BMW’s 3-Series: Managing Platform Design and Development Costs

1/3 September-December 2000 25

BMW’s 3-Series: Managing Platform Designand Development CostsWilliam R. BoultonAuburn University

Michael DowlingUniversity of Regensburg

BMW’s philosophy is to build driv-ing machines that respond faithfully andenjoyably to their driver’s commandswhile also providing the safety, practical-ity, style, quality, reliability, and durabil-ity that help make long-term ownership arewarding experience. It was BMW’spolicy to develop new platforms for itsseries of cars every 7 to 8 years. Newplatforms were not simply modifications,but completely new structures. Accord-ing to Friedrich Nitschke, manager for theBMW 3 series development project, therewere three goals for the new 3 series plat-form that came to market in the 1999model year. First, it was to generate thehighest level of customer satisfaction inits class. Second, it was to utilize the bestprocesses available within BMW. Third,it was to generate the most profit of anyBMW series.

Developing the successor to the third-generation of 3 series was a real chal-lenge, since it had been an extremely suc-cessful series. BMW’s chief designerChris Bangle explains,

There are two ways of doing car de-sign. Either it’s a personality cult,where the designer runs the show andthe car is just an ego toy, or the stylistis more the curator of a heritage andtradition. Then, the challenge lies inunderstanding a marque so well youbecome part of it. It’s quite clear myjob here is to perpetuate a set of icons.Bangle’s latest platform introduction

had been the redesign of its 5-Series, asshown here.

The overall styling of the 3-series wasespecially important, and was a limitingfactor in its design. Models built on thenew platform had to look like a BMWand be recognizable on the highway.Focus groups were used to identify thebasic design features that made a BMWrecognizable, and those traits were main-tained in its design. For example, theBMW logo and “double kidney” grillwere key recognition factors. The 3 se-ries position in BMW’s product line setmany of the basic characteristics for thenew platform, and would determine thebasic characteristics of the models usingthe new platform. For example, the di-mensions, engines and transmissions be-came part of its target definition.

Building on Tradition

The Bayerische Motoren Werke (Ba-varian Motor Works) was created in 1916through the merger of an aircraft makerand a manufacturer of aircraft engines.BMW’s first notable success was the 6-cylinder BMW IIIa engine, which in 1918powered a biplane to 5000 meters alti-tude (16,405 feet) in just 29 minutes. Itsimpressive performance led to strong de-mand for BMW engines. In 1919, aBMW-powered biplane set a world alti-tude record of 9,760 meters.

In 1922, BMW produced a small en-gine for the Victoria motorcycle, and abig truck engine with an advanced over-head camshaft. BMW introduced theBMW R32 motorcycle in 1923 using theBMW Boxer engine [a horizontally op-posed twin-cylinder engine], rear-wheeldrive, and a double-tube frame. Its pio-neering technology was still used in thefour R1100 series models introduced in1993.

In late 1928, BMW acquired theEisenach Vehicle Factory, which pro-duced a single licensed model of

England’s little Austin Seven. As BMW’sfirst automobile, it was known as the 3/15 or “Dixi.” In 1932, a new, larger modelcalled the BMW 3/20 was introduced. Ayear later, BMW introduced its first sportssedan, a 6-cylinder model called the 303.In 1934, the 303’s engine was enlargedfrom 1.2 to 1.5 liters, and the model-nam-ing system was instituted with the 315/1sports roadster. BMW’s 6-cylinder enginewas progressively enlarged, to 1.9 andthen to 2.0 liters, and finally to 3.5 liters.

After World War II ended in 1945,BMW’s Eisenach plant, where all BMWcars had been produced, was in East Ger-many. BMW rebuilt its bombed-outMunich plant and began production ofmotorcycles, then tiny hybrid cars, calledthe Isetta, powered by motorcycle en-gines. Within a decade, BMW introducedsedans and roadsters with V-8 engines,including the 507, one of the Europe’smost coveted postwar collector cars.

BMW introduced the 1500 family carin 1962, with responsive overhead-cam-shaft 4-cylinder engine, front disc brakesand 4-wheel independent suspension.Germany’s unlimited driving speeds en-couraged BMW to introduce a perfor-mance model as well, thereby beginningBMW’s modern sports-sedan tradition.The BMW 1500 spawned a smaller 2-door version and the 2002, which intro-duced Americans to BMW’s sports-sedanconcept. A new 6-cylinder version, likethe Bavaria sedan and 3.0 CS coupe, wereintroduced.

Since 1975, BMW has referred to itssmallest line of cars as the 3-Series. Thefirst 3 series achieved higher performancewith its 6-cylinder engine. The secondgeneration provided more choices, in-cluding 4-doors, convertible and touringmodels. The most successful third gen-eration raised driving performance to aneven higher level, while providing qual-ity comparable to Mercedes-Benz. The

The BMW 528l

Journal of SMET Education26

fourth generation would again undergo aredesign of the various models in the se-ries. The first models to change were thesedans, next the coupe, then the convert-ible and hatchback, and finally the wickedM3 sports cars. This sequence of intro-ductions allowed BMW to extend the lifeof the design by stretching out introduc-tions over the life of the series. The cur-rent 3 series design had been in produc-tion since 1990. The styling of the new3-series would be similar to the larger 5and 7 series cousins.

