What an Ansoff matrix is
An Ansoff matrix is a two-by-two grid for planning growth. One axis is products, existing or new. The other is markets, existing or new. The four cells are market penetration, product development, market development and diversification, and risk rises the further a move gets from what a company already sells, and to whom.
It is named after Igor Ansoff. It does not tell you which move to make. It makes you say which kind of move you are making, and what you do not yet know.
How the Ansoff matrix works
Each cell is a combination of the two questions: is the product new, and is the market new.
Market penetration
Existing products, existing markets. Sell more of what you have to the buyers you already reach: more often, at a better price, or won from a rival. You know the product and you know the buyer.
Product development
New products, existing market. Sell something new to people who already buy from you. The unknown is the product: will it work, can you make it, will your buyers want it from you.
Market development
Existing products, new markets. A new country, a new segment or a new channel. The unknown is the market: will these buyers want it, and what does it cost to reach them. A new country means local prices, shipping, returns, language and support.
Diversification
New products, new markets. Both unknowns at once. Related diversification shares something you are already good at, a supply chain or a brand. Unrelated diversification shares nothing but the owner.
Risk
Each step away from the bottom-left cell adds one unknown, and diversification adds two. That is not a reason to stay put. A company that has already taken most of its market has a penetration problem of its own: there is not much left to take.
Three Ansoff matrix examples
All three companies are hypothetical. Each lists one option per cell, then picks.
A. A subscription coffee brand selling in the UK
- Penetration: move subscribers from monthly to fortnightly deliveries and win back the ones who paused.
- Product development: a cold brew concentrate, sent in the same box to the same subscribers.
- Market development: ship the same beans to Germany and the Netherlands.
- Diversification: lease coffee machines to offices.
It picks cold brew. It knows the buyer and already owns the delivery, so the only new thing is the product. That is also where the risk sits: cold brew needs a different supplier and a shelf life the beans never had.
B. A skincare brand selling in the US
- Penetration: refill packs and a routine bundle for existing customers.
- Product development: a daily moisturiser with sun protection.
- Market development: the same range in Australia.
- Diversification: an ingestible beauty supplement.
It picks Australia. The product does not change, but nearly everything around it does: prices in Australian dollars, local shipping and returns, and rules on what a product page may claim. The cost of a new market is mostly paid before the first order.
C. A project management tool for agencies
- Penetration: sell more seats inside accounts it already has.
- Product development: time tracking and invoicing inside the tool.
- Market development: the same tool for in-house marketing teams.
- Diversification: a marketplace for hiring freelancers.
It picks penetration. Seat growth inside happy accounts is the cheapest revenue it can find. The limit is the ceiling: once every agency seat is sold, the plan runs out, and the next move is one of the other three.
Reading a competitor's Ansoff moves from outside
A competitor will not tell you its growth plan. Its public footprint shows the moves after they start.
New products in the catalogue point to product development, but classify a move by what is actually new. The Paris Hilton x Loop edition is a good test. Loop Earplugs ran its launch week on Instagram and Facebook in the week to 25 August 2026. It is an edition of an earplug Loop already sells, so the product is not really new.
Whether the buyer is new is the question the grid asks. If it sells to Loop's existing fans, it is penetration with a new reason to buy. If it brings in people who never looked at earplugs, it is market development. Loop's range stood at 66 products on 28 September 2026.
New localised sites point to market development. Between 24 August and 3 September 2026, Loop's public sitemap gained a Belgian French section (fr-be) and lost the Dutch one (nl), leaving 14 locale prefixes. The two methods count different things, so neither number is a change on the other. The note on counting localised sites explains the difference.
Compare the sites with where visits come from. In August 2026, modelled figures put 34.7% of Loop's visits in the US, 8.7% in Great Britain, 5.9% in Australia, 5.8% in Spain, 5.6% in Germany and 4.8% in Canada.
A localised site in a country that sends few visits is a bet on the future. A country with many visits and no local site is a market served without the cost. The count of localised versions sits on Loop's benchmark.
New categories point to diversification: a collection that has nothing to do with the core range, a new domain, or job posts for skills the company never needed. The note on job posts as a roadmap covers the last one.
How to use it
- Write down what you sell today and to whom. That is the bottom-left cell.
- List at least one real option in each of the four cells, even the ones you expect to reject.
- For each option, name the unknowns: the product, the market, or both.
- Put a rough cost and a rough size on each, and say how you would test it small first.
- Pick one or two. Say which cell they sit in, so everyone knows which risk you are taking.
- Place your closest competitors' recent moves on the same grid, and check where nobody is moving.
What it cannot tell you
It has no numbers. It does not say how big a market is, how fast it grows, or what a move will earn. For that you need sizing, and for how your current products compare you need a portfolio view like the BCG matrix.
The cells are not as clean as the grid. Is a new flavour a new product? Is a new sales channel a new market? Teams spend time arguing about the cell instead of the move.
It ignores competitors. A new market with an entrenched leader is riskier than the matrix shows, and a new product a rival is about to launch is not new for long.
From outside, you see the move only after it starts. A sitemap section or a new product page shows the decision. It does not show the budget behind it, or whether it is working.
FAQ
What are the four Ansoff strategies?
Market penetration (existing products, existing markets), product development (new products, existing markets), market development (existing products, new markets) and diversification (new products, new markets).
Which Ansoff strategy is the riskiest?
Diversification, because both the product and the market are new. Market penetration is the least risky, because both are known.
Is entering a new country market development?
Yes, if you sell the same product there. If you need a different product for that country, it moves toward diversification.
How is the Ansoff matrix different from the BCG matrix?
Ansoff sorts where to grow next. The BCG matrix sorts the products you already have by market growth and relative market share, to decide where cash goes.
Pick one competitor. See what we find.
Your first benchmark is free. One domain, measured against you, no card and no call.