How Brands Grow

The book begins with a challenge to the marketing profession itself. Marketing is a creative discipline, but Sharp argues that this does not mean it should be free from testable regularities. An architect can be creative, but cannot negotiate with gravity. In much the same way, marketers can create, but they should not ignore observable patterns in buyer behaviour.

The Crest versus Colgate example

Sharp uses an example involving two major toothpaste brands. The smaller brand appears worryingly dependent on people who also buy competitors. Its buyers appear less loyal and less likely to name it as their preferred brand.

The conventional response would be to search for a “brand problem”: weak positioning, insufficient emotional connection, the wrong customer segment or a need for more persuasive advertising.

The book proposes a different explanation. Much of this performance is simply what we should expect from a smaller brand. Lower market share tends to come with lower penetration and slightly lower loyalty. Some metrics that appear to be causes may therefore be consequences of brand size.

This becomes the foundation for the rest of the book: before inventing a story about the data, understand what normal behaviour in the data looks like.

Sales depend on two fundamental quantities: how many people buy and how often each person buys. In theory, a brand could become large with a small number of customers buying continuously. In practice, comparisons between real brands reveal a very different picture.

As market share increases, the number of buyers changes dramatically, while average purchase frequency changes much less. Large brands are therefore large primarily because more people buy them.

The Double Jeopardy Law

A small brand suffers twice: it has fewer buyers and those buyers also buy it slightly less frequently. The important word is slightly. Differences in loyalty are generally much smaller than differences in penetration.

What about niche brands?

The book is sceptical of the romantic idea of a small brand supported by a tiny but exceptionally loyal customer base. Such cases exist, but they are less common than popular marketing literature implies. What appears to be a niche can sometimes simply reflect limited geographical, physical or mental availability.

What about cross-selling to existing customers?

Cross-selling is possible, but Sharp explains why it should not automatically be treated as a massive growth engine. A customer who buys one service from you is not necessarily an untapped warehouse of future revenue.

One of marketing’s most famous mantras says that retaining an existing customer is many times cheaper than acquiring a new one. Another popular claim promises extraordinary increases in profit from very small improvements in retention.

Sharp is not attacking the idea of taking care of customers. He is attacking the exaggerated conclusions built upon it. In particular, he questions mathematical examples in which enormous improvements in retention are treated as easy and almost cost-free.

The real issue is that every brand loses customers. Smaller brands tend to lose a larger percentage of their own customer base, while larger brands lose a smaller percentage. This again reflects Double Jeopardy.

Growing brands still lose customers

This is one of the most useful insights in the book. Growth does not require customer defection to disappear. Growing brands continue to lose customers – they simply acquire a large number of new ones. Conversely, shrinking brands often suffer primarily from weak customer acquisition rather than an extraordinary epidemic of defections.

The “average customer” can be a statistical fiction. With Coca-Cola, for example, average buying frequency may look relatively high while the typical buyer purchases much less frequently. A small number of extremely heavy buyers pull the average upwards.

Every large brand therefore has a long tail of people who buy it only occasionally. Individually they appear almost uninteresting. Collectively, they can represent enormous volume.

80/20 is not a universal answer

The popular version of the Pareto principle claims that 20% of customers produce 80% of sales. The evidence discussed in the book shows substantially less concentration. Across many categories, the most active 20% of buyers contribute closer to roughly half of purchases than 80%.

That distinction matters. If the remaining 80% account for around half the volume, ignoring them suddenly looks much less sensible.

The Law of Buyer Moderation

There is another complication. Someone classified as a “heavy” buyer this year is likely, on average, to buy somewhat less next year. Some non-buyers will make purchases. Some light buyers will become more active. This may occur without any fundamental change in preferences – simply because purchasing naturally fluctuates over time.

This chapter challenges one of the foundations of conventional segmentation: the assumption that different brands naturally belong to different types of people.

Studies covering automobiles, credit cards, consumer goods, services and other categories show that competing brands generally have customer bases that look remarkably similar in demographics, attitudes, values and media habits.

This does not mean that no differences exist. Expensive products naturally attract more people who can afford them. A children’s product may naturally over-index among households with children. But after obvious functional constraints such as these, differences between rival brands are often much smaller than expected.

The surprising gift of this discovery

If competitors’ customers are broadly similar to your own customers, that is good news. It means there is no invisible structural wall preventing you from winning them.

Your potential market is often larger than you think.

If customers of different brands overlap substantially, this also changes the way we should think about competition.

The Duplication of Purchase Law says that brands generally share customers with other brands roughly in line with the size of those other brands. A large competitor will therefore appear in the buying repertoire of many more of your customers than a small competitor.

This means that the most obviously similar product competitor is not necessarily the only, or even the most important, source of shared buyers.

Markets are broader than they appear

Managers often like narrow definitions of categories. Market share then looks larger, competition smaller and targets more comfortable. Real consumers, however, can move between formats, price levels and variants far more freely.

Cannibalisation is not automatically a tragedy

If one company owns several similar brands, they will inevitably take some sales from one another. Similarity alone is not a sufficient reason to close a functioning brand. Economic viability, costs and the value of existing market-based assets matter more.

Buyers really are loyal. They do not select randomly from the entire market every time. They develop small personal repertoires and return to familiar brands. This saves time, effort and perceived risk.

But this loyalty is rarely exclusive. A person can like a brand, buy it regularly and at the same time buy competitors. Sharp describes this as divided or polygamous loyalty.

What about brand fans?

