Issues Specific to E-Commerce

We have chosen to highlight three issues that distinguish the development of e-commerce strategies. They are time compression, importance of externalities and the presence of partners (or the multi-firm view). Time compression has to do with organizations receiving instant feedback from the "market." Success or failure is very visible. Externalities are important in e-commerce because strategic alignment depends not so much on aligning e-commerce strategy with internal organizational and technology strategy as much as being in sync with the strategies and technology investments of partnering organizations. Lastly, the presence of multiple stakeholders reduces the degree of control an organization has on its e-commerce investments. The implication is that competitive advantage has to be balanced with collaborative actions.

Developing an IS strategy is never straightforward. Strategy ^[5] generally connotes a long-term commitment. In addition, strategy also implies commitment to a broad direction—turning away from which is generally very costly. Like organizational strategies, it also takes significant time and resources to formulate an IS strategy. Traditionally, IS strategies have been deployed in order to consolidate investments in diverse applications and technologies. IS-strategy development marks an organizational shift in viewing IS as an operational necessity to reaching for strategic benefits. This formative process is similar to a learning process—whereby a collective insight is generated and assimilated over time. Once such insights are able to provide direction, options and clarity of purpose, they can be documented as a strategy.

The preceding description of IS strategy formulation connotes a deliberate and slow process. This is the first point of departure for e-commerce strategy. An e-commerce strategy receives almost instantaneous feedback after it is deployed ^[6]. If the e-commerce strategy is successful, the fast growth or ramp-up is almost always exponential in terms of transactions and/or workload. This is almost always associated with the need to sharply reduce response or reaction times. This is captured in Table 2 showing the need for scalability. The main implication for e-commerce strategies is that "pure strategies" may not be possible to develop given the short time to react. In other words, e-commerce presents organizations with an opportunity to mass-customize selected products or services.

Table 2: Keys to E-Commerce Scalability: Three Success Stories (Scheier, R. L., 2001).

AVIS GROUP HOLDINGS	HSN.COM (HOME SHOPPING NETWORK)	eBLAST VENTURES
Challenge: Web-enable mainframe systems without rewriting them.	Challenge: Scale from 325,000 unique visitors in January 2000 to 2.3 million in December.	Challenge: Scale from one server to 50 servers and from 50MB to 9TB of managed data in one year.
Strategy: Find crucial data and business rules in current applications and build clean interfaces between them and Web applications.	Strategy: Stick as closely as possible to a tiered architecture; keep to single vendors for server software and hardware, for maximum compatibility. Limit the number of servers to 10 per site to ease data management.	Strategy: Follow a strict development process, and define and enforce strict rules for defining data so common functions can be used across applications.
Technology: Professional services from Rockville, Md.-based Merant PLC, as well as Merant's PVCS Version Manager and Micro Focus mainframe access and development tools.	Technology: Compaq Computer Corp. Web and database servers; Microsoft SQL Server 2000 and Windows 2000 Advanced Server.	Technology: The Rational Unified Process from Cupertino, Calif.-based Rational Software Corp.; tools from Rational such as Rational Rose for application modeling and Rational ClearCase for software configuration management; XML as a translation layer between mainframe and Web applications; BEA Systems'. WebLogic development and management tools
Advice: "Design big and build small."	Advice: "Keep it simple, stupid."	Advice: "By breaking your business into three layers, you can add hardware against each one" as needed.

While Table 2 described a technical challenge (achieving scalability in quick time), feedback related to failure can be equally quick. For Proctor and Gamble, which has been the best and most innovative for decades in developing brands, it was a challenge to build a brand on the web since traditional brand-building models do not work. A few months after the June 2000 launch, the struggling MoreThanACard website was shutdown so a new site design could be created (Gribbins et al., 2001). At MoreThanACard.com, consumers can buy gift packs of P&G products for a new baby (brands such as Pampers and Febreze, a fabric freshener) or a care package (loaded up with Pringles, Jif peanut butter, and other products) to send off to college (Bulik, 2000).

