Life Cycle

[Please send correspondence, and direct telephone calls, to my home address in the first instance]

The Product Life Cycle (PLC) has represented a central element of marketing theory for four decades. Following its development in the 1950s, and its subsequent popularisation in the 1960s, it has remained a stable feature of marketing teaching; despite evidence of its limited applicability and the growing awareness, amongst leading academics at least, of its flawed nature.

This paper critically re-examines its practical value in the light of the twenty year history of 929 brands spread across 150 FMCG markets in the UK. The main conclusion is that the 'life cycle' of the brand leaders is indeed more stable, and much longer, than some of the previous work might have suggested; and that the PLC would have predicted. As a credible aid to marketing decisions the brand life cycle is, therefore, of little value in the majority of FMCG markets - and cannot justify its long held position as a theory which has general, practical applicability across the whole field of marketing.

The Product Life Cycle (PLC) is one of the most quoted, and most frequently taught, elements of marketing theory. Its influence has also been seen in other theories, from new product management to portfolio analysis.

The theory has been has been subject to relatively little public criticism, with only 20% of 271 papers published on the subject between 1971 and 1991 undertaking further research into the subject and only a handful challenging its basic assumptions. On the other hand, proof of the concept has appeared to be surprisingly difficult to find.

Although the PLC concept was first developed from the 1950s (Dean, 1950[1]) , with a number of PhD theses published at the end of that decade, it was in the late 1960s, following the influential article by Theodore Levitt (1965[2]), that much of the definitive work was undertaken. For example, William E Cox Jr (1967[3]) examined 754 ethical drug introductions and found several forms of life cycle, though he was only able to track 2% of these to their 'death'. Two years later Polli and Cook (1969[4]) conducted their widely quoted survey of 76 selected product forms and 37 product classes in the fields of health and personal care, food and cigarettes. Much of the work was though anecdotal, or concentrated on a very limited number of examples.

It was also dogged by problems of definitions. Perhaps the clearest are provided by Kotler (1991[5]) who describes a hierarchy starting with the longest lived cycle, that of the 'demand/technology' (market) life cycle, which contains a succession of 'product form' (product class) life cycles, each of which in turn contains a set of brands with their own 'brand life cycles'. This paper concentrates on this lowest level since this should, as the shortest lived, show the most marked changes over time.

The article by Dhalla and Yuspeh (1976[6]) was one of the few to offer a counter viewpoint; but was, again, justified by a limited number of examples. Even so, they made the important observation that "...clearly, the PLC is a dependent variable which is determined by market actions; it is not an independent variable to which companies should adapt their marketing programs. Marketing management itself can alter the shape and duration of a brand's life cycle." Despite its relative isolation, in some quarters this article seemingly had an influence out of all proportion to its limited objectives; since it is the one most often quoted as the basis for the more recent doubts about the general applicability of PLC theory.

Despite these doubts, product life cycle has become an accepted, and valued, element of basic marketing theory in general - and beyond that it has become one of the building blocks of management theory as a whole. It is now an obligatory part of the staple diet provided by most business schools, and especially of the content of most marketing textbooks. Philip Kotler (1991[7]), for instance, devotes 24 pages to the subject, but includes less than one page on its limitations (and then largely in terms of the false detection of the decline phase, as described by Dhalla and Yuspeh). Other authors of best-selling marketing textbooks give the subject a similar treatment. Thus, a survey of 15 current student textbooks shows that an average of 9 pages (from an overall total per book of 650 pages) is devoted to description of the PLC; with an average of less than half a page given over to its limitations (again usually only in terms of the practical possibilities of extending product life).

It would appear from the above evidence, taken from the literature searches (including 271 papers on the subject published between 1971 and 1991), that the PLC is still a dominant element of marketing theory. The latter part of the paper, therefore, starts by adding to the volume of evidence that the PLC has only limited applicability. But, in a more constructive vein, it then also goes on to look at some of the other lessons which may be derived from the leading brands, in UK FMCG markets.

The data used in this survey was derived from that collected by the British Market Research Bureau (BMRB) as part of their Target Group Index (TGI) survey. Specifically, it compares the results of the 1989 survey with those of the comparable survey in 1969 - thus giving a valid comparison over two decades. The survey techniques used by TGI are well known, and have been validated over many years (O'Brien & Ford, 1988[8]), so these will not be reported in detail here. The essence, in terms of this research, is the size of the TGI sample (which covered 25,000 British adults in 1989) - allowing a large number of brands to be tracked with some degree of accuracy and comparability of data (both in terms of questions asked and sample selection) over the two decades.

