Forget the headlines about Amazon.com and eBay. The really big business travelling through the browser isn't about best-sellers or Beanie Babies. It's about binder clips and ball bearings, electric power and electronic components -- that is, the stuff that businesses buy from, sell to and distribute for each other.
Business-to-business e-commerce over the Internet can be as basic as a manufacturer putting up a bare-bones Web site to let distributors securely order a handful of products; it can be as complex as a distributor offering thousands of customers company-specific pricing and content, complex product configurators and near real-time access to inventory levels for its entire product line. Compared with traditional electronic data interchange (EDI) systems that run across private networks, Internet-based business-to-business e-commerce is seen as less of a headache to implement, especially for companies that want to reach smaller customers and suppliers that can't handle EDI's cost and complexity.
Even so, rolling out a business-to-business e-commerce site is not so easy -- or cheap: Forrester Research (US) puts development costs at $US1.5 million for a basic site to more than $US15 million for the most sophisticated of sites; count on spending another $US700,000 a year to keep up a basic site and up to $US4 million a year to run a high-end site.
Most of the costs are for labour, Forrester says in a recent report, "What Business Sites Cost". In addition to programmers, a successful e-commerce team will need graphic designers, content creators and business analysts. Many companies wind up having to do a lot of application development in-house, tweaking off-the-shelf systems and integrating them with back-end systems. A CIO can expect to spend about 40 per cent of his or her company's e-commerce budget on content management, data cleansing, and legacy-system and supply chain integration.
But even if an e-commerce team can get incompatible legacy systems to work together, there's no guarantee that it can get the company's marketing and IS departments to work together -- or that it can mobilise the company to adopt new business models at Internet speed. "They've got to be willing to change the way they do business and adopt some new business models, and that's hard stuff," says Doug Cale, a partner in charge of Deloitte & Touche LLP's global Secure e-Business services. "That's 70 per cent of the work, where the technology piece is about 30 per cent."
With Forrester estimating that US business-to-business Internet commerce will grow to more than $US1 trillion by 2003 -- about 10 per cent of business sales overall -- companies that do not yet have an effort in the works will likely feel increasing pressure to get one under way, either from customers, suppliers or more nimble Net-commerce-enabled competitors. This story will look at some of the biggest political and technological hurdles that CIOs can expect to face along the way and offer ways to overcome them.
Getting Beyond the Starting Line
Perhaps the biggest obstacle to a company implementing a successful B-to-B e-commerce effort is the company itself -- that is, the company's inability to agree on an e-commerce strategy or to even acknowledge that it needs to develop a strategy. Why? An e-commerce effort cuts across the org chart, so either no one wants to own the effort or several people are vying to lead it. "They spend all their time trying to seize the future rather than trying to build it," says Jeffrey Nickerson, a partner with PricewaterhouseCoopers in New York City.
Other times, a company can run into trouble when it tries to launch a business-to-business e-commerce site without the CIO's input, using what Forrester senior analyst Laurie Orlov calls the "toe in the water" approach: The company puts up a simple site that offers a few products and is not integrated to the call centre or customised, just to see if there's any customer interest -- and since the site doesn't offer much, there usually isn't any interest.
The best way to overcome corporate inertia is to have one person with an e-commerce vision become the project leader but work with a cross-functional team to sort out the roles of each department. The CIO can be a leader of this business-transforming effort, Nickerson says. "In the past, often the CIO's role was really reactive to a bunch of business and operational pressures," he says. "Some CIOs are now saying: 'I recognise we can do a lot more with this technology than we're doing. I'm going to go out and get them aligned on this'."
That happened at Applied Industrial Technologies, a $US1.5 billion industrial supplier based in Cleveland. After the IS department spent some time fooling around with Internet commerce technology in an R&D mode, James Hopper, Applied's vice president of IS, approached the CEO about becoming the leader of a Web commerce effort. "Since IS cuts across all of the functional areas, we felt we were in a unique position to move the e-commerce initiative along," Hopper says. The IS department also had more than a dozen people who it could immediately redirect to e-commerce, while the marketing department would have had to start a whole new group. Plus, Hopper had previous experience running his own business before joining Applied. In June his group launched an order-entry Web site, called AppliedAccess, that lets existing customers order from Applied's one-million-plus item product line; the Web site is seen as a replacement for Applied's old, terminal-based online ordering system.
