Table of Contents

Value creation in the pc industry

Cost structures and at their significance

Elasticity and its significance

Price movements and trends

The future

References

Value creation in the pc industry

The assignment is designed to discuss factors affecting the PC industry. While not dwelling on the history of the industry per se, it is useful to look at how companies in the industry have reacted (or not) to serious industry milestones. As such, one focus will be the rise, fall and rise again of Compaq Computer Corporation (Compaq) from 1983 to 1999. Compaq has been chosen as it is a good example of the effects of ignoring market demand on pricing and capabilities. The case of Compaq shows that to compete in today’s PC market, the focus of the entire company must be on cost and value.

As can be seen from the following charts, Compaq enjoyed meteoric growth from its creation in 1983 to 1990, and from 1993 onwards. It is intended to use the information in the GSM Analysis course to analyse why there was a lull during 1991 and the first half of 1992 in revenue and income growth, as this should be used as a warning for PC manufacturers in today’s market.

• Compatibility with the then IBM Industry Standard

• Portability through the original Compaq Portable

• Reputation for quality and reliability

• Strength in the corporate marketplace

The first and last points are significant - corporate customers saw Compaq as a means of moving away from the dominance of IBM, creating ease of substitutability, while retaining the ability to run software written for the industry standard. This meant that Compaq could justify a price premium through differentiation.

Reproduced from: Compaq Presentation (1995)

In the latter half of the 1990s, Compaq, Intel and others promoted a ‘new’ computing paradigm, based on choice of vendors at all layers of the model. These open systems would provide cost-effective computing solutions to the customers, and would remove the dominance of a handful of dominant, high-end computing companies.

In future, companies would focus on their areas of expertise, bounded by industry standards. As Bresnahan states, “The dominant firms earn rents on their superior abilities – be those technical, in marketing or in management”.[1]

The model above, and the feedback loop below suggest that it then becomes very difficult to compete in any segment where there is a clear dominant platform. For example, Novell tried to compete with Microsoft in the desktop applications segment, with a product that was seen as superior (WordPerfect), but without the marketing presence and market share that Microsoft enjoyed. This was an expensive mistake, culminating in the sale of WordPerfect to Corel, who would also struggle to compete with Microsoft.

Created From: Bresnahan, T.F. (1998) New Modes of Competition: Implications for the Future Structure of the Computer Industry [WWW} http://www.pff.org/microsoft/bresnahan.html 20^th November 2000

As Brenahan writes, “These virtuous cycles are associated with social scale economies, which means that there will be few platforms for any given segment. Yet the same forces also offer excellent opportunities for the sellers of a platform to erect barriers to entry.”

Cost structures and at their significance

Compaq built its reputation working with third parties referred to as 'resellers'. This allowed Compaq to focus on designing and manufacturing product, and creating demand in the corporate marketplace. The resellers provided local logistics and expertise to the customer.

This added a link to the value chain, some cost to the customer, but the benefits to Compaq were significant. Compaq and other manufacturers paid a percentage of sales as a marketing fund, allowing the reseller to develop more skills and perform local marketing. There was also a further cost – as emachines CEO Stephen A. Dukker states, “Compaq and others spend more than 10% of revenues to repay retailers for price cuts and take back unsold inventory.”[2] Significantly, “PCs lose about 1% of their value every week they stay on the shelf”.[3]

The cost model is to set a reseller buy price (RBP) that Compaq would be paid, with the reseller selling out at a price agreed with the customer. The reseller retains the difference as margin, and Compaq would use the RBP to cover manufacturing, research & development, admin costs and profit.

As emachines Stephen A. Dukker says, “Dell showed that it could be successful at a 25% gross margin rather than at Compaq’s 50%-plus”.[4] Emachines work on the basis of 10 to 12% gross margin. As this margin decreases through competition, and the value of PCs falls, providing a lower absolute gross margin, traditional manufacturers must look to diversify and add more value to maintain the absolute value of profits.

As Michael Dell claims, “our company has about 55% of the profitability of the entire PC industry…This does not come from overcharging our customers, but by having the lowest cost structure and the most efficient business model.”[5] But, as a word of warning, there is an opinion that Dell’s way of doing business can be copied or adapted.[6] Dell has also strongly resisted the move to the sub-$1000 segment, where different cost structures and business models may be required.

In the period up to 1991, Compaq believed that they could charge a 10% price premium over market rates and still the corporate customers would continue to purchase. This was true, but as will be explained, this was a near-monopoly position that could not be defended.

Other manufacturers, most notably Dell, chose a direct model of distribution. In this model, Dell provides remote logistics, and provides support through local service providers. This removed the cost of the reseller, but Dell absorbs more cost of sales directly. As Michael Dell notes, “This gives us both a cost advantage by eliminating the cost of the dealer and by enabling us to integrate everything in a one-step process.”[7]

In 1993, the home market started to develop into a serious opportunity. The traditional resellers focused on the corporate marketplace, and did not want to change their business models to address the home market. The avenue to market is controlled by major names such as Dixons, PC World, John Lewis, Argos etc.