BMW’s Forth-generation3-Series PlatformDevelopment

As one journalist said of the third gen-eration models, “BMW’s beautifully pro-portioned 3-series cars are still the designpacesetters in this realm” of small sportsedans. Friedrich Nitschke’s challengewas to develop the fourth-generation ofBMW’s pacesetting tradition. In addition,Nitschke was assigned the task of imple-menting a new planning system forBMW’s 3 series replacement platform.He explained:

I was given responsibility for devel-oping the new 3-series BMW in 1993.This development process (Figure 1)was introduced in BMW for use in thenew 3 series. The target definitiongives the basic characteristics of thenew platform for the series.Separate teams designed the saloon

(sedan), coupé, compact, touring, con-vertible, and M3 sports models to be in-troduced on the new platform. Models in-cluded two or four door versions of fam-ily and sporty cars. For example, the ex-isting lineup included:

Saloon (316i, 318i, 323i, 328i, 318tds,325tds, M3)

Coupé (316i, 318is, 323i, 328i, 328iSports, M3)

Compact (316i, 318tds, 318ti)Touring (318i, 323i, 328i, 318tds,

325tds)Convertible (318i, 323i, 328i, M3)The first model on the new platform,

the four-door sedan, was to be introducedin 1998, using a more powerful 166 horse-power 2.5-liter inline-six engine (323imodels) or the 190 horsepower 2.8-literinline six (328i models). The less power-ful in-line four-cylinder engine was be-

Figure 1: BMW’s Product Development Process

ing dropped from use in the sedans. Overthe new platform’s life, existing modelswould be replaced by the new platform,and then beefed up as an M-type model.

Setting the budget.

The top-down target cost budget forthe project was set as shown in Figures 2and 3. BMW’s marketing departmentset the price range for the 3 series betweenDM 40,000 and DM 60,000, pricing thecompact 316i at around DM 42,000, andthe 328i sports sedan at around DM60,000. At these prices, marketing esti-mated sales between 2 million and 3 mil-lion cars over the platform life. Manage-ment then set the target revenue and profitlevel required for the series at 3 millionunits. Mr. Nitschke explained:

This strict product development pro-cess was introduced with the 3 seriesin 1994. We would not start the devel-opment of the 3 series until we weresure that we could make money withthis car. That means that the plan’scost and target cost needs to be nearlyequal. In the past, we might have somedifference, but we would start the R&Dprocess in hopes of finding ideas toreduce costs. Therefore, we wouldn’tmeet target profits and would have toimplement cost reduction projects assoon as we introduced the new car.

Market Inputs

The concept definition phase of de-velopment relied on market inputs. Mar-ket inputs included data on appeal, cus-tomer satisfaction measures, productbenchmarks, product reverse engineering,customer call center inputs, market stud-ies, product tests, patents, conjoint analy-sis, and SWOT analysis of the competi-tion. The third generation 3-Series mod-els were compacts, but offered minimalpassenger comfort. Some customers had

complained about its cramped rear seatand sub-par interior. The hard plastic in-terior was heavily criticized. New com-petition from Lexus, Infiniti andMercedes caused customers to questionspending upwards of $40,000 on a car thathad a second-rate interior.

Conjoint analysis, shown in Appen-dix 1, revealed that safety and road be-havior were the two most important fac-tors in purchasing a new vehicle. Sincethe early 1990s, BMW had investedheavily, producing the safest and bestdriving machine on the market. For ex-ample, by 1997, BMW planned to have ahead protection system above the frontdoors and within the A-pillar of its 7-se-ries, which inflates an airbag diagonallyacross the side window. Rear sideairbags, and driver and passenger sideairbags were planned for 1996. Dualthreshold deployment was an “intelligent”safety system that kept airbags from de-ploying at a lower crash severity if seatbelts were worn. Passenger seat occu-pancy sensor recognizes if a seat wasoccupied or not. Interlocking door an-choring system anchors a diagonal alu-minum reinforcement bar to the body pil-lar in the event of a serious side impact.The newest systems were typically of-fered on the 7- and 5-series first.Benchmarking the existing 3 Series, theMercedes C series, the Audi 90, and, fornoise, the more expensive Lexus was thebasis for setting feature standards for thenew platform. Mr. Nitschke explained:

A goal for the new 3 series is to be oneof the quietest cars in its class. As of1994, Lexus didn’t have a direct com-petitive model, so we looked at thelarge Lexus for its noise control. Ourtarget was to reach that level of quietin the new series. The existing 3 se-ries is also a benchmark model, sincewe want to improve on it. It is good todrive, but the space is restricted and


the cockpit isn’t great. It is nice, butwe will improve the value to the cus-tomer. These were the inputs from themarket.Mr. Nitschke reviewed the importance

and age of each component.More interior space was a target that

resulted from prior owner responses,as was the need for a more luxuriousinterior. The suspension is an impor-tant item in determining the interiorspace. Changes in rear and front sus-pension provide more interior space.Since we have a rear wheel drive, wecan push the front suspension closerto the front of the car. This is one ofthe reasons that the interior space canbe expanded. We can also bring the en-gine behind the front suspension to givethe car a more balanced weight ratiobetween the front and the rear. We alsodecided to add two centimeters to thelength and three centimeters to thewidth to give additional interior space.Many of the characteristics that thecustomer does not see, like the basicstructure, can be carried over from theprevious model.

The resulting technology analysis pro-vides the characteristics of the new model(Figure 4). Nitschke continued:

We look at the competitors’ noise lev-els, aerodynamics, etc., and create inour head what is needed for the newcar to be better than the old one. Weconsider quality, transportation, safety,comfort, and environmental features.We then compare each feature againstthe old car and competitors’ cars todetermine what we need to do in eachcategory. In quality, we need to be bestin class. In space, we just want to bemarginally better. Everybody knowswhat the key points are for improve-ment in the new car. Our customerinputs made that clear.

The target definition included more inte-rior size and higher quality appointments.According to Dr. Anton Heiss, teamleader for manufacturing,

The old 3 series had a problem gettingfeet under the rear seat. It also feltcramped, so we needed more space be-tween the front seats. We decided toadd space to the rear and between thepassengers. We also decided that thestyling of the dash needed to give amore open feeling.

