Bargain with kings and princes, but after you've blustered and bluffed, never, never, dicker with a vendor that is a gorilla. The beast just might hold the key to your IT planning. The techniques investors use to evaluate the potential performance of technology companies can be adapted by information executives for strategic planning. In high-tech, as investors and information executives know, betting on the future prospects of a company is more risky than in other sectors. The chancy situation requires successful investors to adopt some new rules, which are set forth in The Gorilla Game: An Investor's Guide to Picking Winners in High Technology (HarperCollins Publishers, 1998, available here through Amazon.com).
High-technology businesses continually bring to market discontinuous innovations -- new technologies that promise dramatic increases in price or performance but are not compatible with existing systems. These innovations challenge the marketplace to create a new supply chain for their delivery and support. Sometimes the marketplace responds with a "yea", and such innovations as PCs, laser printers, LANs, relational databases and Web sites charge onto the scene. But sometimes the answer is "nay", and such innovations as pen-based laptops, wireless LANs and massively parallel computers are greeted by a vast indifference. CIOs who committed resources to the former look like geniuses; those who committed to the latter look like the goose at Christmas dinner. The key is not to commit until the technology has "crossed the chasm", that gap in the technology adoption life cycle. Until a fully functioning supply chain arises to support the new offering, an innovation's long-term viability is suspect.
As a CIO, if you sponsor a technology before it has crossed the chasm, then you have to be prepared to write off the entire investment, including the cost of returning to a supportable infrastructure if the technology fails. There may be a business case for taking this risk -- Wal-Mart had one back when data warehousing was no sure bet, for example -- but committing to a new technology should never be undertaken without a contingency plan.
Once technologies cross the chasm, the risk to a CIO is no longer a matter of investing resources in the wrong category but one of investing resources with the wrong vendor. Many markets in enabling technology develop along a winner-take-all model, and CIOs who back anyone but the winner are in for a hard time. When markets replace an old infrastructure, they do so in a big hurry. Business, naturally, wants to get back to business and end the transition quickly. To bring off rapid global deployments, for example, you must have standards, but standards are not to be had when a category is brand new. So the market standardises on the market leader's architecture, making it the de facto standard.
If that architecture is proprietary, as it was in the case of Intel, Microsoft, Cisco and Novell, then the market confers a tremendous competitive advantage on that vendor. It becomes virtually impossible for any rival architecture to succeed. Thus Motorola's 68030 chip, IBM's OS/2, Wellfleet's routers and 3Com's (later to be Microsoft's) LAN Manager all got squeezed out of the market.
Business has no time to wait for the development of a competing standard.
Companies that have proprietary architectural control over a product that has high switching costs are gorillas. We advise CIOs to follow the same rule we offer investors: In a gorilla-game category in enabling technology, once you know who the gorilla is, bet heavily on it and withdraw your support from all of its competitors. This is the domain of what economists now call the law of increasing returns, which states that as a company gains market share, its power and profits increase, rather than decrease, with each additional percentage point of share.
This is a fairly rare occurrence. The more familiar law of diminishing returns usually occurs, and it has a natural curbing effect on monopolies. Compare Microsoft's position in the NT server space, for example, with Compaq's in the PC space. Both are market leaders, both are providing enabling technologies, but Compaq does not have proprietary architectural control over the PC. It is not a gorilla. Dell, Hewlett-Packard or IBM can always field a directly competitive offer. Because switching costs for PCs are relatively modest and low-cost clone companies can sell products at prices that, if Compaq matched, would decrease Compaq's profits, it is possible for a customer to bargain with Compaq, a king. On the other hand, a CIO who bargains with Microsoft in the NT server space does so knowing that if he finally opts to take his business elsewhere, he does so at an enormous switching expense and risks going to a vendor whose platform may shortly become incompatible with industry standards.
The rule for CIOs is plain. Bargain with kings and do business with their lesser competitors -- princes -- but once you have committed to a gorilla company's architecture, you cannot realistically switch away.
And if you commit to the chimp instead, you will have a rough road ahead. As a chimp's customer, you quickly learn what happens when the marketplace commits to one company's architecture and you are on another. Every year, your set of options gets smaller, and the industry standard infrastructure gets stronger, broader and cheaper. The key to avoiding a singularly painful experience is to postpone commitments until you are sure who the gorilla is.
Not all gorilla games have a pure winner-take-all structure. Within the category of gorilla vendors, there is a strong distinction between application vendors and enabling technology vendors. Virtually all applications have significant proprietary content, which makes for high switching costs.
Nonetheless, the power of an application gorilla is much less than that of an enabling technology gorilla because the market can support multiple gorilla-class application vendors with relative ease.
Among ERP applications, for example, SAP is clearly the gorilla, but PeopleSoft, Oracle and Baan all have strong offerings. And if you go down a tier, companies like Lawson, J D Edwards, QAD and Symix all have sustainable, growing businesses. The gorilla in an applications domain may get preferential treatment, but it cannot eradicate its competition the way NT is eradicating Novell's LAN OS or Windows eradicated OS/2 and the Macintosh OS. CIOs can choose applications from a much broader field of candidates and pick a second-tier vendor with confidence wherever its functionality or focus is a better fit with their companies or industries.
Geoffrey Moore, chairman of The Chasm Group, can be reached at email@example.com.
Three Rules to Remember
By Geoffrey Moore
- Don't commit to any technology until after it has crossed the chasm.
- Use normal rules of engagement when dealing with enabling technology kings and princes and application companies of any size.
- Wherever there is an enabling-technology gorilla, get on that bandwagon and no other.
Finally, a quick way to learn which companies are which is to track the market capitalisation of your vendors compared with the other companies in their category. If they are at the head of their pack, you can be reasonably assured of an ongoing stream of investment in the product line. If they fall to the rear, you need to be increasingly circumspect in making future commitments.
Stock market valuation, after all, is simply the reflection of a whole lot of investors' best guesses as to future competitive conditions in your vendors' market. Their judgement is no substitute for yours, but it is worth paying attention to.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.