Consumer behaviour, be it product purchase or media consumption, is a matter of great interest to marketers. Tomes and tomes have been written on this subject. Consumer behaviour of segments within the mass may be as different from each other as chalk and cheese. However, the nuances may be hidden from us if we continue to examine the data at an aggregate level. The averaging out effect that happens when we look at the aggregated numbers hides this sharp differentiation from us.
Vive la difference!
Mass marketing looks at the market as a homogenous group and offers the same marketing mix to all customers. This method allows economies of scale to be leveraged however an undifferentiated offering is unlikely to satisfy all customers. Competitors with products that meet specific consumer needs would eat into the mass marketer’s brand franchise.
With brand proliferation and a bitter share battle being fought in the market place, the trick lies in being relevant to the consumer and in being able to deliver a value proposition that can’t be replaced. Differentiation, practicing the art of Niche marketing, Narrowcasting or Consumer Segmentation is therefore the key. Marketers are moving from an ‘All things to All people’ strategy to an ‘All things to Some people’ or a ‘Some things to Some people’ strategy.
Segmentation studies typically work on breaking up the base TG into segments which are distinct from each other. A good segmentation ensures that there is maximum difference between segments (heterogeneous) and minimum difference within the segment (homogeneous). The following criteria need to be kept in mind while defining segments:
Measurable – the variables used to differentiate the segments should be easily measurable
Viable – the segments should be large enough to justify returns on the investments required to target them
Accessible – the segments must be reachable through communication and distribution channels
Types of Segmentation
The basis of the differentiation decides the kind of segmentation.
Region – by country, state, district, town etc.
Population density – often classified as urban, semi-urban, rural
Population strata – according to size of population : 10 lakh+ towns, 5-10 lakhs, 1-5 lakhs, below 1 lakh
Social class SEC
Psychographic segmentation groups customers according to their lifestyle. It considers a number of influences on buying behaviour – attitudes, expectations and activities of consumers etc. The main types of psychographic segmentation are:
Lifestages – Life-stage is the life-cycle that each household passes through where the priorities, expenditure patterns and media consumption undergo a significant transformation. If variables like age, marital status, children at home, age of children etc. are used to segment, it would give rise to groups like Yuppies (Young upwardly mobile), DINKS (double income no kids), Family with toddlers, Family with Teens, Empty Nesters (Kids grown up and have left home) etc.
Product categories like Education, Healthcare, Financial services need to be Life-stage targeted and plain vanilla demographic TG definitions do not suffice to target the TG too well. Let’s look at how Life-stages would impact targeting for financial services. The Unmarried Independent has just started earning and since he has no dependents and peer group and friends are important, savings are limited and the focus is to spend on hanging out with his friends. Housing loans and FDs are the financials instruments for the Married with no dependents. Insurance schemes and mutual funds linked to education are what would interest the Married with young kids. With an increase in income levels and corresponding increase in the risk propensity, the Married with Teens and the Married with independent kids are more willing to look at shares and equity linked mutual funds and of course retirement plans. Empty Nesters are rarely the target for any financial instruments.
Attitudes, Interests and Opinions – Apart from denoting individual dimensions of personality, psychographics describes lifestyle of the consumer. Typically the analysis centers on the consumer’s Activities, Interests and Opinions. Activities give a sense of how the consumer spends his leisure and work time, interests in terms of how they relate to their immediate surroundings and Opinions in terms of their stance on social issues etc.
If the clustering variables are more lifestyle variables like openness to change, frequency of going out, comfort with technology, save v/s spend pattern etc., we would arrive at segments which were more attitudinal like Strivers, Fun-seekers, Home-bodies, Conservatives etc.
With the same demographic profile, we could be speaking to two very different people. Let us look at a case of an FMCG company having 2 products – fortified packaged atta and the other product, ‘heat and eat rotis or chapattis’. Demographically both these products would be talking to the same TG – Female, 25+, SEC AB. A psychographic segmentation of the demographic TG shows that the ‘Progressive home-maker’ would be prepared to pay a premium for a value-added product, especially if it impacts the health of her family positively. However she is not in the market for convenience foods and ready-to-eat products as these products are seen to be a short-cut compromise which she is not willing to make. She would obviously be the right TG for our fortified atta. The ‘Career-Conscious woman’ on the other hand is secure about her place and worth in the family and does not define herself basis the food she puts on the table. She is therefore more than willing to buy our heat and eat rotis thereby spending less time in the kitchen and more with her family!
This kind of segmentation is based on actual customer behavior toward products. If a brand has to grow its market share, the task would be achieved more efficiently and effectively by a deeper study of the competitor’s brand franchise. Competitor’s brand loyal consumers would be that much more difficult to convert and we would be better off targeting consumers who are non-regular users. Or trying to increase consumption with our light user base. Or reaching out to the non-users of the category ! The entire marketing and communication focus would be quite different for each of these initiatives. Media choices too, would be focused on specific selected segments leading to a more targeted plan.