Figure 2: Integrated Market-orientedTop-down Target Cost Management

Figure 3: Target Cost Budgeting Model

Discussions included leather trim andoptional leather upholstery, power seatsand windows, an on-board driving com-puter, alloy road wheels, ABS brakes,dual airbags, and air-conditioning. Theteam also wanted it to achieve the USA’shighest safety rating. Each module andcomponent was analyzed. Nitschke con-tinued:

We evaluate each component to deter-mine what is good enough, and whatmust be improved. We created a sum-mary of each area. We know that thebody must be new. The customer cansee that it is new, and not a face lift.All parts for the skin must be new.There are about 80 such parts that willbe new. In the structure, we know thatthe new safety laws and crash tests re-quire some new parts. The old struc-ture will not pass these new tests. The

impact tests and roll over tests will bemore stringent in the future. In Cali-fornia, these tests are required. Wethen determine what parts of the struc-ture must be new, and what can be usedfrom the old model, or taken from the5 or 7 series. To improve quality, a de-termination has to be made on whatparts must be new and what is goodenough, so that there is no real qualityproblem.The cost of the proposed design im-

provements then had to be determined.Table 1 shows how the parts’ costs aredetermined in terms of planning, R&Dand investment costs. A vehicle includesabout 4000 different part numbers, or13,000 total parts. An engine, for ex-ample, uses the same parts for each cyl-inder. Each of the 4000 parts are classi-fied into modules and related compo-


Figure 4: Next Generation 3 Series Profile

nents. For example, the interior area in-cludes the cockpit and seats. The seat isone module, and the frame for the seat isone component. Each component in-cludes a number of parts. Each part IDnumber was identified.

For example, Nitschke’s team had todecide if the back seats from the 3 or 5series could be used in the new platform.Each part was either used from the prior3 series, incorporated from the 5 or 7 se-ries, acquired from a supplier, or devel-oped new. Nitschke explained:

We identify the ID for each part, andthen decide what action is needed. Theinside hood, for example, can be usedfor the new series, but the hinge mustbe new. Of the 335 parts in Table 1,we will create 212 new parts. Theairbag generator can be used from the5 series, so that saves planning, R&D,

and investment costs. This analysis letsus determine at an early stage the costrequired for development, R&D, andinvestment in each technology area.We haven’t pre-determined how muchtechnology will be transferred from the5 or 7 series. When we tried to do thatin the past, it didn’t work. The marketmoves too quickly to determine whatparts might be used in a later series.We had developed a new seat for the 5series that can be used in the new 3series. The marketing analysis tells uswhat features we need.The budgets specified the basic ve-

hicle and its standard features. The bud-get provided DM15,000 for the basic ve-hicle and DM5,000 for the standard fea-tures. A small budget of DM 2000 re-mained for features that may be includedlater, but were not specified. Since the

concept definition phase occurred 36 to40 months before launch, some flexibil-ity was needed in deciding on innova-tions. Nitschke noted:

If we think the car is good enough, wedon’t need any innovation expense.However, the research department haslots of ideas for new features. The onlyproblem is that they must be able tobring them to high quality within thetime needed for launch. If we need twoyears for development, then we may beable to include those features in thenew car. If the quality is not goodenough, we don’t want it.BMW’s goal for the basic model is tohave the best quality and crash per-formance in its class. That is part ofthe basic cost. We want the highestsafety ratings in this class of car incrash and roll over tests. The basicsafety in terms of brakes, steering,frame, etc. is built into the basic car.Some of the innovation money can bespent for safety, and not just gimmicks.Additional side impact airbags cancome out of the innovation cost.It is difficult to determine what is agood feature, and what is a gimmick.The rain detector, for example, wasoriginally developed for the 7 series,but the marketing department said itwasn’t a feature that customers wanted.They said that customers would con-sider it a gimmick or a toy. ThenMercedes Benz introduced a rain de-tector, so marketing said we had tohave it. In the third generation 3 se-ries, customers loved the automaticdimming of the interior lights after youclose the door. So that feature will bein the fourth generation 3 series. Weusually have more ideas for innova-tions than we have budget. How doyou decide what is right for this car?BMW’s research and development

center (RDC) employed nearly 5,000 en-gineers, designer, technicians and supportpersonnel in one of the world’s largestadvanced automobile development facili-ties. Since studies have shown that engi-neers develop 80% of their ideas in pri-vate conversations with their colleagues,the individual offices and workstations foreach department are no more than 50meters away from each other. The ideaof streamlining communication has beenapplied to the management structure, too,


Table 1: Technical Scoring Framework

thereby reducing the length of decision-making paths. The five buildings and90,000m2-floor space that comprise theRDC are completely linked in a massivedata communications network. At theheart of this is a CRAY X-MP/28supercomputer and 800-plus CAD/CAMand CAE terminals for development ofnew designs and technologies of tomor-row.

The development teams also considerthe additional options for the new series.For example, other BMW series offeredsuch options as a sonar distance device(for crunch-free parking), automatic trac-tion control, a limited slip differential,metallic paint, sunroof, an anti-theft de-vice, powered front seats with settingsmemory, and CD players.

Cost Control

With 70% of the cost of a new seriesbeing determined during the planningphase, BMW’s new planning process hadhigh visibility. The model profitability isdetermined after all direct and indirectcosts are estimated. Direct costs includethe project costs from planning, develop-ment, investment, and start-up. Income,variable production costs, and warranty

costs are also direct costs. Indirect costsinclude overhead costs for facilities, op-erations, and unit operations. These bot-tom-up planning costs are then comparedagainst the top-down target costs andmodel profitability levels established bytop management. Nitschke continued:

After we complete the technical analy-sis, we total the planning costs, theR&D costs, and the investment costs.We then have a matching processagainst the target goal. We had a dif-ference of between 15 to 20%. We hadbudgeted DM 2000 for innovation, butwe don’t know what that will be at theplanning stage.We have some cost drivers that we canchange. One is technology. For ex-ample, to develop a new suspensionsystem required investment costs ofsome DM 200 million; but, if we carryover some of the old parts, the invest-ment costs are only DM 80 million. Welook at what we can carry over to lowerour total costs. On one hand, we needno R&D and investment for old parts.On the other hand, the seat of the 5series costs DM1000, but we have abudget of DM 800 for the 3 series. Youhave to develop a new seat to realize

the DM 200 difference. We look forways to save investment and R&D, butthe cost of an existing part may actu-ally be higher.During 1994, the matching process be-

tween top-down targets and bottom-upcosts took 3 months.