The second issue has to do with the importance of externalities when developing e-commerce strategies. Issues outside the organization assume far greater importance for e-commerce than compared to any other application area. Essentially, all upstream and downstream partners need to be understood and accounted for when conceptualizing an e-commerce strategy.

Today's approach (or yesterday's) focused on aligning IS strategy to the organizational strategy. The organization was assumed to be an island in the value-chain sea. Information was a resource and had to be hidden and leveraged. This made the process of strategy creation inward-looking and the focus was always on transactions and efficiency so as to achieve process-level effectiveness. However, tomorrow's approach (which assumes e-commerce to be a norm) assumes inter-organizational integration or an extended organization. In that context, not only will an organization be expected to be tightly integrated internally, it will also be integrated with other organizations with access to each others' information. Relationships ^[7] and partnering will form the basis for collaborative advantages and the entire value chain will be more directly responsible for customer focus. The effectiveness measure of the extended-enterprise will be applied to the value chain and not just to the organizational entity.

Case 1. The Failure of Brandwise.com (Kover, 2000) ^[8]

Even if Brandwise had signed up Kmart, there's no guarantee that the site could have made the alliance work. In theory, a retailer was supposed to feed Brandwise up-to-date inventory and pricing information and link its computers to the Brandwise site. Brandwise and the retailer would cooperate to make sure that each order placed on the website would flow smoothly to the store's warehouse and that the appliance would then be delivered quickly to the consumer. Getting all this straight, it turned out, was a nightmare. Some retailers were still in the Dark Ages of technology and didn't have the data the site needed. Others, perhaps preoccupied with their own e-commerce efforts, failed to deliver adequate data. Still others couldn't figure out how to connect their computer systems with the one at Brandwise. The problem was acute with the large national retailers. With many stores in many different locales, these retailers had inventory all over the place, and kept data in many different computer systems. "Their systems weren't set up to feed us automatically," says former CEO Misunas. Coordinating the flow of data among a company's constituencies—customers, suppliers, internal departments, the sales force—is the software problem of the Internet Age; helping solve it has brought great success to companies like Siebel Systems, Oracle, and many others. So, it may seem shocking that this came as a surprise to Brandwise, but remember, Brandwise was launched at the height of consumer dot-com mania, when details seemed far less significant than vision. Another oversight: Brandwise didn't control order fulfillment, but it did manage the call center that handled customer complaints. While the company had to deal with every angry, bitter customer whose refrigerator was late, it could do nothing to speed up delivery. The company claims this was not a problem, but outsiders say such a setup is a recipe for disaster. "Can you imagine if Amazon did the same?" asks Tom Nicholson of Icon Medialab. "It would destroy the brand."

The extended enterprise has many advantages but the concept comes with its overheads. One of those overheads has to do with governance frameworks. Poorly developed governance frameworks can lead to significant inefficiencies as documented in Case 1. Whirlpool, on the advice of BCG consultants, initiated Brandwise.com in a bid to reap the rewards promised by participating in the Internet bonanza. However, Brandwise experienced major problems due to difficulties in interacting with upstream and downstream partners.

The third aspect that differentiates e-commerce strategy from a typical IS strategy is that e-commerce has to both enable and account for strategic alliances. In the absence of strategic alliances the expected payoff from e-commerce initiatives will tend to be low (Adobor & McMullen, 2002). Alliances represent the importance of coordination in the value web of the future. While command and control structures dominate the monolithic organizational model, the "allianced enterprise" (Harbison et al., 2000; de Man et al., 2001) appears to have legitimized itself as a pervasive and dominant organizational form.

Alliance formation preceded the e-commerce era. However, the nature of alliance formation has been influenced by e-commerce. Some existing alliances became stronger while the Internet made other alliances, that were inconceivable, a strategic and operational necessity (see Case 2). The driver for alliance formation is value creation. Case 2 provides an example of the value specific alliances have accorded to the auto industry. E-alliances ^[9] tend to differ from more traditional alliances in that they may be more fluid and short-lived especially if they are premised on immediate pay-offs.