Degree of loyalty – customers who buy one brand either all or most of the time are valuable to firms. By segmenting markets in this way, firms can adapt their marketing in order to retain loyal customers, rather than having to focus constantly on recruiting new customers. It is often said that it is ten times more profitable selling to existing customers than trying to find new ones !
Occasions – this segments on the basis of when a product is purchased or consumed. For example, some consumers may only purchase flowers, wine or boxes of chocolates for celebrating birthdays or Christmas, whereas other consumers may buy these products on a weekly basis. Marketers often try to change customer perception of the best time to consumer a product by promoting alternative uses for a product. For example, recently Kellogg’s has attempted to change the image of cereals to that of an ‘any time’ snack, rather than simply a breakfast meal.
Benefits sought – this requires marketers to identify and understand the main benefits consumers look for in a product. From mass brand like Lux and Lifebuoy, we now have Dove, a moisturizing soap for the older woman, a fairness soap, an anti-acne soap, an anti-bacterial soap and even soap-free face washes. Even a herbal soap like Medimix is now available in 4 variants.
Usage – intensity of usage is another parameter of segmentation – heavy, medium and light users
User, Buyer and Decision-maker – In some cases, the user may not necessarily be buying the product. The Housewife may purchase soaps for the entire household but the users within the household may have influenced the choice – the husband may want a Brut soap, the mother-in-law may want Medimix, the teenage daughter - Clearasil and she may buy Lux for herself. It is important to decide the target audience fairly clearly – the buyer or the user
The buying process for a lot of high value purchases like durables or cars is fairly complex with possibly multiple influencers. The whole family may get involved in the purchase of the much desired plasma screen TV. Some car manufacturers even target kids in the home as a ‘pester power’ influencer in the final car purchase decision!
Psychographic segmentations have a unique problem when examined from the media context. Since the base has been segmented basis psychographic variables, selected ‘hot prospect’ segments may actually have very similar media usage patterns when compared to the base TG and therefore may not result in any significant insights while fine-tuning the media plan.
In media segmentations, the segmentation or clustering variables are media usage patterns v/s the values, attitudes and lifestyle variables used in psychographic segmentation. Readership of Dailies, language of dailies read, magazine genres, channel normally watched, radio stations listened to, internet usage, Language of movies watched etc. would typically be the variables used.
Since the segmentation is done basis media variables, each segment has distinct and well differentiated media habits. Media segmentations are therefore able to overcome the shortcomings of psychographic segmentations as far as fine-tuning media choices is concerned.
Viewing baskets of TV viewers – Media segmentations done on the TV viewing data yield some interesting insights. The industry data is restrictive when it comes to highlighting cross program viewership and the existence of different viewer types. Segmentation studies can actually help overcome this issue. This is invaluable information from a programming perspective for a TV channel. A channel can now plot its program portfolio across the different segments, identify the gaps and the segments where buy-in to its franchise is not happening, analyze a segment mix to decide on ideal programming to build viewership etc.
Intensity based segmentation – Heavy, medium and Light viewer classification helps brands build a more efficient reach construct. GECs may actually be over-delivering on the heavy viewer and under-delivering on the light – the dangers of the averaging effect ! Alternate channel choices would help build reach evenly across all segments
Databases like TGI which collect psychographic data on values, attitudes and lifestyles would be best suited for psychographic segmentation. As TGI also collects detailed media and product consumption data, it is possible to connect the chosen segments to media choices for specific categories or brands.
Surveys like the NRS, IRS and TGI which cover media usage across media like Print, TV, Radio, Cinema and Internet would provide a good base for media segmentation studies with the added advantage of being able to link to Product and brand usage. TAM and aMap would provide a suitable base for TV audience segmentation though having a drawback of not being able to link to brand usage.
Usage & Attitude studies or Household Panels provide sound bases to segment the consumer on product purchase and usage dimensions. While retail offtake data has a drawback of not being able to connect to the last mile or the end-user, it is invaluable information for Geographic segmentation.
Segmentation studies which are based multi-dimensional variables like psychographic and media segmentation, use a research technique called Cluster Analysis. Cluster analysis looks at separating a demographic base into sub-segments where differences or heterogeneity is maximum between the sub-segments and minimal within the segment. Respondents are randomly assigned to a pre-decided number of groups to begin with (k-means clustering where k is the number of groups). They are then reallocated iteratively in such a way as to maximize homogeneity (sameness) within groups and to maximize heterogeneity (differences) across groups. While this process sounds sufficiently complicated, there are various research tools like SPSS which do all the complicated calculations at the click of a button! The only task left to do is to then interpret the results.
Care must be taken to ensure that we do not slice and dice the base so fine that we are left with meaningless information on sample sizes that are too small to lead to any valid analysis. This holds true for any kind of segmentation.
In the hugely complex and complicated marketplace, we no longer have a choice – it’s either differentiate or sink. Segmentation studies are invaluable as they help us peel another layer of the consumer onion and help us understand the drivers of consumer behaviour.
(Arpita Menon is Managing Partner, Quantemplate. Quantemplate is a media analytics company focused on maximising realisation for media owners.)