The actual model profitability was de-termined through a complex computeranalysis. This sensitivity analysis (Fig-ure 5) was used for the matching process.For example, the model profitability fig-ure is the overall profitability of the car.Top management had set a goal to gener-ate a 26% profitability level. As is usualin product planning, the initial bottom-up cost plan exceeded the top-down tar-get by nearly 20%. Nitschke’s team hadto find ways of improving profitability.

We first asked marketing if they couldcharge a higher price or sell more carsbecause of our technical improve-ments. Of course, marketing said thiswas impossible, so we had to deal withthe planned cost factors. In our sensi-tivity analysis, an increase of 1% inprice (say to DM 40,400) can increasethe model profitability by 2%. Whenwe have lower project costs of about1%, we can increase the model profit-


ability by nearly 1%. With this analy-sis, we can determine if it is better tochange the part cost, or the investmentcost. For example, when we have alow volume car like the 8 series, it isbetter to reduce investment and R&Dcosts, rather than reducing part costs.A model like the 3 series, with esti-mated sales of 3 million units over itslife, it is better to reduce the parts costs.Therefore, this sensitivity changes witheach model. This matching process was central to

the new planning process, and requiredserious reevaluation of the planned costs.Nitschke explained:

Until we actually start the develop-ment process, the car is simply in ourheads. We must complete the planningphase within six months, or we delaythe actual development. Since the first matching had higher

costs, we looked again at parts thatcould be carried over. We wanted todevelop a new suspension, but the oldsuspension was very good. Instead ofdeveloping a new suspension, we de-cided to reduce the weight of the oldone. We decided to change the mate-rial of some parts, by using aluminumand magnesium rather than steel. Thatsaved some investment and R&D costs. We began the actual development

phase with 5 to 6% higher plannedcosts than target, but decided to de-velop some new process that wouldhopefully recover those costs in pro-duction and purchasing. For example,we brought some suppliers into thedevelopment team as model life sup-

Figure 5: Sensitivity Analysis for Model Profitability

pliers. Five years ago, we had suppli-ers compete for contracts, so theynever knew who was the model life sup-plier. As a result, we couldn’t bringthem onto the development team. Nowwe asked suppliers to compete on con-cepts, and the supplier with the bestconcept is chosen as the model life sup-plier. Now our supplier actually de-velops the seat. Some of the purchas-ing people don’t like this approach,since they prefer to push the pricedown with competitive bidding. WhenLopez (Opel’s ex-purchasing manager)moved to VW, he started a process thatVW still uses to push its suppliers toreduce prices. We hope that our newapproach will save money in the longrun. For example, since our suppliersare on the development team, we askthem to make changes without addingadditional costs, and they usuallyagree. Before, any change providedopportunity for the suppliers to raisetheir prices. By launch, we hoped tomeet the 26% model profitability goal.

The New 3 Series Platform

The current 3 series models had littlecommonality in parts, except for the en-gines, transmissions and floor structure.The development team wanted more com-monality between the new 3 series mod-els in order to reduce costs of manufac-turing and improve overall quality. Al-though transmissions and engines werelargely carried over from existing mod-els, the chassis was totally new. Accord-ing to Dr. Anton Heiss:

For the new 3 series, we decided tohave more common parts. Everythingup to the instrument panel will be thesame for the sedan, coupe, and tour-ing models. Also, the sunroof is thesame. The doors for the convertiblewill be the same as on the coupe. Thatis more than in the past. All the skinsare totally different between the mod-els. All of the discussion about what will

be carried over and what will be neware styling issues. We decided that wehad to have different styling for eachmodel. On one hand, nobody wants tohave totally different cars. Every carhas to be almost the same shape. Butit is impossible to have the same frontwith the same windscreen, and have abody with a new shape. You have tochange the windscreen angles andcurves to fit the coupe or convertiblestyling. That requires a totally newgreenhouse (cabin) for each model. Wedidn’t want to make compromises inkeeping the look of a BMW, so we mustproduce a new skin and give it a propershape. We want each model to looklike a BMW, and not like a Toyota.To increase rear-seat passenger room,

the all-steel body was about 4 cm (1.6inches) longer. In addition, new frontseats were being developed to give morespace beneath them for rear seat passen-gers to put their feet. The added weightof the larger frame was being held downto 27 kg (60 lbs.) through the use of alu-minum, particularly in the suspension andfront sub-frame assembly. The plannedrollout of the new sedan models wasscheduled for the second quarter of 1998.The coupe model was planned for 1999.A touring wagon and 3.4-liter M3 was tofollow in 2000, with the compact andconvertible models in 2001. M-tunededitions would then supplement modelsto keep the cars exciting through theseven-year life of the new platform.

The new 3 series platform added ad-vanced safe driving technologies to meetfuture safety requirements. First, it wasone of the most balanced cars on the roadwith a near 50/50 ratio between front andrear weight distribution. Second, whileit was a rear wheel drive, it also includedall-season traction and an antilock brak-ing system (ABS). BMW’s corneringbrake control (CBC) was a further devel-


opment and expansion of the antilockbraking system for braking maneuvers oncurves. CBC regulates the pressure inthe different wheel brake cylinders foroptimal breaking. BMW’s advanced ABSsystem included a recirculating pump inthe hydraulic unit that activates in emer-gency braking situations. This dynamicbrake control (DBC) assured maximumdeceleration and the shortest stoppingdistance. A new rear axle drive systemarchitecture and suspension improveddriving performance and body reinforce-ment. About 20 percent of the axle andsuspension parts were remade of alumi-num to save weight and improve drivingcomfort.

