Marketing as a corporate function originated out of sales more than a hundred years ago. Rather than simply trying to sell what the company manufactures (the sales concept), the marketing idea was that it made more sense to start by first examining customer needs and then producing what consumers actually need. As Peter Drucker argued: The aim of Marketing is to make Selling superfluous by understanding the customer so well the product or service fits him/her and sells itself.

As companies began examining customer needs closely, they quickly realise that customers differ in their preferences. This gave rise to the concept of market segmentation, which is about dividing the market into homogeneous groups of customers who respond similarly to the marketing mix. The marketing mix, or the 4 Ps of product, price, place and promotion, are the essential tools available to a marketer to position the company’s offer to the segment that is targeted for sales. In my experience, solving any marketing problem ultimately requires a deep understanding of market segments.

Having observed marketing practice for three decades, I am frustrated that the fundamental concept of segmentation is so poorly understood and implemented in companies. Rather than thinking deeply about customer needs and the resulting segments, marketers slip too easily into defining segments based on demographic variables such as gender, age, education and income. I suspect this happens because it is so easy, through demographics, to identify who falls into which segment: males versus females, for instance, or singles versus married.

The problem with demographic segments is that they rarely reflect differences in customer needs. For product categories such as toys, apparel and cosmetics, demographic segments can undeniably be a valuable starting point. But even here it is rarely enough. One needs to further subsegment the chosen demographic target based on "needs" in order to develop actionable strategy or marketing recommendations.

Segmenting Porsche Buyers

Consider an exercise I often ran while teaching. I would ask my class, “Who buys a Porsche in the United States?” The immediate response was: male, 40-plus, college educated and earning more than $200,000 annually. Yes, it is true that a large proportion of people who buy a Porsche fall into this group. Yet it does not tell us why they buy one: what is the need they are trying to fulfil with a Porsche?

When pushed the class typically argued for the profile we call "top guns" – driven, ambitious types for whom power and control matter. They buy a Porsche to be noticed. But these top guns account for only a quarter of Porsche buyers. There are four other segments that buy Porsche with varying motivations.

Implications for Marketers

1. Marketing versus sales function?
The role of group marketing is to support the sales function with better targeting so that it increases conversion rates. Clearly, except by chance, no dealer or salesperson for Porsche could develop this segmentation scheme. It is the central marketing team of the brand that has this capability and responsibility. Beyond segmentation, marketing must generate the ideal scripts to converse with each segment and, subsequently, roll it out across the dealer network. This will ensure that dealer salespeople do not treat every customer conversation as if they are interacting with a top gun, as all that would achieve is turning off potential buyers from the other four segments.

2. How is the segment to be identified?
Of course, the problem remains that even after a salesperson is trained on these five segments that exist and the ideal scripts are prepared, how is she to know which segment the customer in front of her falls into? The identification of the segment is essential for differential conversations to happen. At the outset we must acknowledge that identification will always be a challenge with needs-based segmentation.While perfection is the enemy of any action, there are methods that get us closer to solving the identification problem. For example, one may profile the segments to see if they differ in some demographic manner. There may be other clues, such as dress or mannerisms which can help identify which segment an individual belongs to. Finally, one may develop a set of questions – fewer the better – which help in discriminating between segments. To keep improving, if a customer database is maintained it should have a column for segmentation that is updated by the salesperson. This ensures more appropriate follow-ups. None of these approaches are ideal, but together they improve the hit rate for identification. Over time, it is to be hoped, companies learn and get better at it. It is such improvement that results in a competitive advantage for the enterprise.

3.Modify product or modify other elements of marketing?
It is important to understand when segmentation will drive the development of unique products for each segment versus when changes will be restricted to the other elements of the marketing mix. In the Porsche example, one would not have singular products for each segment. Instead, it is the sales approach and advertising – perhaps the segments consume different media and, as a result, advertising copies can be tactically placed – that will differ across the segments.

In conclusion, if there is one thing that I have learnt examining marketing problems over the years is that they ultimately boil down to a segmentation problem. Perhaps the segments have not been defined appropriately? Or, one is not targeting the right segment? May be the marketing mix is not aligned to the target segment?

As we say, “who targets best, wins” but, to do that the foundation is an effective segmentation scheme.

Nirmalya Kumar is Member-Group Executive Council at Tata Sons and Visiting Professor of Marketing at London Business School. His Twitter handle is @ProfKumar. This article is written in his personal capacity.