SensoGram 28: Sensographics   We are familiar with demographics (splitting of the population according to age, sex, income etc).  More recently, psychographics has become important to advertisers and marketers (splitting the population on attitudinal or personality variables).  And now we have sensographics - splitting the population according to sensory preferences. Sensographic segmentation is not new; it has always been part of life (Aone man=s meat is another man=s poison@).  If we take a single food category - cheese, for example - we see that it can vary enormously in appearance, flavour and texture; from  soft, bland cream cheese to hard, strongly-flavoured Parmesan.  Individuals will differ in their liking for these products.  This is not surprising, for the products themselves vary tremendously.  But if we narrowed our focus to cheddar cheese only, would we still find that individuals vary in their likes and dislikes?  Or could we find a single cheddar cheese of maximal acceptability to everyone?  Similarly, would we find sensory segmentation with such apparently homogeneous products as white bread, tomato sauce, margarine, or vanilla ice cream? SensoMetrics experience suggests that this must be established empirically (typically it requires a large number of consumers tasting a large number of products).   Sometimes the study can reveal striking segmentation - segments, interestingly, that were not at all obvious a priori.  In other cases there is no segmentation at all, with consumers agreeing on what are the best and worst products within a category.  What are the implications for food marketing?  Let=s take the easy case first.  If there is no sensory segmentation the issue is straightforward:  there will be a consensual bliss point and one optimised product formulation will satisfy the whole market.  (This is perhaps not as easy as it sounds, but at least the task is well defined.)  When there is segmentation, the task is more difficult.  Suppose we find two segments.  This will involve the optimisation of two variants, one for each segment.  While this is a larger responsibility for product development, it is an even more onerous one for marketing.  The catch is that, while these two segments are real, consumers may not know to which they belong!  With a tomato sauce, for example, you may actually like a thin, more acid formulation while I like a thick, sweeter product ... but is each of us aware of this?  It is possible to go through life being aware of what you like and dislike but not being aware of (or even interested in) which descriptors match your preferences.  Marketing communication thus becomes fraught: launching the two variants with appropriate descriptors won’t help unless consumers can match descriptors to their preferences. The only way forward is to encourage trial of the products.  If each variant is truly matched to the preference of each segment, it should be obvious to an individual which one he/she prefers and, hopefully, will continue to purchase.  There are, however, other complications.  If family members fall into different segments, will the grocery buyer be willing to buy two variants of a product?  And what are the implications for product cannibalisation?  Sensographics is in its infancy and not without challenges.