Cost Reduction Strategy

As the new 3 series entered the devel-opment phase, Dr. Anton Heiss was giventhe task of bringing the new model tomarket within the cost targets. Since theinitial design phase had not met the tar-get cost levels, the factory had to makesignificant improvements. Heiss ex-plained:

This was the first series where webrought manufacturing people into theprocess at the beginning of the plan-ning process. This was very helpful.At the beginning, we started with a 5or 6-person team one year before theplatform team was founded. We hadone from marketing, one from finance,one from styling, one from general carintegration, one from interior and er-gonomics, and one from driving char-acteristics.At the beginning of the project, I didall the feasibility studies for manufac-turing. What can we do, and what isfeasible to manufacture? Second, Istudied the manufacturing problems ofthe old 3 series to find solutions fromthe manufacturing point of view. Wewanted to avoid all the old problems.The third job was to calculate the in-vestment for the new manufacturingline. At the beginning, it was a roughestimate, determining what new toolswere needed in comparison to the ex-isting line. Once we decided on thenew version of the 3 series, we mademore detailed studies. We came up withall the manufacturing requirements,what is in the new car, and then beganwith styling the new model. For body-

in-white, we set targets for the num-ber of welding points. The number ofparts used for body construction af-fected that.After studying the old assembly line,

Heiss found over 950 items or processesthat needed to be changed or fixed. Inplanning assembly of the first prototype,about 600 were addressed. About 200were not worth the cost or investment ofthe improvement. The new productionphilosophy was to handle parts or assem-blies only once, or complete them in asingle line. While assembly only ac-counted for about 10% of the total cost,Heiss and his team had cut those costs byover 10%.

Changing Competition

As BMW’s new 3 series prototype wasabout to be completed, a major competi-tor, Audi, introduced its new benchmarkmodel for 1996, the Audi A4. Car andDriver magazine named the A4 one of theTen Best 1996 cars. In fact, Car andDriver gave the new A4 a one-pointhigher rating than BMW’s 1996 328i. TheA4 was central to Audi’s strategy of di-rectly attacking BMW’s product line.Having hired away several of BMW’smarketing executives, Audi was now in-troducing series with numbering systemssimilar to BMW but always symbolicallyone number higher. The Audi A4 wastargeting the BMW 3 series. The AudiA6, introduced in 1995, was targetingBMW’s 5 series.

The Audi A4’s design was very com-petitive with the BMW 3 series. The A4’sstandard items included cruise control,anti-theft system, automatic climate con-trol system, AM/FM/cassette and reardefogger. It included power windows andlocks and dual power outside mirrors—along with a power driver’s seat and tilt/telescopic steering wheel to make driv-ers of various sizes comfortable. Thewheel was even leather-wrapped. En-hanced safety items included dual airbags, anti-lock brakes and side-impactprotection that met 1997 federal stan-

dards. It also included 5-m.p.h. bumpers,rear headrests and a headlight-washersystem. Audi gave a three-year, 50,000mile warranty that includes everything,even scheduled maintenance. Reviewspraised the styling and roominess of theA4, along with its luxurious interior.According to one reviewer, “Walnut in-lays in the dashboard, center console anddoors convey a feeling of opulence notgenerally found in a car that costs lessthan $30,000.” An optional Bose musicsystem was available. The primary com-plaints of reviewers came from the lim-ited legroom in the rear seats, the poorcup holder design, and the difficulty inputting the stick shift into reverse.

Overall, the A4’s styling was consid-ered more attractive than anything beingoffered by Lexus or Acura, the luxury wasequal to big American cars such as theCadillac Seville or Oldsmobile Aurora,performance was on par with top sedansfrom BMW and Mercedes, and safety wasclose to Volvo standards. The biggest con-cern to Nitschke was the A4’s price tag,about $4900 less that BMW’s 328i (seeAppendix 2). Nitschke’s team felt addedpressure, since supplier prices had in-creased as technical improvements wereadded during the development phase.While the new 3 series was only 24months from launch, Nitschke saw Audi’sA4, Mercedes C-class, and the new LexusCS300 as the new benchmarks for the 3series.

Besides the price, team members weresurprised with the A4’s driving charac-teristics. It handled better than expectedfor a front-wheel drive car. Nitschke ex-plained:

The old Audi 80 had been our originalbenchmark, but the A4 quality is muchbetter, and the interior appointmentsand suspension system are much bet-ter. The Audi 80 had been a conserva-tive car that was considered the choiceof civil servants. The A4 is a seriouscompetitor and is a well-developed car.We need to reevaluate our budget forinnovations to determine how we canrespond to the A4. The budget hasbetween DM 500-1000 still availablefor innovations.The team considered redesigning the

new 3 series for more significant cost re-ductions, but was concerned that a rede-sign would significantly delay the plat-


form launch and negatively affect the fi-nal quality of the new model. Further-more, revisions and major alterations tothe development prototype could signifi-cantly affect the model’s profitabilitysince development costs were included inthe targeted costs (Figure 6). Significantalterations could easily add an additional25% to the overall development costs.However, new competitive entries raisedquestions about the new 3-series competi-tiveness for the coming decade, since theseries would be in the market from 1998to 2005.

The team wanted to make sure that thenew 3 series would surpass the A4 and

Figure 6: Cost Drivers in Developing the New 3 Series(million DM)

compete head-on with the Mercedes C-class in buyer satisfaction. They decidedto review existing technology and inno-vations that could be included in the newmodel. Given the A4’s lower price, theyhad to review what features to include inthe base model to justify the higher priceof the BMW, and decide what optionsconsumers would pay extra for. R&D haddeveloped a remote keyless entry with 2stage unlocking power door locks. Driv-ers could press the button once to openthe driver’s door, or twice to open theother doors and the fuel filler flap.BMW’s new remote locking system couldalso include a power remote trunk release.

Cadillac had recently introduced a simi-lar system. The 5 and 7 series powerdriver’s and front passenger’s seats had3 memory settings that could be included.A smart key system could be programmedto remember the seat settings for eachdriver.

Nitschke also considered raising thesafety level of the new 3 series by addingthe HPS (head protection system) airbagfor about 150 DM. The highly accreditedAmerican Insurance Institute for High-way Safety (AIIHS) crashed two 5 seriesBMW’s sideways against an obstacle at20 mph. In the car without the HPSairbag, the collision would have been fa-tal. Using the HPS airbag, the injury rateswere one-fifth those of cars without thesystem, ensuring the driver’s survival.According to Brian O’Neill, president ofAIIHS, “This first of its kind head pro-tection system from BMW demonstrateswhat can be done to protect people in se-rious side impacts.”

