Carmakers have good reason to focus their efforts in the 1990s on distribution. Thanks to fundamental changes in the role of automotive distribution, the existing system has for a long time not met the market’s needs. The result: industry stress at both dealer and OEM¹ level, coupled with notable dissatisfaction among customers.

One way to resolve this conflict is to optimize the current system. Long-term success, however, will require a more radical approach, centered on unbundling and rescaling the four or five key functions within the overall distribution system. Deciding to pursue the unbundling path represents a major strategic choice for any carmaker, but there are steps it can take to ease what will be a very difficult transition.

Weighing costs and benefits

There is much to be gained by improving the efficiency and effectiveness of the channels that move cars from manufacturers to consumers. True, there are plenty of reasons not to open this "Pandora’s box." First, the system seems, at least on the surface, to be working well. Moreover, other OEM challenges—such as taking costs out of manufacturing and accelerating product development cycles—have been more pressing in this period of tough competition and considerable overcapacity. Finally, the system has been awkward at best to tinker with.

Dealership principals tend to be independent entrepreneurs with strong views, and any manufacturer’s efforts to improve efficiency or effectiveness have to work through thousands of distribution points. This area is thus a lot "messier" to work with than, say, one’s own manufacturing facilities, where the managers work for you rather than for themselves, and where the issue is confined to a few dozen locations.

There are even more reasons, however, to take action to reform the typical automotive distribution channel. An enormous amount of money is, after all, tied up in distribution: dealer margins and incentives, advertising allowances, floor plan financing, and numerous other charges can add up to 30 percent of the cost of a new car (Exhibit 1). Small slips in this area can ruin years of hard work in other parts of the business. An error in positioning downstream inventory in the delivery system might cost, in extra carrying charges, much more than the savings from a year-long value engineering project in product design.

Furthermore, as other parts of the carmaker’s business system are being overhauled to squeeze out excess cost, the opportunities to increase distribution efficiency become more salient—as is evident when one compares the relatively slow rate of improvement in carmakers’ sales productivity with the rapid rise in their manufacturing productivity (Exhibit 2).

Buying decisions

Distribution reform obviously affects not only the cost side of the profit equation, but revenue too: our research reveals that a consumer’s sales and service experience with a dealer can influence his or her decision to purchase the same make next time by as much as 40 percent. For luxury cars, the figure is 60 percent or more. As all carmakers improve their product quality to much the same standard and incorporate similar technologies and features into their vehicles, the degree of dealer influence will probably climb still higher. Such trends weaken the product’s ability to differentiate the brand, thereby increasing the burden on the channel to do so.

Thus downstream distribution performance can today offer greater leverage, on revenues as well as costs, than the upstream areas of design and production that were the focus of carmakers’ efforts in the 1980s.

Why is this true? Why has distribution become the next best place for carmakers to direct their competitive energies? The underlying cause is that the typical Western car distribution system has not evolved to keep up with changes in the commercial realities of the car business—changes so slow that industry participants may not even have noticed them taking place, but so sweeping that they are fundamentally "changing the game," making the current distribution system effectively obsolete.

The lost rationale

The original reasons for the existence of the current distribution system have largely disappeared

The original reasons for the existence of the current distribution system have largely disappeared, but the system itself has yet to fully react. A brief glance at the history of automotive distribution makes this clear.

The current distribution system is essentially one in which OEMs sell vehicles to numerous independent franchisees that preferably deal exclusively with one OEM and usually operate a fully-integrated, standalone location handling new cars, used cars, parts, service, and some financing. As the automotive industry evolved between the 1930s and the 1950s, this original system developed alongside it in response to its needs. The result was a distribution function characterized by five key features:

1. Dealerships were the favored channel of the OEMs, since there were no economic direct-distribution options (partly due to poor logistical infrastructures), and inexperienced customers needed a lot of information before buying—information that could best be provided by a local dealer.

2. Independent dealers (dealerships owned by dealer principals) were preferred by OEMs, as setting up an entirely company-owned system would involve massive amounts of capital.

3. Numerous outlets in the hundreds and then thousands were necessary, partly because customers trusted local merchants more than national companies, but mostly because cars broke down so often that many local repair points were needed, and because populations were widely dispersed across large areas.

4. Single-branding: dealer exclusivity was usually enforceable by OEMs, given that a dealer could thrive on the volume from just one make. Motoring was spreading rapidly, demand for cars was high, and competition was often mild.