Speaking from a year-and-a-half's experience as an e-commerce leader, Hopper says he's found it a lot tougher than he thought it would be to try to be "Amazon.com within a Barnes & Noble type of structure". For example, Applied is now considering building another Web site that would target new, non-traditional customers who don't already have relationships with the company's local branches; those orders would be placed with credit cards and fulfilled through regional distribution centres. The source of debate between Hopper's group and the marketing and pricing departments is pricing. The company does not set national retail prices; instead, pricing varies by region and by customer. "The hard-charging, Internet time-dependent IS group wants to put a very competitive price out there," Hopper says. "The traditional brick-and-mortar business says: 'We have to understand who the customer is before we can charge the price.'"Smoothing Channel Conflict Hopper's not the only IS leader to find that the CEO's buy-in on a business-to-business e-commerce effort doesn't guarantee the project a turbulence-free flight. Just ask Margaret Bouline, vice president of IS at Dallas-based Aviall, a $US400 million aviation parts distributor. Back in 1996, a few weeks after an upper-management shake-up brought in a new CEO, Bouline hounded the CEO until he approved a $US250,000 budget for a Web-based order-entry system. Little did Bouline know that her project's success would depend on a constituency far from the corner office: Aviall's 300 sales reps around the globe. Fearing their jobs were at stake, she says, they badmouthed the site to customers and told them not to use it. Bouline was caught off-guard. "We didn't take into account that the sales reps would view it as a hostile move and paint the much bigger picture that 'this is the first step in eliminating our jobs'," says Bouline.
Bouline tried pitching the site as a boon for the salespeople, since letting customers take care of routine orders and quotes online would free the sales reps to develop deeper relationships and win new customers. The reps also received commissions for orders placed over the Web. But they still resisted the site. Bouline finally figured out why: Aviall's sales reps view the great customer service they offer as a way to distinguish the company from other distributors. "What we were doing wrong was saying: 'Here's this wonderful e-commerce system, and here's how it's going to benefit you and Aviall'," Bouline says. "We should have been saying: 'Look at what a wonderful tool this is for your customer'."
For the second phase of its e-commerce effort, designed and implemented with the help of C-bridge Internet Solutions, a Massachusetts-based Internet consulting and services company, Aviall took a much more cross-functional, customer-centric approach to rolling out an improved site. The company hammered out its new e-commerce strategy in conjunction with an overhaul of its overall sales strategy. The committee overseeing the sales revamp, which includes Bouline, the vice presidents of sales and marketing, and the CFO, identified target market segments, brought in customers and Aviall business people from those segments and asked them what problems they had and how e-commerce could help solve them. And the pitch to the sales reps has also stressed the customer-service benefits. Every rep has also received a kit explaining how to use the new Web site, which was launched in August, and what speaking points to use to sell it, along with a canned demo on a CD-ROM for reps who did not feel comfortable demonstrating it live.
Channel conflict concerns -- over sales rep commissions, pricing or who will be responsible for servicing customers drawn in over the Web -- are fairly common when a company starts to use the Web as an order-taking channel, analysts say. While maintaining the status quo, such as continuing to give sales reps commissions, is one way to keep the peace, that strategy is only temporary, cautions Kathy Biro, cofounder and CEO of Strategic Interactive Group, a Boston-based Web consultancy. "Eventually, one has to tag sales based on where the value was created," Biro says. "The problem these companies have is that if they do that out of the box, the salesmen will eat their lunch."
The challenge for companies is to spell out the role of the Web site so that different channels complement each other: The site, for example, can serve customers or sell products that can't be cost-effectively served or sold through the traditional sales force. To optimise channels effectively, companies must measure the Web's impact on other channels -- that is, find out whether the Web is serving customers the company reached through other channels but doing so at a lower cost or whether it is actually bringing in new customers and new revenue -- and change the role of the other channels accordingly. "If I don't do that, all I'm doing is adding cost," Biro says.
