Product Variety, Life Cycles, and Rate of Innovation – Trends in the UK Automotive Industry Matthias Holweg Lean Enterprise Research Centre Cardiff Business School, Wales, UK Tel: +44 29 20 87 45 44, Fax +44 29 20 87 45 56, Email:
[email protected] Anthony Greenwood Cap Gemini Ernest & Young London, UK Email:
[email protected] Abstract This contribution aims at predicting future trends in the automotive industry in terms of product variety, model range and product life cycles in the UK automotive market. The analysis is based on empirical data dating back to 1960, which is extrapolated into future trends. The findings indicate no consistent trend in regards to product variety, which seems to be merely driven by vehicle manufacturer’s policy, but could identify a clear trend towards shorter product life cycles and increased model ranges offered in the UK market. These trends in consequence will force vehicle manufacturers increasingly to design and introduce vehicles that are far less dependent on economies of scale, as the overall volume per model is predicted to decrease drastically. Current efforts of the vehicle manufacturers to meet these requirements, the product platform strategy for example, could also be identified. Introduction The automotive industry, heavily constrained by its worldwide production over-capacity, not only faces fierce competition in all its major markets, yet has to deal with a customer who is proven to be more and more demanding (ICDP, 1998), both in terms of choice as well as model range. It is often claimed that both competition in the marketplace and customer expectations have led to an ‘infinite variety’ offered by the manufacturers and significantly shortened product life cycles which then in return provide even shorter ‘market time windows’ for the vehicle manufacturer to achieve the sales volumes necessary to support the economies-of-scale induced by the enormous development costs of a new vehicle. However, neither in terms of product variety, i.e. the specification derivatives of a model, nor the model range, i.e. the total number of different models offered, nor in terms of the development of product life cycles over time, is comprehensive primary data available. Hence this paper sets out to analyse the product variety, model range and life cycles of six major vehicle manufacturers offered in the UK automotive market since 1960. The objective of the research is to establish trends, extrapolated from historic data, to predict the scenarios and market characteristics vehicle manufacturers (VMs) will have to meet in the future. Life cycle and model range requirements dictated by the marketplace are key inputs for the strategic model line-up policy, and product variety is a key factor in determining future distribution strategies of the vehicle manufacturers.
Also, the effect of the new plate change system, which was a major cause of distortion in the UK auto market, will be investigated. The research questions in detail are: 1. What product variety has been offered by volume manufacturers in the UK since 1960, and what trends can be observed? 2. What model ranges are offered to the UK customers, how did this offer develop since 1960, and what trends can be seen? 3. Is the rate of innovation - in terms of product life cycles - increasing in the UK volume automotive market? The analysis is based on empirical analysis of secondary data from various sources: the product variety data was obtained from private archives of historic motor vehicle information material, the registration data was supplied by the Society of Motor Manufacturers, and various historic motormagazines were consulted, with issues dating as far back as 19601. The research presented forms part of two major automotive research programmes, the 3DayCar Programme2 and the International Car Distribution Programme3. 1. Product Variety Product variety is defined as the number of vehicle permutations offered to the customer of a particular model. This variety is generally market-specific, as all major markets have different standard specifications (i.e. the air-conditioning might be offered as standard in the UK, but not in Germany, etc.). Product variety as such has not been studied in detail, although initial studies have been carried out by MacDuffie et al (1996) or Clark and Fujimoto (1991), who proposed initial metrics on how to assess variety and complexity in the auto industry. Their work mainly focused on the actual assembly operation, although the real impact of product variety is on the vehicle distribution strategies, as ICDP (1998) showed. The main impact of variety is that the more variety is offered, the less likely the customer will find the right vehicle amongst finished car stock. Hence the more variety offered, the less successful sales sourcing from stock will work, or the higher the discounting risk, whereby the customer is sold a vehicle which compromises on his original specification, using sales incentives. To assess the trends in variety, the authors analysed the historic development of four major volume cars offered in the UK market from 1964 / 1980 onwards – the Vauxhall Viva, Austin 1100 (BMC ADO16), Ford Cortina and VW Golf (introduced 1974) and their successors. The variety is measured on the number of bodystyles, engines, paint colours, etc. offered, and also shows the total number of permutations offered and the sales volume in the UK for that particular year and model. (see table 1). It may be noted that the Austin 1100 was produced under other brands in the UK - Morris, MG, Riley, Vanden Plas and Wolseley. These variants, which added further complexity to the product 1