5. Full integration of functions within a single dealership made sense because each function (new-car sales, used-car sales, parts sales, and repair service) was profitable at the scale of an individual dealer. Volumes of business were sufficient (thanks, for instance, to the need for frequent repairs) and fixed costs low enough to give relatively small dealerships good returns.

These five features, so critical in the early development of the industry, no longer apply today, due to new process and product technologies, as well as changes in both the supply and demand sides of the automotive market. Taking each feature again in turn:

Direct shipment of a customer-ordered car is within striking distance for OEMs

More and more purchase decisions are made before the customer reaches the dealer

Distribution methods more direct than franchised dealerships are now conceivable. Direct shipment of a customer-ordered car is within striking distance for OEMs as their "build to order" capabilities improve and as transport logistics become more reliable. In turn, customers are now much better informed about car choices, thanks to better information technology and increasing consumer sophistication. As a result, more and more purchase decisions are made before the customer reaches the dealer, thereby reducing the latter’s role as a provider of information.

OEM-owned distribution is still inherently costly, but becoming cheaper, both in absolute (if direct shipment can be used and expensive real estate and inventory investments avoided) and in relative terms. (As dealerships’ interests diverge from OEMs’, control problems arise—as when, for instance, a dealer decides to shift sales emphasis from new cars to higher-margin used vehicles. OEMs must often lay out large monetary inducements to dealers to change their behavior, whereas OEM employees could simply be directed to make a change.) Even when ownership is not feasible, OEMs are taking steps to align dealers’ interests with their own, such as forming more highly-integrated dealer councils.

Fewer outlets are now needed to sell and provide services for a given volume of cars. In sales, now that national brands have proved their quality, they no longer need a local dealer’s credibility to survive. (In fact, dealers are often seen as less trustworthy than an OEM.) As for servicing, customers will accept a longer but less frequent journey to the nearest repair facility now that cars are so much more reliable (Exhibit 3). Finally, population shifts from the countryside to cities and suburbs has made many rural dealer points economically redundant.

Multi-branded dealerships have become the norm in the United States and are starting to gain ground in Europe. As brand proliferation and competition increased in the 1980s and 1990s and as volume per make and margin per new car declined, dealers became desperate for better profitability at the new car level. In the attempt both to preserve total profit and to limit the risk of decline in any one brand, dealers began to hold a "portfolio" of makes. Other factors influencing this change are the extension of car ownership cycles (due to older cars’ increasing reliability, and to new cars’ declining affordability), diminishing brand loyalty among customers, and a demand for the convenience of choosing from several brands in one location, as in other kinds of shopping.

The "de-integration" (or unbundling) of dealership functions has become viable in cases where the scale of efficient operations has increased beyond the size of a typical single dealership. Particularly in parts and servicing, there is now a need to pool activities from several dealers’ catchment areas in one location.

There are three main factors contributing to this trend. First, diagnostic and test equipment for car repair is becoming too costly for small dealers to afford, and the proliferation of service parts has also increased their working capital requirements. Second, high-volume repair shops have achieved prices below dealer-profitable levels through specializing in, for example, brakes to achieve scale economies. Their low prices have won them market share from traditional garages and dealerships. Finally, the volume of repair business generated by each new car sale has fallen as cars improve in quality, rendering much of dealers’ parts and service capacity redundant and uneconomic.

Despite the existence of such counter-trends as OEMs’ extension of vehicle warranties, which drives more service and parts business back to dealerships, and the tendency for the average repair to increase in price, which may keep total revenue to the dealer high, routine maintenance will come to represent a dwindling source of income for dealers (Exhibit 4).

A further setback for dealers is the migration away of finance and insurance services from the local scene (where the dealer would refer a buyer to a local bank for a significant fee) to the national level (where the dealer refers to the OEM’s captive finance company). Local institutions cannot match the OEM’s scale and range of financial offerings, and thus another traditional source of dealership income has been eroded.

The stress in the system

Now that the existing distribution system no longer meets the industry’s economic needs, the result is stress at all levels.

Dealers have already seen margins on new cars evaporate and have lost parts and service business to independent outlets

Dealers have already seen margins on new cars evaporate (especially in down cycles) under the pressure of competition, and have lost parts and service business to independent outlets. This makes it difficult for them to afford the investment they need to upgrade their operations. In the attempt to recover profitability, many have turned to multiple brands and multiple locations. Meanwhile, bankruptcy rates for American dealerships are at historical highs.

