Segmentation, targeting and positioning (STP): choosing who to serve

A common way for a small business to waste its marketing budget is to try to appeal to everyone. The message ends up fitting nobody in particular, and the money is spread across groups that were never going to buy. Segmentation, targeting and positioning (STP) is the standard way to make the opposite choice on purpose: decide who you serve, and what you want those people to think of you.

The three steps run in order. Segmentation splits a market into groups. Targeting picks which groups to pursue. Positioning decides how you want to be seen by the groups you picked. Each step depends on the one before.

Step one: segmentation

Wendell Smith is generally credited with bringing the idea into the marketing literature in a 1956 article, according to Wikipedia’s overview of market segmentation. The aim is to divide a market into meaningful groups of current or potential customers who would respond differently to what you offer.

Most segmentations draw on a handful of bases:

Demographics are the easiest to get and often the weakest predictor. In a 2006 Harvard Business Review article, Rediscovering Market Segmentation, Daniel Yankelovich and David Meer argued that segmentation had drifted towards building advertising characters, and that lifestyle personas were “no better than demographics at predicting purchase behavior”. They urged companies to segment for a specific decision, such as which market to enter or what to build, and to focus on behaviour that can be changed or needs that are not being met.

A segment is only useful if you can act on it. The Wikipedia overview lists tests a segment should pass: it should be identifiable, substantial enough to be worth serving, accessible through channels you can reach, responsive in a different way from other segments, workable with the resources you have, and reasonably stable over time.

Step two: targeting

Once you have segments, you decide how many to serve and how differently. The common options are:

• Undifferentiated (mass) marketing: one offer and message for the whole market. This works for basic goods with little variation in need, and needs scale.
• Focused (niche or concentrated) marketing: one segment, served very well. This is where most small businesses should start.
• Differentiated marketing: two or more segments, each with its own offer or message. This costs more because every segment needs its own work.
• One-to-one (hypersegmentation): tailoring to individuals, usually through data and automation.

To choose, score each segment on two questions. How attractive is it (size, growth, willingness to pay, cost to reach)? And how well can you win it (fit with what you already do well, strength of competitors, your credibility)? A segment that scores high on both is your first target. A large segment you cannot win is a distraction, however tempting the numbers look.

Step three: positioning

Positioning is the place you want to hold in your target customers’ minds compared with the alternatives. The term was popularised by Al Ries and Jack Trout through trade articles in 1969 and 1972 and their 1981 book Positioning: The Battle for Your Mind, as Wikipedia’s article on positioning records. Their phrase for it was “an organized system for finding a window in the mind.”

A useful tool here is the perceptual map. As Wikipedia’s entry on perceptual mapping explains, it plots how customers see brands in a category on two dimensions, such as price and quality. Maps built from customer surveys are more reliable. Intuitive maps drawn from a marketer’s own knowledge are quicker but “not based on consumer data”, so treat them as hypotheses.

Maps help you spot gaps. Be careful, though: an empty corner may mean nobody wants that combination. Check for demand before you move into it.

Positioning is usually written down as a short internal positioning statement covering the target customer, their need, your category, your main benefit and how you differ from the main alternative. It guides copy and decisions but is rarely published word for word.

An example (illustration only)

A small coffee roaster sells beans online across the UK. It has been selling to “coffee lovers”, which describes almost everyone. The founders list four behavioural segments and score each from 1 to 5.

They pick home espresso enthusiasts as a focused target and keep gift boxes as a seasonal side line. Their intuitive perceptual map uses two axes: price per bag, and how much detail the seller gives about freshness and origin. Supermarket brands sit at low price and low detail. Speciality subscriptions sit at high price and high detail. The roaster aims for mid-price with high detail, printing the roast date and recommended espresso settings on every bag. Before committing, they ask fifteen existing customers to place the brands on the same map, to see whether customers share their view.

Common mistakes

• Segmenting only by demographics. Two 35-year-olds with the same income can want very different things. Start with behaviour and needs.
• Choosing segments you cannot reach. If you do not know where a segment gathers or how to speak to it, it fails the accessibility test.
• Targeting too many segments too early. Each extra segment needs its own message and often its own offer.
• Positioning on something customers do not value. Being “the only one” with a feature matters only if buyers care.
• Treating the map as fact. An intuitive map is the team’s opinion until customers confirm it.
• Changing position every quarter. Customers need time and repetition to form a view of you.

When to use STP

Use it when you launch, when growth stalls, when you add a product line, or when your marketing feels like it is talking to everyone. Revisit it when your market shifts, for example when a new competitor arrives or customer needs change. Wikipedia’s positioning article notes that repositioning can be a high-risk strategy, so make the change deliberately and stick with it once made.

STP is less useful as a one-off workshop output that nobody acts on. The test of a good segmentation is whether it changes a decision: where you spend, what you build, or what you say. If it changes none of those, go back to step one.