The cost parameters would be affectedby changes to the new 3 series in light ofthe Audi A4’s introduction. Of course,the 1996 benchmark models had nowbeen introduced, and could be used toassess the new design. The question nowwas, “How far should the team go in al-tering the 3 series model to be competi-tive?”

Appendix 1: Conjoint Analysis of Critical Product Qualities and Components


Appendix 2: Comparisons of Benchmark Vehicles


Prof. Dr. Michael Dowlingwas named to the professorship for In-novation and Technology Manage-ment at the University of Regensburg

effective July 1,1996. Previouslyhe had been anAssistant Profes-sor and AssociateProfessor withtenure at the Uni-versity of Geor-gia. Prof. Dowling

was born in 1958 in New York, USA.He studied at the University of Texasin Austin (Bachelor of Arts with HighHonors), Harvard University (Masterof Science) and University of Texasat Austin (Doctor of Philosophy inBusiness Administration). As an ex-change student he also spent two yearsat the Ludwigs-MaximiliansUniversität München. He has workedas a Research Scholar at the Interna-tional Institute for Applied SystemsAnalysis in Laxenburg, Austria and asa Research Analyst with McKinsey &Company in Düsseldorf Germany.

His research interests include thestrategic management of technologyin the private sector, especially in theinformation technology industry, andthe relationships between technology,public policy and economic develop-ment. He has published articles inStrategic Management Journal, Man-agement Science, California Manage-ment Review, Research Policy, Busi-ness Horizons, Columbia Journal ofWorld Business, and Telecommunica-tions Policy.

He has received academic honorsincluding Phi Beta Kappa, a Rotary In-ternational Fellowship, a FulbrightScholarship, a Sarah Moss Fellowshipand a SEL Foundation Fellowship.

He has also consulted with variouscompanies and organizations includ-ing Bell South Corporation, South-western Bell, and UBM Consulting.

About the Authors

Dr. William Boulton is theC.G. Mills Professor of StrategicManagement in the College of Busi-ness at Auburn University since 1990.

His research isfocused on Glo-bal, Technology-based Competi-tion. He wasVisiting Profes-sor at the Uni-versitat Seminarder Wirshaft

(USW) in Germany in 1999 and TheKeio Business School in Japan dur-ing the Fall Semesters of 1992 and1993. He was a Fulbright researchscholar in Japan in 1986 and was avisiting scholar at Japan’s prestigiousInstitute for Fiscal and FinancialPolicy in 1993.� He had completed nu-merous research reports for U.S. gov-ernment agencies on electronic manu-facturing technologies and industrialdevelopment policies of Asian coun-tries that include Electronic Manu-facturing and Packaging in Japan(1995), Semiconductor and Elec-tronic Manufacturing in the PacificRim (1997), Information Technolo-gies in the Development Strategiesof Asia (1999), and Electronic Manu-facturing in Hong Kong and China(2000).� He is author of numerous ar-ticles and cases covering topics ofcompetitive strategy, strategic plan-ning, boards of directors, and technol-ogy and innovation management. Healso co-authored The Japanese Elec-tronics Industry (1999), and authoredThe Resource Guide for the Man-agement of Innovation and Technol-ogy for the AACSB (1993), and Busi-ness Policy: The Art of StrategicManagement (1984).


Teacher’s Note1

BMW’s 3-Series: Managing Platform Designand Development Costs

This case is about the development ofBMW’s new 3-series platform, which wasinitiated in 1994 for introduction in 1998.It is based on interviews with the projectleader and key team members. The casedescribes the key tools used in the con-cept design and target costing processesof product development. It summarizesmany of the tools being used for productdesign and development by leading edgefirms, such as BMW, including targetcosting, market analysis, product assess-ment, module evaluation, and technologyassessment. It also raises the problems ofdynamic competition during the develop-ment process. Key management ques-tions include:

1. What are the critical inputs to prod-uct design and development? Howdo you integrate product designwith and development with costmanagement?

2. What happens when competitiveproducts are introduced during theproduct development process?What is the impact of designchanges on your costs?

3. How do you maintain strategic flex-ibility and manage product costs ina dynamic world?

Intended Audience

This case was designed for use in ex-ecutive development programs and in ad-vanced courses in product development,technology and innovation management,and strategic management. It provides acomprehensive view of the concept de-sign process, which makes it a good in-troductory or summary case. The caseprovides the instructor with an opportu-nity to stress the dynamics of the com-petitive conditions under which productsare being developed. For the BMW 3-series, Audi introduced its new A4 modelin 1996 to compete directly with BMW3-series. The case also compares newbenchmarked vehicles, includingMercedes and Lexus models.

Teaching Objectives

This case is set in the dynamic worldof automotive design. The teaching ob-jectives are related to the interactions be-tween product design and cost manage-ment during the product developmentprocess. There are several directions thatyou can take. One is to focus on the prod-uct design and development process it-self. For example, you can put the classinto teams and have them begin planningfor the next platform, since it is now timeto start the process again. You can thenadd the competitive challenge of short-ening the total development cycle to andindustry goal of 24 months. Two is tofocus on the basic characteristics of thecase itself. Either approach would ad-dress the following questions.

1. What do you consider the importantcharacteristics of a BMW? If youwere to appoint BMW’s next gen-eration 3-series project team leader,what basic guidelines would you setfor the team? Why are they impor-tant?

2. How would you evaluate BMW’sproject management tools? Howdoes the top-down target costingsystem work? Is it effective? Howdo you handle the gap between topdown and bottom up cost assess-ments?

3. How can you determine a newmodel profile? How do you deter-mine what level of satisfaction itwill generate?

4. How do the marketing inputs helpdefine the project targets? What arethe criteria?

5. How effective is the technology andmodule analysis? How does it helpin defining the new model? Whathappens if the bottom-up costs donot meet the top-down target? Howdo you bring the budget into finan-cial balance?

6. What do you do about new competi-tive introduction during develop-ment? Do you modify the model?Do you protect the budget?

Discussion of Teaching Questions

1. What do you consider the importantcharacteristics of a BMW? If youwere to appoint BMW’s next gen-eration 3-series project team leader,what basic guidelines would you setfor the team? Why are they impor-tant?