Many OEMs are unhappy too. They have to spend enormous sums on distribution just to keep up with the competition. Many feel they are losing control of their networks: multi-branding threatens their once exclusive relationships with dealers; rationalizing sales points is slow and subject to delay; and even vehicle pricing is vulnerable to tremendous pressure from dealers keen to attract customers into their showrooms.

Customers are frequently unsure whether they got a fair deal on a new car, annoyed at delays in repairs, and concerned about used-car quality

Finally, customers are frequently dissatisfied with the present system—unsure whether they got a fair deal when buying a new car, annoyed at delays in repairs, and concerned about used-car quality. Despite massive efforts by both OEMs and dealers, few commercial activities generate less customer satisfaction than visiting an automotive dealer.

OEM, channel, and customer alike are discontented; all are looking for new answers. Such temporary fixes as rebates or promotions, often employed during market downturns when the stress is greatest, will not suffice; the roots of the problem are structural, not cyclical, and will persist long after the recovery of any market. Tinkering at the margin or waiting for sales volumes to return will not do. Fundamental problems need fundamental solutions.

The optimization route

Optimizing the current system is a necessary step—but only the first one

One approach to the distribution issue is to overhaul and optimize the current system. This is a necessary step—but only the first one, and not enough in itself to resolve the problem in the long term. Squeezing greater efficiency from the existing system and increasing its effectiveness as much as possible are already on the agenda of many OEMs in Europe and the United States. Their initiatives include:

Assisting dealers to improve their profitability so as to forestall multi-branding or loss of dealer interest. Manufacturers are sending their own personnel out into the field to advise dealers on best practice in sales, parts, and service. The problem is the manufacturers themselves lack the experience in dealership operations to make this advice consistently helpful.

Most OEMs have announced aggressive delivery targets

Reducing delivery cycles by increasing the speed and responsiveness of the manufacturing process, with a view to raising customer satisfaction and lowering system inventories. Most OEMs have announced aggressive delivery targets: for example, Chrysler in the United States is aiming at a total time of 20 days from customer special order to vehicle receipt.

Launching campaigns to increase customer satisfaction and loyalty, including overhauling parts networks to improve delivery response and on-shelf availability, experimenting with "no haggle, no hassle" single-pricing schemes, installing "customer service for life" programs, requiring dealers to invest in customer service assets to qualify for new model allocations, and paying bonuses based on dealers’ customer satisfaction ratings. Other initiatives aim to build loyalty directly to the manufacturer. Examples here include affinity charge cards like that of General Motors, Honda’s offer of tour packages for new car buyers in Japan, and—in the most extreme example—Toyota’s campsites for the use of purchasers of its sport-utility vehicles.

Adopting leasing programs (in the US at least) both to bind the customer more tightly to the manufacturer—or dealership—and to make the flow of sales more predictable.

Aligning OEM and dealer interests through the establishment of more truly democratic dealer/OEM "councils" and advisory boards, and revising franchise agreements where possible.

Trimming excess dealer points by purchasing and closing dealerships when they come onto the market and de-listing dealerships that do not meet standards of customer satisfaction. Thanks partly to change in the economics of distribution and partly to geographic shifts in population, as many as 10,000 sales points in the United States may be superfluous, according to OEM calculations, and the story is similar in Europe. Some OEMs are pushing this process aggressively; others are content to let it develop at its own pace.

Structural impediments

One flaw in the "optimizing" approach is that diminishing returns set in

Despite the local and short-term successes that these efforts may achieve, they cannot ultimately overcome the structural problems within the distribution system. One flaw in the "optimizing" approach is that as more and more OEMs adopt initiatives like those described above, diminishing returns set in.

Customer loyalty initiatives, in particular, may be doomed to low rates of success, since to increase customer loyalty within the current system necessitates spending more and more effort on a weaker and weaker leverage point. Certain models of car already achieve such high homogeneous scores on the customer satisfaction index in the United States that incremental spending designed to boost levels still further is likely to achieve little benefit. Also, it is hard to improve customer satisfaction when the customer’s car is so reliable she rarely needs to visit the dealership at all!

Another drawback with optimizing efforts is that they are generally difficult and slow to implement. OEMs pursuing them may be overtaken by firms taking more revolutionary paths. The trimming of surplus dealer points in the United States, for instance, has moved at a pace that can only be called glacial. Moreover, the fundamental change in attitude and behavior that most such programs require has not yet spread sufficiently widely among either dealers or OEM channel-management personnel. Dealer resistance is far from being the only obstacle; the OEMs themselves must reverse years of "push the volume" tactics if they are to convince dealers that they genuinely want to reform the system.