Indirect manufacturers such as Compaq had been used to setting reseller buy prices, and the quality of product and service ensured that customers would continue to purchase. Moreover, Compaq had become used to delivering product when it was available, rather than necessarily when the reseller or customer requested. In entering the retail market the cost structure and business approach has to be radically changed.

For example, in 1993, a major High Street outlet was in a strong enough position to dictate to Compaq the margin that they would receive on product - c. 45%. Compaq was forced to sell to a major High Street outlet at anticipated sell-out price less 45%. Another major difference is that because of the volume of suppliers, the major High Street outlet issue a delivery timetable - if the delivery lorry does not arrive at the main depot at a certain time on a certain day, that delivery is missed until the next scheduled slot.

Making better use of the existing manufacturing facilities has helped manufacturers bear the price cuts. For example, Compaq achieved manufacturing Economies of Scale in Houston manufacturing by moving from 60 hours per week in 1993 to 24 hours per day manufacturing from 1994. This allowed a better ROI to be achieved on existing capital equipment and premises.

Elasticity and its significance

For simplicity, I have assumed a linear demand curve, although this is not likely to be the case in the PC industry. For example, when Intel reported a profits slide in 1998, despite a move to lower price points for its processors, the elasticity of the market was questioned – as LaFountain of Needham & Co. asked of Intel: “If the move to lower price points was so widespread, why aren’t we seeing greater unit growth?”[8]

As has been stated, until 1992 Compaq could charge a 10% price premium on the market price. This caused a problem, as Compaq CEO Eckhard Pfeiffer explained – “Sticking to a high-performance strategy was reducing us to a niche player. You don’t get volume (this way), so we decided to refocus our PC effort.”[9]

It is fair to assert that the PC should be classed as a ‘normal’ good, that is, an increase in income results in an increase in demand. The other feature of the PC market is that there is a long refresh rate. Typically, based on experience, organisations will replace PCs every three years, but the home market may exhibit a longer refresh cycle.

So, if Compaq was able to charge a 10% premium on the market price, surely they would be earning supernormal profits of 10% above their ordinary profit, particularly as the market was dominated by IBM and Compaq? Unfortunately, this was not the case.

The absence of serious competition through the late 1980s, had removed the competitive pressures of a free market that promote efficient, profit-maximising behaviour.

In order to maintain the 10% price premium, Compaq engineers and marketers had engineered more and more capability into revisions of product. The methodology was to create a superior product, work out how much it had cost to manufacture, then sell it for the market price plus premium. This ‘engineering to ability’ rather than ‘engineering to cost’ led to superior but cost-prohibitive product, especially by late 1991. As an example, the steel lids of PCs were specified as having 90^o corners – lid manufacturers would ask 90^o plus or minus what? The response was plus or minus zero, adding tens of dollars onto the price of a lid. Other manufacturers would accept a tolerance, resulting in reduced costs and sell price to the customer. This technique was clearly not technically efficient and represented inferior economic efficiency, as it was not producing at least cost.

The desire to control the desktop attracted a large number of aggressive competitors, turning a difficult business situation into a fight for survival.

The turning point came when Compaq realised that it could not justify a price premium – signified by the fall in net income and revenue in 1991. A new approach was adopted, setting a maximum engineering cost per product, and using this to improve purchasing agreements with component suppliers with anticipated higher volumes as the carrot. As Compaq’s Ross Cooley said, “value engineering is in and over-engineering is out.”[10]

Compaq, DELL, IBM and the other market leaders moved to a system of Just In Time (JIT) manufacturing, reducing short-run production costs through reduced handling, and stock-holding. As IBM’s Bob Moffat indicates ”Dell touches a box 18 times in its facility. We touch it 11 times. Fewer touches equal faster delivery of build-to-order boxes.”[11]

Gone are the days when Compaq and IBM would provide 30-day ‘price-protection’ to the resellers, meaning that if the price of stock fell they would be compensated. This has been reduced to 14 days. For IBM, “that change brought roughly $400 million in additional revenue in the first quarter.”[12]

According to 4^th Wave, we are entering “the fifth era of computing.”[13] Fourth Wave talks about the an information appliance, shaped by two forces – “free MIPS (Millions of Instructions per Second) and Pervasive Connectivity”.

4^th Wave Summarise the five eras as:

Reproduced from:  4^th Wave (1999) Information Appliance Market Report [WWW] http://www.fourthwave.com/report/execsum.htm 20/11/00

4^th Wave has performed a study and designed a model to show a GDP/Price Elasticity relationship. They indicate that Information Appliances may have to be priced at the $70 price point to achieve penetration and profitability. This has a significant effect on an industry where the entry level price is currently around $600. 