Autocar, Motor and UK Motor Show Reviews. www.cardiff.ac.uk/3daycar 3 www.icdp.net 2
range, are not included in the analysis. It’s successor, the Allegro was also offered under another brand in the UK, as a Vanden Plas model. Vauxhall
Austin
(Rover)
Ford
VW
Model Viva I - HA Viva II - HB Viva III - HC Astra I Astra II Astra III Astra IV 1100 MkI 1100 MkII Allegro I Allegro III Maestro Maestro 200 200 25 Cortina I** Cortina II Cortina III Cortina IV/V Sierra Mondeo Mondeo Golf I Golf II Golf III* Golf IV
Year 1965 1967 1970 1983 1984 1993 1998 1964 1967 1973 1979 1983 1987 1991 1998 1999 1964 1968 1972 1982 1983 1993 1999 1980 1985 1995 1999
Bodystyles Powertrains 1 2 2 4 3 5 4 10 4 11 4 11 4 13 1 1 3 4 2 4 3 6 1 5 1 4 2 8 2 9 2 8 3 3 3 5 3 7 3 6 3 9 3 7 3 8 2 6 2 6 2 11 2 9
Paint/Trim Comb. 17 20 23 28 22 30 44 10 12 n/a 11 n/a 105 44 60 106 14 n/a 28 275 110 51 92 26 29 93 211
Options Total Variations UK Sales [year] 1 59 n/a 4 728 100,220 8 15,848 76,338 10 1,495,104 62,570 26 53,575,680 56,511 15 76,972 108,204 41 55,425,024 81,494 3 240 n/a 3 864 131,282 3 448 28,713 6 1,056 59,885 11 n/a 65,328 7 1,132 43,815 17 21,792 68,122 10 14,960 64,928 18 2,742,656 1,170 5 2,688 n/a n/a 2,880 137,873 12 702,464 187,159 28 219,576,000 135,745 35 1,278,852,000 159,804 19 315,072 88,660 16 171,584 77,183 7 7,216 n/a 8 2,192 31,145 8 16,968 44,111 22 154,964 63,715
Table 1: Historic Product Variety, four Major Volume Models As can be seen, the data does not show a general trend towards increasing or decreasing variety. In fact, each of the four models analysed show a different individual pattern. Ford has been rationalising their product variety since the mid 1980s, as for example the 1982 Cortina offered more than half of the paint-trim combinations of the S-Class Mercedes today, yet the 1999 Mondeo is available in only 0.013% of the number of permutations available on the 1983 Sierra. Vauxhall also tried to rationalise their product variety by packaging of options, however with the Astra IV, the large range of options has led to a sharp rise in variety, a total figure similar to the Astra II. For Rover, the variety over the last three decades has been below 25k, and only with the takeover by BMW has a revised option policy exploded the range to 2m+ (this of course, may now change due to Rover’s new ownership since May 2000). For VW, as a non-domestic manufacturer, the variety offered has to be seen in context, as the variety offered in Germany for example might be significantly higher. Nevertheless, VW has steadily increased the variety, offering a wide range of paint-trim combinations and options. The inconsistency of the historic trend also manifests itself in the current variety offered, as shown for seven volume cars in the UK in 1999. Total variety ranges from 408 to 3.9*10^12, paint-trim combinations range from 40 – 250, although no major differences in numbers of bodystyles, paints and powertrains could be observed.
Product Variety - UK Market 1999 No of Bodystyles No of Powertrains No of Paint-Trim Combinations No of Options
200 150 100 50 20 10 GM Ford Astra IV Focus
55m
1.1m
Peugeot VW 206 Golf IV
7.5k
155k
Nissan Primera
820
Honda Accord
408
MB E-Class
3,930bn
Total Variety in number of build specifications
Figure 1: Product Variety, UK 1999 In summary, no major trends on product variety could be established, as the variety offered seems to be guided purely by vehicle manufacturer policy. 2. Product Life Cycle Dynamics It has been frequently claimed that the life cycles in the auto industry have been significantly decreasing over the last couple of decades. The analysis undertaken below shows the life cycles of 20 volume model lines. The life cycle is defined as the time-span from market introduction to model replacement or major facelift4. The shows the year of introduction (X-axis) and length of model cycle (Y-axis). As can be seen, the moving average of 5 years shows life cycles of 3-4 years during the 1960s, which then increase drastically to 7 years in the early 1970s. The reason here is the fact that vehicle complexity was comparatively low in the 1960s, hence development cycles were short, and the successor models generally would share many components*, whereas the models introduced from around 1975 onwards were stand-alone and individual developments. The ability to easily change the appearance of ‘sheetmetal’ (external body panels) and interiors meant that in the 1960s and the 1970s the American owned manufacturers - Ford GB, Ford of Germany, Vauxhall and Opel were able to replace models on a regular basis very economically, and had far shorter model lifetimes, whilst the European manufacturers such as VW, Fiat and Renault preferred to keep a model in production longer and then replace it with a ‘completely’ new model with very different components. The best example of this approach was Citroen, with longlived models such as the 2CV, DS, GS/GSA and CX.
