Market definition

Market definition is the boundary a product strategy draws around the buyers it intends to serve, and it is the first of the choices about where to compete.

Theodore Levitt gave the argument its most quoted form in Marketing Myopia, published in Harvard Business Review in 1960. The American railways declined, he wrote, because they had decided they were in the railway business when the business they were in was transportation. Derek Abell turned the same observation into a method in Defining the Business, published in 1980, by splitting the single boundary into three.

A boundary drawn in one sentence decides a great deal. It sets the denominator under every figure that follows, which means the market size, the share the plan assumes and the revenue the business case promises. If a product manager draws the boundary carelessly, nobody outside the team can check any of those numbers.

The sections below set out the boundary a definition draws and the four things a usable definition contains. One market is then defined four different ways, so the count each boundary produces can be compared against the other three. Finally they take up the argument over how wide to draw a boundary, and the arithmetic that decides whether a market is large enough. The page ends with where each syllabus puts the material.

The boundary a market definition draws

Levitt's railways drew their boundary around a technology. That boundary counted every competitor running on rails and none of the lorries and aircraft that took the freight away. Product teams make the same mistake today with a product category. If the market is defined as the buyers of refrigeration maintenance software, the count covers only the organisations that have already bought some. Every operator running the job from a spreadsheet falls outside that boundary. That group is larger. It is also easier to sell to.

Published classifications draw the boundary the same way, around suppliers. The United Kingdom Standard Industrial Classification has been in force since 1 January 2008, and it sorts businesses by what they produce. Its North American counterpart is revised on a similar cycle, and the 2022 revision created 111 new industries out of the 156 that had existed in 2017. Both are built for counting economic output. Neither was built to say which buyers share a problem. So a market size lifted from either of them answers a question nobody on the product team asked.

Derek Abell replaced the single boundary with three in 1980. A business is defined by the customer groups it serves, the customer functions it performs for them and the technologies it uses to perform them. Choosing a span on each of the three settles the boundary. A careless choice on any one of them makes the other two impossible to settle.

Those three spans explain why a market definition is a choice and never a discovery. A supplier of refrigeration maintenance scheduling can widen any one of them.

• The customer group, from grocery chains to every operator of chilled stock.
• The customer function, from scheduling engineers to preventing stock loss of every kind.
• The technology, from software to an engineer in a van.

Every one of those widenings is available, and each one costs something different. Only the organisation making the choice can say which cost it is able to carry.

What a usable market definition contains

Choosing the three spans deliberately still leaves the question of what the written definition has to say. Four things separate a definition that changes decisions from one that only reads well in a deck.

Who the buyers are, described by something that predicts behaviour. An industry code and a headcount are both easy to get and both weak predictors. A shared situation predicts better, such as operating across more than one regulatory regime, running a fleet the operator does not employ or holding stock that spoils within four hours of a compressor failing.

How many of them there are, with a stated method behind the number. A figure with nothing behind it is a wish. Most business cases fail quietly at the arithmetic.

What the buyers do about the problem today, which is the real competitor in most markets and is usually a spreadsheet, an agency or nothing at all.

How the buyers make a purchase, which is the input to distribution strategy. A segment buying through resellers and a segment buying on a corporate card need different products. The difference goes past marketing and into what the product itself has to do.

How a buyer makes a purchase is the part most often handed to marketing and then forgotten. It decides what the product has to do on its first day. A product built for a corporate card rarely survives the first procurement review it meets.

One market defined four ways

Those four elements are easiest to judge against a real market. One example is a product that schedules refrigeration maintenance for grocery chains in the United Kingdom. Four boundaries are available for that product, and each one produces a different count.

Only the boundary that names both a problem and a buyer can carry a plan. The other three are the ones that actually get drawn, because none of them costs any work. A product category is what the product already is. An industry is what a published report already counts. A problem on its own makes the opportunity look large without ever saying who would sign for it. None of those three boundaries was chosen.

The argument over how wide to draw a boundary

The table argues for the narrowest of the four boundaries. That advice has never been unanimous. Levitt's advice in 1960 was to define the business broadly, and it has been argued with ever since. Geoffrey Moore gave the opposing answer in Crossing the Chasm in 1991. A product entering a market should name a beachhead small enough to dominate and then serve it completely, because a reference from one satisfied buyer only reaches buyers who look like that buyer.

A second objection comes from marketing science. Byron Sharp argued in How Brands Grow in 2010 that brands grow by reaching all the buyers in a category. He also argued that buyers see little difference between brands, and that the data gives little support to the segmentation and targeting most marketing plans rest on. John Dawes, a colleague of Sharp's at the University of South Australia, added the figure most often quoted from that work in 2021. Writing for the LinkedIn B2B Institute, he put the replacement cycle for services such as banking, software or telecoms at roughly five years. That leaves about five per cent of potential buyers in the market in any given quarter.

Sharp and Moore are describing two different purchases under one word. Sharp describes how an established brand grows in a large consumer category, where almost everybody already buys and the limit is memory. Moore describes how an unknown product wins its first reference customers, and there the limit is credibility. A product manager writing a strategy for a product nobody has heard of stands in Moore's situation. The useful form of Sharp's objection is that a narrow definition is temporary. The narrow boundary is the entry boundary, and widening it later has to be a decision somebody makes on purpose.

Large enough for the business the product has to become

Widening a boundary later is also what the arithmetic has to allow for. A definition has to carry the business the product will have to become, and that is a longer horizon than the year it is serving now. A segment can support the product today and still cap out below the revenue the company needs in three years. That segment is a dead end, and it will look healthy for two of those three years.

The check is arithmetic, and it is worth doing before anything else is written. Under the boundary that names a problem and a buyer, 6,820 stores sit inside the segment. Once the chains with no maintenance contract are removed, 5,456 of them qualify, and an annual licence of £1,400 a site puts the whole serviceable market at £7.6 million. If a three year plan calls for £20 million of annual revenue, the company cannot reach that figure inside this definition even at complete dominance. Either the definition is wrong, the price is wrong or the plan is.

That figure is a ceiling, so the plan has to fit underneath it. Naming who the buyers are is the first half of the work. Counting them, with a method another person can check, is the second half. The count is what turns a definition from a sentence into a constraint.

Where each syllabus puts it

The Pragmatic Framework places market definition in its Focus category, beside Distribution Strategy, Product Portfolio and Product Roadmap. The instruction there is to map needs to target markets, to analyse which segments to pursue actively and to make sure the targeted segments are large enough to support the current and future business of the product. A second instruction in the same category asks which channels best match the buying preferences of those markets. That is why a usable definition has to say how a buyer makes a purchase. Pragmatic PMC examines the material beside market problems and buyer personas, so a candidate meets it as the front half of discovery.

The Product Strategy Practitioner course puts the same work at the opening of its module on where to compete. The boundary is the same boundary. What changes is what sits underneath it. The three pages after this one count the group this page has named, divide it and then contest it.