All this is not to deny that there is plenty of room for improvement in the current system. It certainly makes sense for every OEM and dealer to scrutinize the merits of "optimizing" programs. They should, however, keep careful watch on the real payoffs of these initiatives, which are likely to vary by OEM, country, and even market segment (for example, basic versus luxury).

Changing the game

If optimizing the current system is not enough to maximize competitive advantage, what are the options for changing the game entirely?

New channels will probably share a common tendency to unbundle or de-integrate functions

While it is likely that a range of new channels for automotive sales and service distribution will eventually emerge, most of these solutions will probably share a common tendency to unbundle or de-integrate functions. The goal will be to take each function separately and design it in the best possible way from scratch, rather than to try and improve each function as it exists in the current fully-integrated structure.

Considered in the light of this goal, many of the primary units in the current system—the dealerships—are too small in scale and too fragmented in their market impact and management focus to be economically sustainable. Unbundling dealership functions, then carefully rebuilding and refocusing them, is the only way to overcome this problem. Such a program will:

• achieve economies of scale (for instance, by gathering the new-car sales volume of several dealerships into one larger franchise to pool and thus reduce sales overheads);
• achieve efficiencies of focus (by building local expertise in a single specialty, such as used-car sales, for instance); and
• develop better marketing effectiveness (or better matching of services to market needs) rather than offering every function at every dealer point regardless of demand (not every sales location, for instance, needs its own repair service).

Functional unbundling is not, however, the same as "going direct," although direct sales of cars by manufacturers can be split off from other distribution functions as part of the process. Nor does unbundling necessarily entail unbundled ownership: there is nothing to prevent a dealership, for example, from splitting its business into dispersed functional pieces, while the same individual retains ownership of all of them. Exhibit 5 illustrates two conceptual paths of unbundling functions involving multiple and single ownership.

How to unbundle

The process outlined above is already taking place on a trial basis at various locations around the world. Some examples of feasible unbundling initiatives are listed below. They are not recommendations, since a particular OEM would need its own tailored program; nor do they need to be executed concurrently (one might, for example, divorce car sales from service, but keep new- and used-car sales under one roof).

New-car sales

For greater efficiency and effectiveness, new-car sales can be split off from other activities in a number of ways. First, OEMs could sell more directly to car purchasers with less intervention from the sales channel—provided that customers can be reassured that parts and service support will be available and reliable. The possibilities include:

• An OEM with a new low-end product in a market like the US distributing its cars through discount merchants such as Sam’s Clubs, with service provided by a national chain of independent garages such as Sears. Savings derive from using existing retail facilities with minimal overhead.

  An OEM with a new low-end product in a market like the United States might distribute its cars through discount merchants such as Sam’s Clubs

• An OEM with high-speed flexible manufacturing delivering direct to the consumer (as some fleet sales operate already), eliminating local facilities and inventory cost.
• A manufacturer entering a market by selling direct to customers, while contracting with dealers of other brands (or using independent specialist chains) to provide service support.
• Using dealers as intermediaries, but cutting their inventories back to a few cars for display only and delivering vehicles directly to customers from regional car compounds.
• Operating through directly-owned dealerships, as occurs in Germany and elsewhere.

Steps toward more direct sales are already evident worldwide in fleet sales, discount sales to OEM employees, buying-group sales, and sales through highly specialized dealers. One dealer in central Ohio, for instance, sells only cut-price Corvettes, by phone, to buyers in the Midwest; he acts essentially as a low-cost broker, and services few of the vehicles he sells. Some dealerships, too, are experimenting with more direct links to customers, calling on potential buyers at their homes, or using non-employees as commission agents in their own workplaces.

Consumers rate visiting a car dealer as even less pleasurable than a dental appointment

Many customers would, of course, balk at direct approaches, fearing to spend large sums of money on an item where no "store"—or only a distant one—exists to back up the purchase. OEMs, in their turn, might worry that their brand image would be eroded by direct sales. However, given that consumers rate visiting a car dealer as even less pleasurable than a dental appointment (Exhibit 6), manufacturers should perhaps ask themselves how much they have to lose by cutting poorer-performance dealers out of the loop.