QFD (quality function deployment) isdirected at meeting customer require-ments. There is a basic difference be-tween developing new models, and con-tinuing a classic design or icon. The ba-sic 3, 5 and 7 series of BMW are wellknow styles in the marketplace, and mustcontinue their heritage if they wish tokeep their customers coming back. Forexample, if the Jeep was radicallychanged in style, it would no longer be aJeep. Many of the 1950s British sportscars have lost their heritage, as did Ford’sThunderbird and Mustang, and Nissan’s280Z. Maintaining a classic design al-lows for an extended life cycle that maygo well beyond normal products. BMWis attempting to build classic designs thatwill live for at least seven years. In sucha case, what are the design parameters?

A picture of the BMW 528i is placedon page 1 of the case to show BMW’slatest series design. The question posedto students should be, how similar or dif-ferent should the new 3 series look? Howmuch similarity should there be across theseries 3, 5 and 7. In what ways shouldthey differ? This should lead to a discus-sion of the basic characteristics that thedesign team must use? For example:

What characteristics make it a BMW?The “double kidney” grill is key. TheBMW logo is very distinctive. Thestyling is sleek and the raised bodyover the wheels is distinctive forBMW.What should be the differences be-tween a 3, 5 and 7 series? They pri-marily differ according to size, weight,and features. The 5 series is heavierand larger, making for more comfort-able driving at high speeds (in Ger-

•

•

1 Funding was provided by the Peter Curtius Foundation for USW.


many, it is the car for company salespeople). The 7 series is much largerand more expensive. There is about a$15,000 price difference between eachmodel.There can be open discussion aboutwhat car gets the latest technology.Since each series lasts for 7 years, thelatest model generally gets the latesttechnologies. However, that closes thegap between model features that canbe used for marketing. The latest se-ries tends to pull sales from older se-ries, unless new “sporty” or “special”features can be added to existing mod-els.Finally, response to competition andcustomer complaints is critical withnew models, since you will live withthem for seven years. The old 3 serieswas too small in the rear seat and be-tween seats, and lacked an elegant in-terior suitable for its high price andimage.

2. How would you evaluate BMW’sproject management tools? How doesthe top-down target costing systemwork? Is it effective?BMW’s new project management pro-

cess is systematic and is intended to in-crease the discipline of the project team.The easiest way to understand it is byviewing Figure 2 in the case.

The first step is to understand the mar-ket. Interviews provide input for Con-joint Analysis. A good example of Con-joint Analysis for consumer use is avail-able by accessing http://www.autotrader.com/ on the Internet andfollowing their Decision Guide throughthe decision-making process. You canthen identify many of the elements thatgo into the decision-making for car buy-ers. AutoConnect will help you find thecar(s) that best fits your decision criteria.

3. How do the marketing inputs helpdefine the project targets? What arethe criteria?

Market analysis and customer satis-faction inputs provide the criteria for im-proving on the existing product. Com-petitive benchmarks give a minimum re-quirement for a new series, plus specificcustomer inputs that come from customersurveys and satisfaction reports. This setsthe minimum standards for improvementof a new series. From this, the vehiclefeature assessment provides the specific

•

•

Figure 2: Integrated Market-oriented Top-downTarget Cost Management

targets for design and technical improve-ments. You can see from Figure 4 in thecase that overall safety and acoustic com-forts were especially poor in the old 3series.

You can also see that little effort wasplanned to improve design and road per-formance characteristics.

Finally, the market volume and pricemust be decided to provide a forecast fordetermining the overall revenue target forthe new series. Without this number, it isimpossible to generate a total cost target.

4. How do the marketing inputs helpdefine the project targets? What are thecriteria?

Since there are planning, develop-ment, and other Non Recurring Coststhat must be determined in addition to

Production Costs, the Targeting Cost-ing parameters must be determined by theproject development team (Figure 3 in thecase). Warranty and volume mixes arealso important in determining total costs.Because of the variations in volume andcost allocations, historical market data isuseful. BMW has a long history sellingits 3, 5 and 7 series and can determinerelative volume based on past sales num-bers.

In the final analysis, tradeoffs betweenvarious scenarios can be produced usingSensitivity Analysis shown in Figure 5 ofthe case.5. How effective is the technology and

module analysis? How does it helpin defining the new model? What hap-pens if the bottom-up costs do not

M


Figure 5: Sensitivity Analysis for Model Profitability

meet the top-down target? How doyou bring the budget into financialbalance?The ultimate cost structure will be

driven by the degree of changes made inthe new series. Every component and partthat is changed has some additional costelement. Any components that can be mi-grated from an existing series reduces thecost of change. The more desirable thenew series, the higher the sales volumeand ultimate profitability. A 1.0% in-crease in price will generate a 2.0% in-crease in profitability. A 1.0% increasein volume will generate a 0.5% increasein profitability. A 1.0% reduction in de-velopment costs will increase profitabil-ity by about 0.5%. A 1.0% reduction inproduction costs will improve profitabil-ity by over 1.0%.

Over 70% of costs are determined inthe design process. Depending on whattrade-offs are made in design, 30% of theremaining costs can be controlled by pro-duction. At the end of the design pro-cess, the deficits in achieving the targetcosting goals are put on the shoulders ofmanufacturing. This is what was donewith the 3 series.

6. How effective is the technology andmodule analysis? How does it helpin defining the new model? What hap-pens if the bottom-up costs do notmeet the top-down target? How doyou bring the budget into financialbalance?

The real cost analysis of the new se-ries must come from the technology as-sessment. This begins with an assessmentof the key spaces/modules (Figure 6) ofthe car, and then moves into the individualdiscussion of all 4000 unique parts (Table1).

The degree of change, planning cost,R&D requirement, and investment asso-ciated with each part change is assessedto provide the bottom-up cost factor. Itis also here that ultimate cost control willbe made. When the initial matching pro-cess begins, the design team must movefrom the “ideal” to the “reasonable” de-sign. For example, the engine used in the528i was used in the new 3 series. Thesuspension system, which was quite ad-equate in the old 3 series, was only modi-fied, rather than made new. You can loweroverall development costs by using moreparts from existing series. This requires

Figure 3: Target Cost Budgeting Model

Figure 4: Next Generation 3 Series Profile


continuous trade-offs, and continued re-view of the conjoint and competitiveanalysis to determine the impact.