Almost every famous brand can produce a spectacular fan: somebody with a tattoo, collection, online community, personal story or intense emotional attachment. The problem is scale. These people are memorable precisely because they are unusual.

The book even examines Apple and Harley-Davidson, traditional examples of “cult” brands, and shows that their actual repeat-purchase and cross-brand buying patterns are substantially more ordinary than the legend suggests.

The Natural Monopoly Law

Large brands attract a disproportionately greater share of light buyers of the category itself. If somebody buys a soft drink only once a year, the probability that this one purchase will be from the largest brand is relatively high. This further explains why growth requires reaching occasional buyers, not just the most enthusiastic ones.

This is perhaps the most famous and provocative chapter in the book. Conventional marketing insists that a brand should possess a unique, meaningful and valuable difference.

Sharp and Jenni Romaniuk distinguish between two very different ideas: differentiation – customers perceiving the brand as meaningfully different – and distinctiveness – customers rapidly recognising which brand they are seeing.

The evidence discussed in the book reveals surprisingly low levels of perceived uniqueness for many highly successful brands. This does not prevent people from buying them and displaying moderate loyalty towards them.

Distinctive assets

Colour. Shape. Logo. Symbol. Sound. Character. Typography. Tone of voice. Visual composition. Anything capable of quickly saying “this is that brand” without requiring someone to read the brand name carefully.

The value of such an asset depends primarily on two things: how many people associate it with the brand and how uniquely they associate it with that particular brand.

These associations are not built during a single campaign. They require years of consistency. If colours, visual style, message, composition and tone are changed every year, the company repeatedly pays to rebuild recognition that it has itself destroyed.

If advertising worked solely as an immediate salesperson, we would expect sales to jump as soon as advertising starts and collapse as soon as it stops. For established brands, such a clean pattern is often absent.

Sharp proposes two important explanations. First, advertising often protects future sales that would otherwise be lost. Second, advertising effects are spread over time. Somebody may see an advertisement today but not encounter the relevant buying situation for weeks or months.

Memory is the bridge

Between advertising exposure and purchase lies memory. Advertising therefore needs to succeed at least at two things: getting noticed and being correctly linked to the advertised brand.

A beautiful commercial that everyone remembers but half the audience attributes to a competitor can be excellent creative work and a terrible investment.

Reach versus frequency

One of the book’s central recommendations is to pursue broad reach over time rather than excessive repetition to a small group of people. Advertising needs to reach the millions of occasional buyers who rarely think about the brand.

Continuous presence is also more compatible with the way memory decays than short, intense bursts followed by long periods of silence.

A price promotion has one enormous psychological advantage for managers: the result is visible. Reduce the price, sales jump and the graph looks impressive.

The problem is that the effect usually ends when the promotion ends. A large proportion of buyers purchasing on promotion already know and buy the brand. Afterwards, they generally return to their normal buying behaviour.

Revenue is not profit

This is where accounting language becomes brutally useful. A discount does not merely reduce the selling price. It reduces the contribution margin on every unit sold – including units customers would have purchased without the discount.

Imagine a product with a selling price of 100 and variable costs of 70. Contribution is 30. Reduce the selling price by 10%, to 90. Contribution falls to only 20.

To preserve the same total contribution, you now need to sell 30 ÷ 20 = 1.5 times as many units – in other words, 50% more volume.

And that is before adding the cost of the promotion itself, promotional materials, commissions, logistics, additional servicing and any competitive response.

None of this means “never run promotions”. They may have logistical, commercial or distribution objectives. The correct question is: what economic problem are we solving, and at what cost?

The theory behind loyalty programs is extremely appealing: reward customers for repeat purchases and they will buy more frequently, stay longer and avoid competitors.

The empirical studies discussed in the book find an effect, but generally a weak one. The reason is almost built into the mechanism itself: loyalty programs most easily reach people who already buy regularly and already have the strongest incentive to join.

The company therefore risks rewarding behaviour that would have occurred without the reward.

Another statistical trap

If loyalty-program members are compared with non-members, the members will probably look more loyal. But that does not prove the program made them more loyal. More loyal customers may simply have had a stronger incentive to join.

This does not make customer databases useless. They can have substantial analytical value. A large database and a powerful growth mechanism are simply not the same thing.

After challenging loyalty strategies, segmentation, differentiation and price promotions, the book arrives at its own positive theory of competition.

Mental availability

Mental availability is the probability that a brand will be noticed or come to mind in a particular buying situation. It is more than simple awareness. The question is whether the right links in memory are activated at the right moment.

Ninety percent of people may know that you exist and yet almost nobody thinks of you when the need actually arises.

Memory works as a network. A brand can be associated with a place, situation, need, colour, product, person, time of day, occasion or experience. The richer, more relevant and fresher those associations are, the greater the number of opportunities for the brand to come to mind.

Physical availability

Being “available” does not mean that a highly motivated customer can find you after twenty minutes of searching. It means being easy to buy: in the right places, at the right times, in suitable formats, through convenient processes, with adequate distribution and minimal friction.

For services, this can mean being easy to find in search, having a functioning website, a clear offer, easy appointment booking, convenient payment methods, suitable opening hours and appropriate geographical coverage. For physical goods, it can mean channels, stock levels, sizes, packaging, shelf presence and distribution.

Market-based assets

Sharp treats mental and physical availability as assets that take time to build and can sustain sales for long periods. This is why an old, familiar and widely distributed brand can remain enormously valuable even when its current marketing is mediocre.

The seven rules