From within a range of 230 FMCG markets and market segments reported on over the twenty year period by TGI, 150 were chosen as being suitable for comparison. Thus, more than 60% of the overall market/segments recorded by TGI were included in the research. Fifty four other markets/segments were rejected due to changes in definition between the two surveys which meant that their results could not be directly compared. Eleven markets/segments were dropped over the period by TGI and 25 new ones were added; and these too could not be used for comparison. There was no obvious evidence that the other markets/segments not chosen displayed significantly different characteristics to those chosen.

As the original population of markets/segments was clearly non-random (having been chosen by MRB on a 'convenience' basis, as being those of interest to the key buyers of research reports) no attempt was made to treat the selected 150 markets/segments as a statistically viable sample. These markets did, however, cover a wide range of products - from instant coffee and newspapers to anti-freeze and indigestion remedies - so their scope was significantly wider than any of the earlier surveys reported above, and the results may reasonably be considered to provide more generally applicable lessons.

Within the 150 markets/segments chosen 929 brands were tracked - these being those with the highest, 'most used', penetration (typically as first, second or third brand in 1969 and/or 1989) within these markets/segments.

To minimise the problems caused by comparison of absolute measures across disparate markets/segments, in the main only relative measurements of brand positions within these markets/segments and/or relative changes between 1969 and 1989 were considered. In practice, the most useful measure was found to be that of the rank order of brand 'penetration'.

It should be noted that the 'penetration/share' measure used was that of penetration as the brand 'most often' used or equivalent. This represented only an indirect indication of the traditional (volume based) brand share. It did, however, allow a comparison over a particularly long time period without any problems caused by shifts in retail distribution patterns or methods of sampling and recording techniques. It is likely that this ('most often used') penetration is indicative, at least of rank order, of changes in brand shares - where the typical brand leader recorded twice the level of the second brand and three times that of the third.

Consolidation of the statistics across the 150 markets/segments was mainly on the basis of simple counts/averages. No attempt was made to weight the results to allow for different sizes of markets/segments or brands.

The most immediately obvious result from this analysis is the dominance of the brand leaders. The first three brands on (unweighted) average took 82% of the 'most used' penetration in 1969 and 72% in 1989. The first brands by themselves took 46% in 1969 and 39% in 1989.

The average ratio of first:second:third brands was 10:5:3. These averages were quite stable, the 1989 average ratios being within 5%, of those recorded in 1969. They compare well with the 4:2:1 ratio suggested by the Boston Consulting Group (Henderson 1985[9]); though it should be noted that the latter were for volume share based figures.

Perhaps the most significant outcome, however, was the length of brand life. The majority (53%) of the 150 brand leaders in 1969 remained the brand leaders in their respective markets/segments in 1989.

At the other extreme, a mere 7% had declined below fourth place and only 1% had been discontinued. Where these brands account for such a large proportion of overall sales, this absence of evidence of the end stage of the life cycle by itself undermines the assumption that the PLC has general applicability.

The changes in position of the leading brands over the two decades are shown in Table 1.

The second placed brands were marginally less stable, but still with 39% overall either holding position (22%) or increasing to first place (17%). Only 22% moved to below fourth place and just 7% of even these second brands disappeared from the record. Third place was marginally less stable again, with 34% holding steady or growing (7% to first place and 14% to second), but 40% falling below fourth place (and 18% disappearing completely).

Thus, of the 451 brands holding the first three positions in 1969 (and accounting for 82% of 'most used' penetration) 91% were still alive two decades later - and 68% remained in fourth place or better! This again contradicts the popular predictions of PLC theory - and demonstrates the lack of general applicability for the theory.

Even though there were relatively few brands involved there is some anecdotal evidence, judged subjectively, of factors which might be observed during the introduction and growth stages of the life cycle. Almost all of the 19 totally new brands reaching the position of brand leader by 1989 appeared to incorporate a significant degree of new technology; for example Duracell offering longer-life batteries or Gillette Aerosol Shaving Foam replacing cream products. Some of this was also, though, in the form of new formulations to match changed consumer tastes (with Heinz slimming soups replacing biscuit products) or even of image (with the more exotic L'Oreal replacing the relatively mundane Amami as a hair setting product).