Case 2. Formation of E-Alliances in the Auto Industry (Enos, 2001; Rothfeder, 2002)

DaimlerChrysler and Union Pacific formed a web-based company to track vehicle shipments from assembly plants to dealers. Under the terms of the agreement, Union Pacific formed Insight Network Logistics (INL), whose sole purpose is to set up a 20-person network control center to track every car Chrysler makes in the U.S. from the factory to the dealer. INL coordinates the mix of railroads and trucks that Chrysler uses to distribute its vehicles, managing the schedules of the companies involved to be sure they're tightly synched and that the automaker receives the most efficient and cost-effective timetable for its shipments. INL posts this information on an intranet so every car and truck in distribution can be tracked simply by typing in the vehicle identification number. Because Union Pacific will essentially be overseeing much of Chrysler's distribution logistics, whether UP rail cars are carrying the vehicles or not, the railroad is, in effect, taking on the unlikely role of a technology services provider. The benefits to Chrysler? The automaker claims it will save about $280 million over six years and reduce by 25 percent the 12 days it takes to move a vehicle from assembly plant to dealership. "This is a nimble move for a railroad, to set up a subsidiary to create a logistics system so quickly," says Tony Hatch, independent railroad analyst and owner of ABH Consulting in New York. "It shows a commitment to technology to help the top and bottom line, and it also shows an understanding of customer needs."

Other automakers have also turned to the Web to track vehicle shipments. For instance, Ford partnered with UPS Logistics Group, a subsidiary of United Parcel in 2000 to create a Web-based vehicle tracking system. The companies said in February 2001 that the initiative had already sliced four days, or 26 percent, off Ford's vehicle transport time. As a result of the reduced transport time, Ford said it has saved $1 billion in inventory costs and more than $125 million in inventory-carrying costs. General Motors made a similar deal with transportation holding company CNF in December 2000. This points to the near necessity of alliance formation in some industries.

However, not all alliances end up as successes. For instance Trilogy Software Inc. and Ford Motor Co. have halted plans to launch a company that was to be Ford's e-business arm. The new company, announced in February 2000, was to be based in Austin and employ 300 people within a year. Both companies decided to steer the initiative, which loosely went by the name "Drive.com," into more of a strategic alliance (Bronstad, 2001). The initial idea of creating a separate entity was discarded as Ford's and Trilogy's business models and strategies kept evolving. This only goes to show that e-alliances can be fluid and subject to constant change. While this may accord much needed flexibility, it can take away in terms of commitment.

There are increasing numbers of cases that can provide anecdotal support for strategy formulation. However, since each organization is unique in its requirements for e-commerce (see the Intel case), we will propose a framework that will help us understand strategic e-commerce issues by utilizing the notion of extended enterprises and then, using the theoretical precept of complementarity as the basis of value creation (with respect to, interfirm alliances).

^[5]Especially the design model that sees strategy as a fit of strengths/weaknesses and opportunities/threats. The central idea is that planning and design of the strategy precedes implementation.

^[6]This is closer to the emergent model of strategy where strategy emerges as a part of its operationalization. The premise here is that the full strategy can not be known in advance and that it evolves and crystallizes as a response to a firm's adaptive behavior to external forces.

^[7]For instance, when indicating dealers to adopt high-speed Internet connections, Ford is mindful of strained factory-dealer relations in the past. Hence, Ford's approach to getting dealers to adopt high-speed Internet access is in the form of a suggestion. "While there has been no edict to our dealers that they have to have high-speed Internet, it has been very strongly recommended," said Dave Rayner, Ford's senior manager of dealer and vehicle systems. "But it's not confrontational at all" (Kisiel, 2002).

^[8]Brandwise.com was Whirlpool's e-commerce (business model) strategy to sell directly to the consumers their ideal refrigerator and connect to a nearby retailer.

^[9]Strategic alliances formed primarily as part of an e-commerce strategy.