4
Major facelift is defined as substantial alteration of the vehicle, including revision of the bodywork and major components or modules * As an example the Ford Cortina II shared the engines, running gear and floorpan structure of the Cortina I.
As cars became more complex due to technical advances and the impact of legislation, and with the more difficult economic conditions from the late 1970s onwards which affected the VM’s ability to invest in model changes, the American owned companies began to lengthen their model cycles, e.g. the Ford Escort III/IV ran from 1980 to 1990 with one major facelift in 1986. These firms are now forced to shorten their model life cycles due to competitive pressures. However – as shown below - across all manufacturers, since the 1970s, the product life cycles have decreased steadily from 7.5 years to 5 years, and it is predicted that the average age of production for the European volume manufacturers will decrease from 4.2 years in 1994 to 3.1 years by 2005.
Product Life Cycles for the UK Market 1965-2000 New Product Introduction to Major Facelift / Replacement 12
11 Toyota Corolla Honda Civic
10
Honda Accord Nissan Micra
9
Nissan Bluebird-Primera VW Polo VW Golf I-V
8
VW Passat Peugeot 104 - 206
7
Peugeot 304-406 Peugeot 504-607 Ford Fiesta
6
Ford Escort-Focus Ford Cortina-Mondeo
5
Ford Zephyr-Scorpio Vauxhall Viva-Astra Vauxhall Victor-Vectra
4
Austin1100 -Rover 25 MB 'Fintail' - E class
3
AVERAGE 5 per. Mov. Avg. (AVERAGE)
2 1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
Figure 2: UK Product Life Cycles 1965-2000 Summing up – life cycles are decreasing across all manufacturers, and the trend analysis suggests an average life cycle of 5 years in the year 2000 for the models analysed. This trend is backed up by the findings of J.D.Powers LMC (2000), who calculated an average age of production for the European vehicle manufacturers of 3.1 years by 2005, with individual product lives ranging from 2 to no more than 5 years for volume cars.
3. UK Model Proliferation Considering product variety on its on however neglects another element of the offer to the customer – the range of models offered per manufacturer. Whereas in the past a saloon, an estate and possibly an economy or sports car were offered to the customer, nowadays the model range has expanded into many diverse niche segments, such as sport utility vehicles (SUV), multi-people vehicles (MPV), etc.
Model Range offered in the UK Market 1960 - 2000 14 12 No of Models
Toyota
10
Nissan/Datsun Honda
8
Peugeot Volksw agen
6
Vauxhall Ford
4
2 per. Mov. Avg. (European) 2 per. Mov. Avg. (Japanese)
2 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 Figure 3: Model Range offered in the UK Market
This trend clearly can be seen in figure 3, which shows the number of independent models offered by 6 VMs. The moving averages over 10 years show both for Western, as for Japanese manufacturers, a strong trend towards diversification. The Japanese VMs started off with a very limited offering in the late 1960s, however overtook the Western manufacturers in the 1970s, and now offer on average 3.5 models more than their Western competitors, the majority of which are imported from Japan5.
5
UK built vehicles: Toyota: Corolla, Avensis (prev. Carina), Nissan: Micra, Almera and Primera (prev. Bluebird), Honda: Civic, Accord and CRV (From mid 2000).
UK Production Sourcing The trend towards a greater model range offered by the vehicle manufacturers in return influences the sourcing of production. Figure 4 shows the total volume and origins of the UK new vehicle registrations. The total volume is averaging at 2m vehicles / year, although it is suspected that up 300-500k vehicles are pre-registered in the UK every year. UK Market 1995-1999 Total Registrations Domestic and Import 100.0%
2,500,000
90.0% 80.0% 70.0% 60.0% 50.0%
2,000,000
40.0%
Total Registrations Domestic Import EU (excl UK)
30.0% 20.0%
Japan
10.0% 0.0%
1,500,000 1995
1996
1997
1998
1999
Figure 4: UK Market Volumes and Sourcing of Production Considering the increase in model range, the content of UK sourced production is declining, as the sales spread over an increasing number of models - with only a limited number being produced in the UK. Hence the reason why the UK domestic sourcing has been constantly declining since 1995 is partly due to the fact that the model range rate of innovation has been greater than the rate at which more models were produced in the UK.