7. What do you do about new competi-tive introduction during development?Do you modify the model? Do youprotect the budget?

Flexibility in design is critical in a dy-namic industry like automobiles. Whilefreezing the design is always critical forcost control, having a strategy for re-sponding to market changes is essentialfor long term competitiveness. The in-troduction of the Audi A4 and theMercedes modified C-class requiredsome reassessment. BMW engineers arefamiliar with most new designs, but havenot experienced the revised driving char-acteristics of competitor’s cars. It is criti-cal for the new 3 series maintain its repu-tation as the “Ultimate Driving Machine”.Team members were surprised at Audi’sdrivability. For a front-wheel drive car,it was better than they expected. Its priceand features were also better than the teamexpected, forcing team members to recon-sider what options to make standard forthe added price. Mercedes’ new C-classhad quad front lamps, which causedBMW’s design team to change the frontlamp design slightly, giving it a slightlysculpted look to its headlamps.

From a teaching perspective, the ques-tion is how many changes you can affordto make at such a late date. The cost ofnew releases can increase the cost by

Table 1: Technical Scoring Framework

25%. A new design costs as much as thefirst. Each decision impacts the ultimateprofit sensitivity. But a better competi-tor design could also reduce expectedsales, and thus profitability. Ask the stu-dents what risks they want to accept?Lower sales, or higher cost?

You can see the added cost of alterationsto an established model in Figure 6.

The figure shows how closely the new3 series looks like the current 5 series dis-played in the case. The latest develop-ment of the new 7 series also follows thisdesign theme, as Bangler continues todevelop a set of BMW icons. The newBMW X5 SUV also has the same frontend design. There is no mistaking theBMW for a Toyota.

Other Readings

Dr. Friedrich Nitschke, Market andProcess Oriented Cost Management forProduct Design in Automobile Manufac-turing. (Hamburg, Germany: Verlag Dr.Kovac) 1998. Dr. Nitschke used the 3-series project for his Ph.D. dissertationat Rostock University.

Figure 6: Cost Drivers in Developing the New 3 Series(million DM)


The New BMW 328i: GoodThings Come in Threes

I’ve always been a big fan of BMW’s best-selling 3 Series—the M3 is still the ve-hicle I’d buy tomorrow—for the cars’ per-formance, usability, and quality (not tomention manual gearbox). In fact, theonly things I’d fault the current genera-tion entry-level Bimmers on is conserva-tive lines, lackluster interior styling, and

a compactback seat. Buthey, with thiskind ofnimble han-

dling and racer-boy mentality camou-flaged in a well-built everyday driver,who’s complaining? Not me, since Idrove the next-generation 3 Series lastweek in Chicago. With the new 328i,BMW has fixed those problems. Polishedthem, in fact, to a high gloss.

On hand to articulate the understatedyet handsome changes was BMW’s headof design, Chris Bangle. An Americanwith the energy of a mad scientist and thedemeanor of a stand-up comic, Banglepointed to two features—the sweepingvertical cut lines that accent the rear por-tion of the passenger compartment andthe scalloped, eyelike headlight casings—as the keys to the redesign. The first waslargely aesthetic, while the second wasthe result of a 30% increase in headlightpower. “We raised the hood and dippeddown low under the lights to create aneye effect,” explained Bangle, magicmarker dancing over his sketch pad. “Thecompetition, however, does cartooneyes,” he continued, drawing a wide-eyedMickey-Mouse-Mercedes pair.

The other looks-enhancing change isthe car’s wider stance, which gives it acrouched, ready-for-action look—andeven more stability. And ready it is: Forseveral hours, I got to push the new 3through Wisconsin’s rambling, emptyback roads. The second I sat down, thenew interior styling wrapped around me,beckoning me to speed with its simple,elegant, driver-friendly controls, and aninteractive feel. All the buttons are at justthe right spot to hit without groping whiledriving, and the orange cascade panellighting gives you, as Bangle puts it, “awhole surface to relate to, not just aswarm of fireflies.” Exactly.

My favorite features were the excep-tionally sweet, easy-to-use five-speedgearbox and the four-wheel ventilateddisc brakes (the largest in class, up againstthe likes of the Mercedes C-Class and theAudi A4). The BMW also expanded itsinterior space from 86 to 91 cubic feet—more than the Mercedes and the Audi.The 328i has a new system, CBC (cor-nering brake control) that continueswhere ABS leaves off. Here’s how itworks—and I speak from experience:Say, for example, you’re screamingaround a blind curve in the middle of farmcountry, and suddenly you realize you’regoing too fast, as evidenced by the rearend of your car starting to swing aroundalongside you. When you brake, the CBCkicks in, automatically calculating thecurve’s degree of sharpness, and then dis-tributes brake force to the outside wheelsto stabilize the car.

Shall I go on? BMW increased bodystiffness by 50% and the front end’s en-ergy absorption by 80%. There are eightairbags, a programmable key that controls40 driver preferences, more low-endtorque, and a trunk rug that reverses torubber to accommodate muddy gear orhelp groceries stay in place while testingthe aforementioned braking system. HasBMW messed with its best-selling carline? You bet. And the results are fabu-lous. It’s good-looking, safe—and mostimportant, it offers up as much fast yetunintimidating driving fun as I’ve had ina while.Price: Base: $33,400 Loaded: $38,200Highlights: 0 to 60: 6.6 sec. Engine 190hp 2.8-liter inline six cylinder engine,multi-programmable ignition key, in-creased interior space, more extensiveoptions packages.Who’d Buy It: Anyone serious about acar’s performance, quality, ability tohandle real-world loads—and good looks.See the BMW website at www.bmw.com.

Review by Sue Zesiger, senior editor of FortuneMagazine, July 31 - August 6, 1998.

Documents

BMW’s 3-Series: Managing Platform Design and Development Costs