Today there are 165 multi-brand auto malls in the US handling 7 percent of the country’s vehicle volume

Auto malls represent another non-traditional approach to new-car sales. In a mall, a group of dealers pool their new-car sales departments into one large location to capture scale economies. The auto mall is another manifestation of multi-brand shopping; customers who are used to shopping in department stores that display many competing brands are beginning to look for the same kind of experience when buying a car. From the dealer’s point of view, the risk of losing a buyer to a different brand in the mall is apparently counteracted by the increased traffic passing through each dealer’s section. Today there are 165 multi-brand auto malls in the US (up from only 70 in 1985), handling 7 percent of the country’s vehicle volume.

Activities that support new-car sales, such as finance, insurance, and extended service plans, can also benefit from unbundling. Financing can certainly be provided over the phone, in the same way that American banks already offer instant over-the-phone loan qualification, and automotive insurance in both the United States and the United Kingdom is now commonly executed by phone.

Overall, the unbundling of new-car sales and its related finance and insurance business could bring costs down through the elimination of much local (dealer-driven) overhead, reductions in notably inefficient local and regional advertising, the achievement of greater scale economies in such activities as financing, and cuts in local inventory carrying costs. Direct-sales unbundling options offer OEMs greater control of car sales, and also reduce "frictional" losses (by, for instance, limiting the game playing that usually takes place through holdbacks, special dealer incentives, ad allowances, margins, and so on). Such initiatives do, however, entail major investment in skills that OEMs do not necessarily have at present—a requirement that represents a significant barrier to implementation.

Used-car sales

Unbundling sales of used cars would allow specialists to target the unique needs of this market. Facilities could be shared by numerous dealers so as to build enough volume to support the hiring of top used-car valuers, who are responsible for driving most of the profit in this field.

At least one used-car specialist with a new format is operating in the United States today: the consumer electronics firm Circuit City has several 300-used-car lots employing standardized valuation and repair methodologies and no-haggle pricing. Other used-car dealers are setting up automated free-call telephone numbers that customers can use to find out which vehicles are currently available on their lots, significantly cutting advertising expense.

Low trade-in valuation is a major source of customers’ dissatisfaction with new cars

Separating new and used (trade-in) car transactions would also relieve OEMs of a headache: the fact that low trade-in valuation is a major source of customers’ dissatisfaction with new cars, creating widespread mistrust of dealers. Specialist used-car firms might be able to overcome this problem if buyers realize that their only business is used cars, that their interest is in satisfying customers to encourage repeat purchases, and that they are not playing games with trade-in and new-car prices. Dealers that did not participate in such specialist used-car lots would complain, of course, since they would lose the handsome margins they generate from used-car sales.

Service and parts

To achieve the best scale, service and parts departments can be split off from sales points and combined with other dealers’ or brands’ service facilities. Larger multi-brand (or, in big metropolitan areas, shared single-brand) service facilities can afford all the test equipment they need (not easy for many average dealerships today), and can focus on the key factors of the business (for example, technician utilization and inventory optimization). In spare parts, except for centrally controlled inventories like those at Lexus, it is rare to find a dealership with sufficiently broad and shallow inventories to execute rapid fills at low cost. There is a clear need for regional parts depots: some US dealerships today have virtually abandoned new-car sales in favor of wholesaling parts to other dealers whose own inventories are inadequate.

Both the decline in repair work (as cars become more reliable) and the rise of independent national chains of repair shops testify to the importance of restructuring parts and servicing. The repair chains vary by country: in Italy, for example, they take the form of small, fragmented "Mom and Pop" operations. OEMs would, however, be well advised to note the high volume and deep specialization typical of the successful independent repair chains, and they should ask themselves if they can realistically match either through single-

dealership repair shops. A recent study by Consumer’s Union in the US reports that dealership repairs are virtually always significantly more expensive than those performed by independent garages.

It is worth noting that, although after-sales parts and service work has a dramatic impact on customer satisfaction, satisfaction itself is relative. Through diligent efforts, an OEM may reduce the customer’s dissatisfaction with the service experience. But the fact remains that the customer is dissatisfied about having to undergo it at all.

The ideas above are feasible in today’s automotive industry. However, restructuring the distribution system, as opposed to "tuning it up," is a bold decision that requires careful consideration. Many OEMs today are avoiding the issue, citing the intractability of dealer principals or the strength of local franchise protection laws as their excuse. The transition to a new system will indeed be very difficult, but the advantages accruing to the first mover in this regard should be great enough to make the struggle worth it.