Michael Lehman, the finance chief of Sun Microsystems, a high-end computer-maker that rose and fell with the dotcom boom of the 1990s, recalls a typical customer visit from about a year ago. A delegation from America’s navy, a big client, sat down in Sun’s Silicon Valley offices. The admiral’s first questions were “How is your business?” and “How is your stock?” The uncomfortable burden of his interrogation was: “Will Sun survive?”
Things are different now. After five years of losses, in which both revenue and market share fell, Sun is growing again (see chart). It is winning customers from its rivals and, after a profitable quarter, is hoping to post its first full-year profit since the boom. Sun’s bosses have even set themselves an ambitious performance target: operating profit, now less than 3 percent of revenues, is to be 10 percent of revenues by 2009.
All this is great for Jonathan Schwartz, who took over as chief executive from Scott McNealy, one of Sun’s co-founders, a year ago this month. People at Sun are wary of giving too much credit to the new boss — Mr McNealy stayed on as chairman — and avoid using words like “turnaround”. But Benjamin Reitzes of UBS, an investment bank, says that Sun’s situation reminds him of Xerox’s swift revival five years ago, also under a new boss. Laura Conigliaro of Goldman Sachs, another investment bank, expects those 10 percent margins to be “a milestone rather than an endpoint”. Overnight, as it were, the setting Sun has begun to rise again.
Surprisingly, Sun’s resurgence appears to be based not only on the boring stuff of lay-offs and cost-cutting, but also on a vigorous plan to grow in an industry that is supposedly stagnant. The market for humdrum corporate servers supporting run-of-the-mill business applications is shrinking, in effect, since the price of the things is falling faster than the growth in volume being shipped. But, says Greg Papadopoulos, Sun’s technology chief, there is a parallel “universe” that is expanding quickly. It contains unusually demanding customers such as FedEx, which needs big computers to route trucks and aircraft efficiently; drug firms that need to model complex molecules; weather and climate forecasters; and fast-growing “Web 2.0” start-ups that handle huge amounts of data, such as video and blogging sites.
Mr Papadopoulos, borrowing from astronomy, calls this expanding constellation of clients a “red-shift” market. As the universe expands, light from galaxies moving away from Earth appears shifted toward the longer (and hence redder) wavelengths of the spectrum. Light from stars and galaxies moving towards Earth, by contrast, would be blue-shifted. Sun’s strategy is to keep fighting rivals such as Hewlett-Packard (HP), IBM and Dell for the traditional “blue-shifted” customers while also targeting the more interesting “red-shifted” customers.
How does Mr Schwartz propose to do this? In part, by being much more pragmatic about alliances and technological combinations than was Mr McNealy, who harboured emotional enmities with rivals. In the past, Sun sold only proprietary boxes that included its own chips and Solaris, its own operating-system software. In recent years, under Mr McNealy’s aegis but with input from Mr Schwartz, Sun opened up, selling computers based on processing chips from AMD, and making Solaris open-source — ie, available without charge. As boss, Mr Schwartz has opened up even more. In January he gave news of a partnership with Intel, the world’s biggest chipmaker and a former rival.
Mr Schwartz hopes for three things from such openness. First, to entice more firms to become first-time Sun customers by letting them run Solaris free on IBM, HP or Dell hardware, or on cheap Sun boxes based on Intel or AMD chips. Second, to boost the popularity of Solaris as a “platform” for other software. And third, to encourage red-shift customers to buy Sun’s most powerful machines without fear of proprietary “lock in”.
This strategy has risks. Vernon Turner at IDC, a market-research firm, warns that customers may decide to use Solaris, but only on non-Sun hardware. Worse, red-shifting companies might decide to emulate Google, an internet company that cobbles its systems together from large numbers of low-cost machines, rather than buying expensive kit from a single vendor. For the moment, however, Sun is rising.