Suppose you lead sales for a regional food brand, and the board has set next year's revenue target at 20 per cent growth. Four proposals are on the table: sell harder in the supermarkets you already supply, take the product to a new region, launch a new product line, or start a different business altogether. Which one goes first, and what must be proven before any money is spent? H.I. Ansoff's Ansoff matrix puts the four proposals into four boxes so you can see which unknown each one is hiding.

What are the four growth paths in the Ansoff matrix?

The matrix asks two questions. Is the product existing or new? Is the market existing or new? Each combination is one growth path (Ansoff, 1957; Ansoff, 1965). Reading from the bottom left: more of the existing product to existing customers is market penetration; the existing product to new customers is market development; a new product for existing customers is product development; a new product for new customers is diversification.

Penetration and market development carry less risk because the firm already knows either the product or the customer. The matrix marks penetration as low risk, the two middle boxes as medium and diversification as the highest.

How would a food brand place four proposals in the matrix?

A hypothetical example. The brand, its sales and every figure below are invented to show the method. They are not data from any real business. The brand makes jam and dried fruit in Can Tho, sells through supermarkets in Can Tho and Ho Chi Minh City, and has revenue of VND 40 billion this year. A 20 per cent target means VND 8 billion more.

Ordered by how much evidence each needs, the sequence in this example is penetration, market development, product development, diversification. The reasons:

The two middle boxes can swap places. Here Hanoi comes first because it needs only a trial in a few dozen stores and no change at the factory, while a snack bar needs a recipe and new equipment. In another firm the order could be reversed. The rule is to test first whichever unknown is cheapest to check.

Added up, the first three paths give VND 7.5 billion, close to the 8 billion target. If those figures hold, you do not need the fourth box to hit next year's target, and the VND 0.5 billion gap can be closed by pushing penetration further. If they do not hold, that is the moment to look at the cafés, and to look with evidence rather than hope.

Why does diversification need the most evidence?

Because it changes the product and the customer at once. Each of the other three boxes keeps at least one thing you already know, so past sales data still helps you forecast. In the diversification box almost nothing carries over.

You can test your own proposals with the table below. For each one, fill the second column with real figures you already hold, then see how long the third column gets.

Diversification does fit sometimes. It fits when the current market has hit a ceiling, when the firm has assets or skills that transfer, and when it can afford to be wrong. It does not fit when the only reason is this quarter's target. Relatedness is worth measuring: two activities are related when they share technology, production or distribution, and the more related they are, the easier it is to share resources. In the example above, fruit supply is the common ground. Running a café and serving walk-in customers are not.

Which box are Disney's two big acquisitions in?

The public documents do not say. The placement below is my own reading, and I chose this example because it shows the answer moving with how you define product and market. The facts come from two press releases issued by Disney itself, covered in the Disney acquisitions case.

On Marvel, the release of 31 December 2009 says Disney had completed its acquisition of Marvel Entertainment. Marvel shareholders were to receive a total of US$30 a share in cash plus approximately 0.7452 Disney shares for each Marvel share. The release describes Marvel's assets as a library of more than 5,000 characters featured in various media over 70 years, and businesses in licensing, movie production and publishing (The Walt Disney Company, 2009).

On 21st Century Fox, the release of 19 March 2019 says the acquisition would take effect on 20 March 2019, with a total equity value and a total transaction value both of approximately US$71 billion, against Disney's US$59.4 billion of revenue in fiscal year 2018. Disney said the deal would increase its international footprint and expand its direct-to-consumer offerings, and expected at least US$2 billion in cost synergies by 2021 (The Walt Disney Company, 2019). That is an expectation stated at the time, not an outcome.

Now try placing them. If the product is Disney-branded content and the market is Disney's existing audience, Marvel adds content for the same audience: product development. For Fox, the aim of a bigger international footprint suggests market development, while the direct-to-consumer aim suggests a new product. One deal can sit in two boxes depending on which stated purpose you read.

Both releases also call Disney a diversified entertainment company, with five business segments in 2009 and four in 2019. That describes how the segments are organised. It is not Ansoff's diversification box. The usable lesson: before placing a proposal in a box, write down your definitions of product and market. Two people reading the same release can place it differently and neither be wrong, if their definitions differ.

Going deeper: where does the Ansoff matrix come from, and where does it stop working?

The site's library gives two origins for the matrix: Ansoff's article "Strategies for diversification" in the Harvard Business Review in 1957 (Ansoff, 1957), and his 1965 book Corporate strategy, subtitled as an analytic approach to business policy for growth and expansion (Ansoff, 1965). The title of the 1957 article shows its subject was diversification. I rely only on publication details and the library's description, and do not quote either work.

The first limit is that risk does not rise evenly across the grid. The framework library lists this as a common pitfall. Real risk depends on internal capability, not only on the box. A firm with the capability to run retail outlets may find a café far less risky than a firm that has never run a service business. Penetration, the lowest box on the grid, can still fail if rivals respond. The grid has no axis for competitors.

The second limit is that each axis has only two values, "existing" and "new". Does new packaging or a new flavour count as a new product? Does a neighbouring province count as a new market? The matrix does not say. You define it, and as the Disney example shows, the definition decides the box.

The third limit lies in how deals are justified. Large acquisitions are often defended with synergy. Disney's release gives a specific figure and names its source: cost savings from operating efficiencies when the businesses are combined (The Walt Disney Company, 2019). When a synergy claim names no source like that, ask what the source is.