This predominance of 'technological' or 'formulation' change (even if only as perceived by the consumer) was less evident in those 39 new products only achieving second place (with only 67% of these displaying obvious 'product advantages') and even less in those 52 only able to reach third place (44%).

A reflection - perhaps - of the value of the brand name, and/or of the (resulting) domination by the well-resourced large corporations, is the fact that no less than 99 of the brands appeared in two or more markets/segments; between them providing almost half of the total number of products in the leading (first to third) positions.

Three quarters (79%) of these 'extended' brand names only occurred, however, in two or three markets/segments - usually very closely related ones. The power of brand names was also reflected in the fact that existing, 'extended', brands took first place in 60% of the new markets/segments added between 1969 and 1989.

Not unexpectedly, one feature to be observed gver the twenty years was the emergence of own brands as strong 'multiple brands' in their own right. The eight retail organisations featuring in this way won 142 places in the list of 1989 leading brands (15% of the overall numbers monitored). More important, the leading retailers in each of the two major sectors (Sainsbury in supermarkets and Boots in drugstores) accounted for 86 of these products (9% of the overall); providing 14% of the second brands and 22% of the third brands (though only 3% of the first brands) in the 150 markets/segments monitored in 1989.

In terms of a test for brand stability, the simplest analysis of all - the simple mean for each of the groups being investigated, produced as much information as the more sophisticated analyses; as well as giving sets of figures which were easier to comprehend. The resulting detailed analysis of the main factors which appeared to most differentiate performance is shown in Table 2.

As is shown by from Table 2, the 'market' growth rate (defined, of course, in this context as penetration overall for any brand) appears - at least at first sight - to be the best predictor of changes within a market; as is also implied by the theory behind the use of the Boston Matrix. Thus, the 53% of markets where the brand leaders remained at the number one position over the two decades recorded an overall 1989 penetration which was on average 98% of that of 1969.

At the other end of the spectrum, those new brands which achieved first place in 1989 did so in markets whose penetration on average grew by 35% over the two decades.

On the other hand, even this growth rate represents only 1.5% per annum cumulative over the two decades; well below the 10% level which has been traditionally considered significant in Boston Matrix analyses.

More surprising, perhaps, is the fact that a similar pattern, of market growth higher than normal, also emerges for brands which disappeared from the record (the eleven brands disappearing in this way after being in second place in 1969 were in markets/segments with an average growth over the two decades of 31%). An alternative explanation, therefore, might simply be that as a result of the increased competitive activity in these markets/segments, where a battle for leadership was taking place, an increase in overall penetration was created.

In Table 2, again, a slight relationship may be observed between performance, in terms of rankings, and the ratio of the penetration of the second brand compared with that of the first. In other words, as might be reasonably expected, the better the advantage the brand leader held over its rivals the more stable its position was likely to be. Thus, the average of this ratio in markets where brand leaders moved from first to second or third position were respectively 0.62 and 0.8. This compared with the average for markets where the brand leader remained unchanged of 0.45.

A similar effect was observed in relation to the ratio of first to third brands, where this ratio for the stable brand leaders was on average 0.25; compared with 0.42 for those moving down to second position and 0.56 for those falling to third place. This was also reflected in the distribution of the results, where only 37% of markets/segments with the changes in brand leadership showed a third/first brand penetration ratio of less than 0.5 compared with 79% of those in markets/segments with unchanged brand leaders.

This finding, that brand leaders with a more dominant position are more stable is not, perhaps, unexpected. The Boston Matrix implies as much - though, once more, the dividing line (traditionally drawn at 1:1) appears rather different in this case (at 2:1). As mentioned earlier, though, this is in line with other work from the Boston Consulting Group (Henderson 1985[10]).

The data for this research has its limitations, consequently restricting the range of analyses which can be utilised. It also is specific to one country (the United Kingdom) and one 'industry' (FMCG). Nevertheless, the findings clearly support the view that the product life cycle is limited in its applicability. It is a tautology that products are created and later die. What is relevant in the context of marketing is whether this is of practical use to the marketing manager. The fact that this research indicates that the average length of life, for the brands which represent the major share of volume in the wide range of FMCG markets investigated, considerably exceeds two decades means that in practice this theory may bear little relation to the pragmatic needs of the marketing manager whose time horizon is more likely to be a maximum of two years.