Conclusion The outcomes of the study show that, despite frequent claims of ‘infinite variety’ offered by the vehicle manufacturers, no coherent trend on the number of derivatives could be established. For some manufacturers in fact the research findings reflect their efforts to reduce variety, which often is disguised using option packaging, yet the figures indicate that product variety is solemnly driven by individual VM policies and does not follow a coherent market trend. In terms of model range and life cycles however, a clear trend towards shortened life cycles could be established, complemented by a trend towards increased model variety per vehicle manufacturer, fostered by the trend to expand into niche markets, such as the multi-purpose vehicle (MPV) and sports-utility vehicle (SUV) for example. Model Proliferation and Platform Use in the European Auto Industry 350
4.5
4 300
Ratio Bodytypes/platform 3.5
250 3
200
2.5
No. of platforms in use (all Europe) No. of body types offered (all Euope)
2
150
Av. Volume by platform ('000) 1.5
Av. Volume by body ('000)
100 1
Av. No. bodytypes/platform
50 0.5
0
0 1990
1992
1994
1996
1997
1998
1999
2000
2001
Figure 5: European Model Proliferation and Life Cycles, Source: Smith, Salomon & Barney The conclusions that can be drawn from the data are that the rate of innovation is clearly accelerating, with more models offered, that remain on the market for shorter periods than in the past (Figure 5).
This puts high pressure on the VMs to: !
Develop and launch vehicles within a shorter timeframe, as current product development times of 3+ years would not permit to compete in a market which innovates every 2-3 years
!
Produce vehicles profitably at lower economies of scale. As the life cycle decreases and sales spread out over many models, the average volume per vehicle decreases sharply (also Figure 7). Hence the breakeven point needs to be achieved much earlier on than previously. The common way of achieving this is to deploy a platform strategy, and the predicted average volume per platform shows this trend clearly. Current platforms will host more models than previously, as for example the current VW A4 platform, which hosted nearly 2m vehicles in 19996.
!
Align the distribution strategy to the product variety offered, as a misalignment inevitably leads to high discounts needed to sell the vehicles. In effect a VM offering high variety would need to embrace a build-to-order strategy, whereas a low-complexity manufacturer might be able to utilise central vehicle stocks efficiently.
The rate of innovation in the automotive market has increased significantly, and the VMs will have to face the challenge of responding to customer demand and competitors’ action in an ever more responsive fashion in the future, whilst being constrained at the same time by much shorter sales windows and lower overall volumes per model to cover their costs. References Clark, K.B., and Fujimoto, T. (1991), ‘Product Development Performance’, Harvard Business School Press MacDuffie, J.P., Sethuraman, K., and Fisher, M.L. (1996), ‘Product Variety and Manufacturing Performance’, Management Science, Vol 42, No 3 ICDP (1995), ‘Supply and Stocking systems in the UK Car Market’, Solihull, UK ICDP (1998), ‘European Supply and Stocking Systems’, Solihull, UK J D Powers – LMC (2000), ‘European Model Life Cycle Dynamics’, Oxford NN, ‘Autocar’, Various historical issues, now published by Haymarket Motoring Publications NN, Historical Sales Brochures published by Ford Motor Company, Vauxhall Motors Ltd, Rover Group Ltd (and predecessors), Volkswagen AG and Daimler-Benz AG. NN (1975), ‘The Future of the British Car Industry’, Central Policy Review, London, HMSO NN, ‘Observers book of Automobiles’ Various years, published by Frederick Warne and Olysager Organisation. Ward, A., Liker, J.K. and Christiano, J.J. (1995), ‘The Second Toyota Paradox: How Delaying Decisions Can Make Better Cars Faster’, Sloan Management Review, Vol 36, No 3 Womack, J, Jones, D, and Roos, D (1990) ‘The Machine that changed the World’, Rawson Associates, New York 6
1,904,755 vehicles were produced on the VWA4 platform in 1999. The platform included VW Golf, Beetle, Bora/Jetta, Audi A3 and TT, Skoda Octavia and Seat Toledo and Leon. (Source: Automotive World, June 2000)
