Ansoff Matrix
The Ansoff Matrix is a four-quadrant grid that crosses the two variables of a growth decision: the product, existing or new, and the market, existing or new. The crossing names four directions, market penetration, market development, product development and diversification. H. Igor Ansoff set out the form in Strategies for Diversification (Harvard Business Review, 1957), then developed it in Corporate Strategy (1965). It answers one question, that of direction: it puts competing growth options on the same footing, each of which arrived with a case of its own, and makes visible what each one forces the organisation to acquire. Sizing the target market, checking the capacity to execute and putting figures on the move belong to other techniques, which take over once the direction has been shortlisted.
Goal
The Ansoff Matrix serves one decision, the direction of growth: by which move an organisation intends to grow. It sorts the growth options in play into four quadrants according to what they demand in product and in market, then puts the executive team in front of a single picture where proposals argued separately until then can be compared. The deliverable is the completed grid, each quadrant carrying one or more options worded concretely enough to be investigated.
The value of the technique lies in the criterion it imposes. An option is judged here on two questions whose answer can be checked: does the organisation keep its product, does it keep its buyers. From that pair come the quadrant and with it the list of what remains to be built, a manufacturing capability, a distribution network, a certification, a sales force. A list of growth ideas becomes a comparison of gaps in capability.
The grid does not size the demand it targets, which is what market analysis does; it does not judge whether the organisation can execute the move, which is what business capability analysis does; it does not put figures on anything, which is what financial analysis and the business case do. Upstream, a SWOT analysis often supplies the gap in position for which the matrix seeks a growth answer. Downstream, the Balanced Scorecard turns the chosen direction into measured objectives.
Usage
When to use it
- An investment budget to allocate: several growth directions compete for the same envelope and one or two of them have to be shortlisted.
- Following a SWOT analysis: a gap in position has been named, the matrix supplies the menu of growth answers.
- A three- or five-year strategic plan: frame the candidate moves before opening the market studies and the investment files.
- A mixed portfolio of initiatives: sort the projects already under way by direction to see where the effort is going.
- An isolated diversification proposal: put it back among the three other directions before investigating it on its own.
When not to use it
- The product-market pair is imposed: a public service mandate or a concession with no room to move on the offer and the buyers, run a process analysis.
- The direction is already settled: the question is the size of the target market and then the return, run a market analysis and then a business case.
Description
The matrix crosses two binary axes. The first is the product: does the organisation sell what it already sells or something it does not yet make. The second is the market: does it address the buyers it already serves or others. The two axes cross into four quadrants, and each quadrant carries the name of a direction and its own economics.
- Market penetration (existing product, existing market): sell more of the current offer to current buyers, through sales effort, promotion, price, loyalty or taking share from a competitor. Neither the offer nor the customer base changes, so the organisation works on what it already holds.
- Market development (existing product, new market): carry the current offer to buyers it does not serve, a new geography, a new segment or a new channel. The manufacturing base stays; customer knowledge, distribution and often certification have to be built.
- Product development (new product, existing market): offer a new or extended product to the customer base already won. The commercial relationship stays; product competence and industrialisation have to be built.
- Diversification (new product, new market): a new offer to new buyers. Neither the manufacturing base nor the commercial relationship carries over, which makes this the quadrant that demands the most learning and takes the longest to produce a result.
Where the technique comes from
Ansoff published the form in 1957 in Strategies for Diversification, a Harvard Business Review article on the question facing a company whose growth is levelling off in its own market. He sets out there the four product-market strategies and their names, still used as they stand. Corporate Strategy, published by McGraw-Hill in 1965, gives the first extended treatment and places it inside a complete method of strategic decision. The matrix has circulated since in varied guises, including three-axis versions that extend this original form.
What separates the four quadrants
The four directions read as four distances from what the organisation already knows how to do, and that distance grows from penetration to diversification. This is where the rule of thumb about rising risk comes from, which makes diversification the most exposed direction. That order is a piece of reasoning about the learning to be financed. It ranks the quadrants by what remains to be acquired and says nothing about the expected return or about the probability of success of a given option: an adjacent diversification, carried by a manufacturing capability that transfers, is handled more safely than a penetration on a saturated segment where every point of share is paid for in discount. The order steers the discussion; the trade-off is then made on figures the grid does not produce.
The procedure
- Name the growth objective
A gap with a figure and a date: take revenue from one amount to another in three years or win a stated number of share points on a stated segment. With no gap named, the grid fills with ideas and ranks none of them. - Define the current baseline
Write down what the current product and the current market cover, in criteria that can be checked: catalogue references, segments served, channels, territories. This definition decides the quadrant each option falls into, and two people who have not written it down will file the same option in two different quadrants. - Populate the four quadrants
An option is worded in one sentence naming the offer, the buyer and the channel. Sweep all four quadrants even when three of them look beside the point: an empty quadrant often signals a habit rather than a dead end. - Qualify each option by what it reuses and what it lacks
For each option, write down what it reuses (product competence, customer relationship, channel, certification) and what it forces the organisation to acquire. - Shortlist
Keep one or two options, usually from two different quadrants, so that a move built on what exists and a move that reaches beyond it are investigated together. - Hand over
Each option kept goes to the techniques that size the demand, check the capability, put figures on the move and set the objectives that will track it.
The pitfalls that hollow out the exercise
Decorative quadrants
The grid is filled in afterwards to dress up a direction already chosen. Three quadrants get a line out of politeness, the fourth takes the project that is already running. The remedy: populate before discussing and open the trade-off only once every quadrant holds at least two options precise enough to be investigated.