Taming the Technology
Beyond the strategic snafus that can stand in the way of a business-to-business e-commerce effort, CIOs have to overcome serious technological hurdles, chief among them, legacy-system integration. At Aviall, for example, the sales force was segmented by what products they used: airline sales used one legacy system, general aviation used another, and the Australian group used another, Bouline says. The company's first order-entry Web site did not even attempt to integrate with these home-grown, back-end systems; orders entered online generated e-mail forms that went to sales reps, who rekeyed orders into the appropriate system -- hardly an efficiency booster. But when Bouline embarked on a Y2K-influenced, $US8 million infrastructure upgrade, she chose an Internet-ready ERP system offered by Minnesota-based Lawson Software, called Lawson Insight, that would make it easier to roll out Web ordering for the second-generation site. In the future, she hopes to be able to use the system to link to supply chain partners -- for example, to let a supplier see what inventory Aviall has so that the supplier's sales rep can sell off excess inventory in Aviall's warehouse.
At office supply distributor W B Mason, CIO Peter Dupre found that using an off-the-shelf e-commerce package from Ironside made the task of legacy-system integration easier. "They give you a shell of a data gateway that has all the routines written in, and you have to write the data cases," Dupre says. "It was the hardest part [of the rollout], but we did it in 30 days." At Applied Industrial, meanwhile, Hopper's e-commerce team turned to Extensible Markup Language (XML) to integrate its home-built Web site and its Cobol order-entry processing system. Unlike HTML, which just marks up text, XML lets developers write custom tags to identify objects. Beyond using XML to solve its interoperability problems, Applied made the decision to re-engineer its processes to be XML-centric -- that is, to build code components that can be easily combined or reorganised. It is now starting to use XML as a way to bridge the gap between its systems and suppliers' and customers' systems. "By sharing XML definitions and access points with our trading partners," Hopper says, "we can perform systems integrations on a much more rapid time line, independent of the systems or platforms with which we are interfacing."
The legacy-system integration issue becomes even more complex for companies that have data about individual customers scattered across several different databases and want to get a single view of a customer's business with the company. David Stoltzfus, CTO of Logical Design Solutions, a New York-based Web consulting company, gives this example: A large financial services institution may offer several products -- brokerage, mutual funds, insurance, credit cards, savings accounts -- but each product may be managed by its own legacy accounting system. There's no easy way to link all of these accounts together to let a customer see his or her entire portfolio of products without having to be authenticated for each account. "It's the single most expensive problem to fix," Stoltzfus says.
To overcome that problem, many financial services organisations have started to create an enterprisewide "customer information file" that sits on top of legacy systems and has cross-reference keys to a customer's accounts. That way, Stoltzfus says, the customer can log in once and access all accounts. To implement this kind of solution, companies need to make changes to their legacy systems so that changes in customer information also make changes to the customer information file; they also need to do some serious data scrubbing and extensive testing and verification.
Another hurdle is offering 24/7 availability to data, since most back-end systems are designed to run some type of batch cycle at night, Stoltzfus says. "That becomes a really big issue because the expectations on the Net are that it is always available," Stoltzfus says. Data replication is one way to work around the data availability problem. But CIOs must decide whether it's worth the expense of replicating data every 15 minutes or whether it can wait every 15 hours.
Don't make the mistake of thinking that business-to-business sites don't need to offer personalisation, Forrester's Orlov says. "Business buyers have company characteristics that distinguish one from another, and they also have individual roles and responsibilities, authorisation levels, and products they buy over and over again," Orlov says. "Without any form of customisation, you can't trigger cross-sells or upsells."
In addition, security -- confidentiality, privacy, protection of proprietary data through cryptography and authentication -- is crucial to building trust with business partners. In many cases, notes Deloitte & Touche's Cale, a company may not be able to anticipate how sensitive data is to its customers or partners. His advice: "If you're setting up an extranet with a business partner's data on it, you need to assume the worst and provide strong protection."
Navigating these strategic and technological hurdles may sound like a tall order to some CIOs. But to others, the challenges of implementing a sophisticated business-to-business commerce site represent the next frontier for them personally as well as for the company. "For years and years, we've been automating back-office systems, like accounting and purchasing," says Applied Industrial's Hopper. "What are [we] going to do for the next hundred years that's going to add value to the business?"
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.