Price movements and trends

In 1981, the average product life cycle was 5 years. By 1999, the average product lifecycle had reduced to 6 months.[14] This has a significant effect on research and development, as any investment has to be recovered over a far shorter timescale, resulting in higher prices or increased volume expectation. In 1999, IBM, HP, Compaq and Dell spent a total of almost $10 billion on research and development. Using Dell as the base point, Compaq spent 4.4X, HP spent 6.5X and IBM spent 14X. While this does not exactly make Dell a ‘Free Rider’, it is clear that Dell gains commercial advantage from the development undertaken by the others.

Since 1992, the PC industry has delivered price cuts of approximately 23-30% per annum. The market demand in the USA as an example has increased from 48 million units in 1993 to 70 million units in 1999. Despite the growth, the major manufacturers are being attacked by:

• a number of Japanese manufacturers

• the convergence of technologies (e.g. using television or WAP phones for internet access)

In the business market, price cuts have tended to lead to customers spending the equivalent amount, but purchasing more PCs. This market is almost saturated, and future revenue is based on the customers refresh cycle. While some customers show loyalty, as the PC has become commoditised, the PC manufacturers suffer from the substitution effect. As all PCs give similar performance and functionality, buyers have the ability to look for the best deal at that time, resulting in frequent switching between manufacturers, with relatively low switching costs.

The home market represents a significant growth market for PCs. This is a notoriously price sensitive market – the key turning point was delivering a sub-£1000 PC. The home market requires products to be positioned at price points. Look into the local PC World or Dixons and the price tags will be set at £799, £999, £1199, £1399 etc. The reasoning is that this market is prepared to spend an amount of money on the purchase, but expects to gain maximum value at that price point.

The emergence of new competitors at lower price points creates a continuing challenge for the traditional PC manufacturers. Stephen Dukker has benefited from creating a company, emachines, focusing on the sub-$600 PC. Dukker uses the same argument that Compaq used against IBM in the 80s and 90s – “In the past, PC makers were very careful not to sell low-priced boxes that could cannabilize other products, or if they did, they wouldn’t build very many of them.”[15]

The future

Moves to extend the reach of the market to TV set-top boxes, internet-attach devices, and personal data assistants (PDAs) are all attempts to widen the scope for product demand and supply, ensuring a growth opportunity for the future.

Manufacturers are now partnering with internet content providers and portals to ensure that they are not excluded from the information age. But, in response to Microsoft’s overtures to the cable companies, “the cable industry realises that the monopoly rents it has been collecting for years could soon multiply in value; and it is not about to hand them over to somebody else.”[16]

How long before the zero-price PC? The price trend for the PC is towards zero, but how can the manufacturers survive? The answer is to partner with companies, for example, free PC in return for:

• annual subscription to an internet portal

• committing to receive electricity from an electricity provider for a period of time

• accepting advertising from a number of companies on a regular basis

• participating in market research

• purchasing servers, networks and services from the manufacturer

• providing commercial computing power when the PC is idle, as demonstrated by the SETI@home experiment

It is fair to assume that a move to free PCs will see further rationalisation in the industry, and that this will threaten PC-only manufacturers. Hence the rush to move to information and services from hardware.

For the users of IT, “the Information Age is largely a world of high fixed and low marginal costs.”[17]It costs large amounts of money to set up an information-based business, but delivering new products and services can be produced cost-effectively thereafter. Moreover, Cox argues that whereas previous goods and services were rivalrous, depleting with consumption, Information Age products and services will continue to be available to everyone after individual consumption.

The PC industry is moving to an Information Industry, driven by relatively new company CEOs – Gerstner at IBM, Fiorina at HP, Capellas at Compaq. Companies that have grown through manufacturing PCs now have to find an economically viable means of making normal profits, through service and added-value. Each must remove years of inward focus and drive the companies to different ways of doing business. Witness Fiorina’s intention to reorganise HP into co-operative businesses and ride the second wave of the internet through Applications Services Provision and the creation of online market exchange mechanisms.[18] For HP, this has led to 15% revenue growth despite serious competition in its UNIX heartland.

For manufacturers already associated with the Internet, such as Sun Microsystems, revenue increased 42% in August 2000. Its biggest threat is probably an attack from the Linux operating system on its core UNIX offerings.[19]

Compaq is finally coming to terms with the 1998 acquisition of Digital to extend its reach in the services/one-stop solution provider segment. After more than $2 billion losses since the merger, Compaq gained Digital’s cash, Digital’s tax carry-forwards and proceeds from the sale of AltaVista approximating $2.5 billion. This equates to a neutral position on the purchase price of $9 billion.[20] As it struggles to reconcile loyalty to the channel with the need to sell direct, whether the combined company can make the transition to the Information Age remains to be seen.

The PC manufacturers grew in the late 1990s with the message that the best customer solution is ‘choice’ – choice of manufacturer, operating system, applications, network and support supplier. How ironic that the industry is now maturing and attempting to provide closer to the whole direct solution in the manner of an IBM or DEC of old.

References