What only looks new
The classification turns on what has to be acquired: a competence, a certification, a sales force, a logistics chain. An extra colour in the catalogue acquires nothing and is still penetration; so is one more reseller in the same canton. The two axes are binary while reality comes in degrees. The acquisition criterion, written down before the options are filed, is what holds the boundary.
The grid mistaken for a decision
It ranks directions. It says neither what the target market is worth, nor whether the organisation can reach it, nor what the move returns. A direction chosen on the grid alone is a direction chosen without a figure.
The exercise done once
The grid photographs the state of the product-market pair on one date. A competitor entering, a regulation changing or a channel collapsing move the quadrants without anyone reopening the document. Running the exercise again at every planning cycle and at every marked shift in the market costs half a day.
Diversification as a catch-all
The quadrant brings together moves of different natures: buying a supplier, opening an activity adjacent to the core business or launching an offer with no connection to it. Treating them as a single option flattens the distinction that counts at this stage, what the move reuses. Split the quadrant into named options, each with what it reuses and what it lacks.
AI considerations
The clearest help is in populating the quadrants. From the product catalogue, the customer list by segment and market intelligence, a language model produces a first set of options per quadrant, which addresses the quadrant left empty out of habit. It also files an existing portfolio of initiatives into the four quadrants and flags the wordings too vague to be classified, which exposes the projects nobody can name a buyer for. On the market development quadrant, it summarises the entry conditions for a territory, certification requirements, applicable standards, structure of distribution, which gives a basis that can be checked.
Judgement stays human. What "new" means for this particular organisation is decided on its competences, which the model does not know: it will produce plausible options with no internal support. Qualifying each option by what it reuses and what it lacks is done by hand. The rule of thumb about rising risk is a piece of reasoning, not a measurement; a model returns it as though it were one. The Swiss framing has to be given to the model as well: trained on a worldwide corpus, a model proposes Germany or the United States where Swiss certification, the size of the home market and the structure of distribution call for other moves, and it overlooks the cantonal segments that separate two market development options. Shortlisting commits resources and belongs to the executive team.
Examples
A Bernese manufacturer of ski touring bindings, some sixty people and a revenue of CHF 14'000'000, is looking to take that revenue to CHF 18'000'000 in three years. The options sent up by sales management and the design office are filed into the four quadrants.
Ansoff Matrix · completed grid
Four directions for a Bernese maker of ski touring bindings
| Markets × products | Existing productstouring bindings in the catalogue | New productsan offer to design and industrialise |
|---|---|---|
| Existing marketsSwiss alpine retailers | Market penetrationDemo days and display stands at the retailers of Verbier, Zermatt and the Engadine, with a volume discount over the season. | Product developmentA line of touring boots sold to the same retailers, alongside the bindings. |
| New marketsbuyers the company does not serve | Market developmentThe current binding range carried to Austrian and French retailers, through one agent per country. | DiversificationMountain bike components machined in the existing workshop, sold to bicycle assemblers. |
The existing markets row brings together two moves that keep the customer base and part company on the competence that has to be built; the existing products column brings together two moves that keep the workshop and part company on the network that has to be set up. The diversification quadrant is the only one whose wording reuses neither the product nor the customer base, which is enough to place the direction before any costing.
Visualisations
The technique produces two figures. The first is the structure: two axes that cross and four quadrants whose meaning comes from where they fall. It is drawn as a grid of four quadrants, where the geometry carries the lesson: an option is qualified by its product and by its market, and the two together decide its quadrant; a gradient running from the penetration corner to the diversification corner shows the growing distance from what is already held. The second is the deliverable, the completed grid, made of rows and columns: the two axes in the margin, the option and its qualification in each quadrant.
Cost
| Phase | Level | Justification |
|---|---|---|
| Preparation | Medium | Gathering the options already in circulation, writing the definition of the current baseline in criteria that can be checked and bringing the decision-makers together. Rushed, this framing produces quadrants that overlap. |
| Execution | Low | A session of two to three hours is enough to populate the four quadrants, qualify each option by what it reuses and what it lacks, then shortlist. |
| Documentation | Low | The grid fits on one page. The cost moves downstream, into the market studies and investment files that the shortlisted options set off. |
Tooling
The technique is run as a workshop and calls for no dedicated tool. A whiteboard, four quadrants drawn on it and sticky notes are enough in a meeting room; a shared whiteboard such as Miro or Mural carries the same grid for a distributed executive committee and still lets someone move an option from one quadrant to another, often the useful moment of the session.
The substantive work happens around the session. The product catalogue, the customer relationship management system and the ERP supply the definition of the current baseline, segments served, channels, territories, references sold, without which the quadrants stay arguable. For the market development quadrant, Swiss and regional market sources, the Federal Statistical Office, industry associations, trade publications, give the material that market analysis will take up. A spreadsheet holds the qualified grid beyond the workshop, one row per option, with what it reuses, what it demands and the study it sets off.
Sources
- Ansoff, H. I., Strategies for Diversification, Harvard Business Review, vol. 35, no. 5, September-October 1957, pp. 113-124: the original publication, where the four product-market strategies and their names are set out.
- Ansoff, H. I., Corporate Strategy: An Analytic Approach to Business Policy for Growth and Expansion, McGraw-Hill, 1965, ISBN 0-07-002627-1: the first extended treatment of the matrix, placing it inside a complete method of strategic decision.