A more difficult question is that which looks to the difference between brands and products (the theory is, after all, carefully labelled product life cycle theory). But, if we follow Kotler's (1991[11]) hierarchy, we would expect product form life cycles to be even longer still - so the same practical considerations would apply. The picture is often confused by the introduction of what may be best described as product feature cycles; and then often even more confused by those in new product development who refer to these planned changes/upgrades in features as 'product life cycles'. These may be relevant in some markets, such as computers, where the technological developments force rapid changes in such product features; though even there the brand which contains these transient elements (say IBM), and often still is the dominant element of the overall product package, tends to be very long lasting - and the true product (a mainframe computer, say) even longer lasting still. In any case, in many markets, not least the wide range of consumer markets, the brand is the most significant and identifiable element of the product or service package. Only in a few consumer related markets, such as ethical pharmaceuticals, is the brand tied to a physical product in such a way that the changes in formulation determine its life - though it is interesting to note that it is just these markets which have preoccupied a number of the researchers in the field of the PLC.

The fact that more than half (56%) of the leading brands (first/second/third) in 1969 remained amongst the leaders (first/second/third) two decades later in 1989, in markets which are typical of Western developed countries and are supposedly the most susceptible to the use of the PLC, - and no less than 83% of the 1969 brand leaders remained amongst the first three brands in 1989 - clearly shows that the theory has little practical value in most FMCG markets in the UK; and this important limitation is also likely to apply in many other markets. It significantly reinforces the criticisms voiced by Dhalla and Yuspeh others. It must, therefore, be argued that this (PLC) theory cannot any longer be justified in terms of general applicability - and should be relegated to usage (albeit valuable usage) in special circumstances.

In terms of predictors as to what, rather than the PLC, will forewarn of future problems, there are several to choose from. Paradoxically, a growing market (and probably one growing as a result of competitive activity) may lead to a change in leadership. Brand leaders with less than a two to one dominance over the second brand would be well advised to frequently look over their shoulder. All brands should beware of genuine technological changes (or at least those perceived significant by customers).

On the other hand, it is fair to note that, despite the demonstrable lack of general applicability for the theory as whole, the major lesson of the PLC - that change is to be ignored at the marketing manager's peril - still holds true. The specifications of few of the brands looked the exactly same in 1989 as they had in 1969.

Acknowledgments: The author gratefully acknowledges the support of the British Market Research Bureau in making available the invaluable historical data which forms the basis of this research.

Table 1 Movements in the Rankings of Leading Brands over 20 Years in 150 UK FMCG Markets

[1] Dean, Joel (1950) "Pricing Policies for New Products", Harvard Business Review (November-December) 45-53

[2] Levitt, Theodore (1965) "Exploit the Product Life Cycle" Harvard Business Review (November-December)

[3] Cox, William E. Jr. (1967), "Product Life Cycles as Marketing Models", The Journal of Business, 40(4), 375-384.

[4] Polli, Rolando and Victor Cook (1969), "Validity of the Product Life Cycle", The Journal of Business, 42(4), 385-400.

5Kotler, Philip (1991) Marketing Management (Seventh Edition), Prentice-Hall, Inc., Englewood Cliffs, New Jersey

[6] Dhalla, Nariman K. and Sonia Yuspeh (1976), "Forget the Product Life Cycle Concept!", Harvard Business Review, (January-February), 102-110.

[7] Kotler, Philip (1991), Marketing Management (Seventh Edition), Prentice-Hall, Inc., Englewood Cliffs, New Jersey

[8] O'Brien, Sarah and Rosemary Ford (1988), "Can We At Last Say Goodbye to Social Class", Paper presented to the 31st Annual Conference of the Market Research Society

[9] Henderson, Bruce D. (1985), "The Rule of Three and Four", Boston Consulting Group, Boston.

[10] Henderson, Bruce D. (1985), "The Rule of Three and Four", Boston Consulting Group, Boston.

[11] Kotler, Philip (1991) Marketing Management (Seventh Edition), Prentice-Hall, Inc., Englewood Cliffs, New Jersey