From Computer Business Review, a sister publication
Each of the new businesses created by the Unisys restructuring this January is already a significant way down the path to becoming the high-growth, wider margin businesses which the company needs to build if it is to successfully escape the declining margin, low-growth trap that it found itself during the late 1980s.
Draconian restructuring
To some extent, this can be attributed to the draconian restructuring of the old business, which now essentially survives as the Computer Services Group (CSG). But it is also due in no small measure to the successful establishment of the other two service-oriented groups. These services companies between them now represent 60% of the company, and have helped it reverse its ratio of products sales to services revenue from 57:43 in 1991 to 43:57 today. That in itself is no small achievement but as Unruh himself concedes the job isn’t finished yet. His personal goal is to complete the transformation of Unisys before he steps down which, given his belief that most organizations should change leadership at least every ten years, means that he only has four years left. If Unruh is to retire a happy man, he has to quickly show that the ‘three businesses/one company’ model will work. Even when the new strategy was first announced last October, a number of industry commentators speculated that what Unisys was preparing itself for, was dismantlement and sale. Unruh denies this emphatically, but as Greenway Partners’ proposal showed, the notion may take some time to go away. If, in the meantime, Unisys suffers any unexpected reversals or fails to meet its promises, it could easily be back high on the agenda. This time, Unisys was fortunate that Greenway Partners probably chose the wrong platform for their assault. Al Kingsley, Greenway’s senior managing director, argues that the three-business model will not work, because it is disadvantageous to the Information Services Group (ISG), potentially the jewel in Unisys’ crown. We believe that the company is losing business because of the setup it has now, says Kingsley. If you walk into a customer at Unisys Information Services they’re going to assume you will push your own equipment. You cannot expect a computer manufacturing company to be a successful computer services company. Digital Equipment and IBM would certainly join with Unisys in disputing that. Kingsley also argues that, with Unisys’ share price in the basement, it is unlikely to be able to attract the skilled systems integration people needed to make a success of such a business. It’s very different to attract talent with a $6 stock, he says. Neither argument is original and, fortunately for Unisys, most independent observers appear to think the Greenway analysis lacks validity. Bonnie Digrius, vice president of Gartner Group’s consulting and systems integration service, believes that customers – and vendors for that matter – are increasingly less concerned about the badge on the hardware that they offer as part of a larger systems integration contract. That’s just a marketing ploy services vendors use against each other, she says. For a return to profitability, analysts will be looking most closely at the performance of ISG – something that has been marred by problems of the units’ own making that resulted in it accounting for an undisclosed proportion of last year’s $624.6 million losses after several of its systems integration contracts turned sour. Over-staffing also contributed to the red ink, according to ISG president, Larry Russell. Unisys is coy about why the projects went wrong, and claims that they represent less than 10% of the units’, overall business which has been growing at somewhere north of 30% compound since it was established in 1991. We took on contracts which were simply impossible, says Russell. Or we took on relationships which from a client or technology perspective were just not reasonable. So we’ve stumbled a little bit. It was a modest price to pay to have built
a business which effectively goes a long way to solving the margin problem that all mainframe computers face, Russell maintains. The price was too high, says Gartner’s Digrius, who believes that ISG had more than a few problem contracts last year. In the systems integration business, she points out, people are everything and last year Unisys lost a succession of senior people. Victor Millar, ISG’s former head, defected to AT&T Global Solutions (now NCR), and both of his successors have also quit: Sandy Sanderson going to Oracle Consulting and Jim Corey also to AT&T.
The first priority
Russell concedes that Unisys has got to get the risk dimension out of this business. The first priority is to improve execution and bring this business to a reasonable level of profit, about 10% return on revenue, pre-tax, fully- loaded, and still maintain a growth rate of 20%. To do this, Unisys ISG must go head to head with more firmly established competitors, such as IBM, EDS and Andersen Consulting. In its favour, the ISG business already has an established presence in its chosen vertical sectors (transport, communications, public service, financial and commercial) and, as Russell says, although his group is small compared with IBM or EDS, in a fractured market where 150 companies are chasing $150 billion worth of business, being a $2 billion player is not a mark of weakness. Gartner’s Digrius believes ISG has a chance to of making its targets, but only, she stresses if it tightens its management processes. Systems integration is such a robust market, anyone can make money, she says. If you can’t [as ISG failed to do last year] you’ve got major problems. If ISG is the least mature of Unisys’ businesses, but potentially the most profitable, Global Customers Services (GCS) is already the business which comes closest to meeting the company’s strategic ideal, even though it still depends to a great extent on traditional systems maintenance contracts for the bulk of its revenues. In reality, the traditional maintenance aspect of the business is diminishing quickly as GCS moves its resources into desktop facilities management and network planning, implementation and management. Unisys’ figures shows that core computer maintenance revenues have declined consistently since 1990, dropping from about 80% of GCS’ revenues to around 50% last year. Maintenance also actually now takes up little of Unisys’ resources as most of the work is now subcontracted to local suppliers so that GCS personnel can concentrate on the ultimately more sustainable businesses higher up the food chain. In 1995, desktop services accounted for 12.5% of GCS revenues and the fast growing networking services brought in 25%. However, the efforts Unisys has made to establish itself as a services company have been counter-productive in some respects. Not least because it has fostered a widespread perception that the company has backed away from being a computer technology provider. That was never an accurate perception, says Unruh. We do not try to be in every segment of technology provision, but technology isn’t any less important, its just no longer sufficient in really creating value for a client.
Systems revisited
Thanks to its history, Unisys ought to have a headstart in this area, and among its established customer base it still does. Outside of that customer base however, it is Unisys’ credentials as a systems builder which have suffered most over the last 10 years, and many of the advances that it has made are among the company’s best kept secrets. For this reason, as well as making a strong recovery and healthy profit, the greatest challenge facing the newly focused Computer Systems Group (CSG) group will be to re-establish the company’s reputation as a technology innovator, while coping with the costly development structures. CSG seems to have recognized this already and is firing on all cylinders. In April the group announced its Clearpath strategy. ClearPath is Unisys’ attempt to halt the flow of its traditional customer base to non-Unisys Unix and Intel-based servers. The servers and future product plans show how users of Unisys proprietary A Series and 2200 mainframes will eventually be able to switch their applications and processing to a single unified architecture. ClearPath is a line of CMOS mainframes hybrid servers and Intel-based servers. The hybrid machines allow users to run a mix of mainframe, Unix and NT applications. And the multi-processor Intel-based servers will run mainframe applications under emulation, as well as Unix or NT applications. Unisys has also undertaken a major re-branding of its whole PC product line. And in June, CSG plans to deliver a new high-end server line, a 10-way version of its 6100 SMP Unix machine which runs Microsoft Windows NT. These announcements suggest that, even in its present financially straightened position, Unisys is committed to investing in the development of key technologies. As Charlie Burns, director of research for Data Center Strategies says: Unisys often comes up with really good ideas, their problem has been marketing them. Now though, claims CSG president Alan Lutz, the company’s systems products business may have the most to gain from the new structure and the opportunity for focus. Though the CSG people have always been computer designers and manufacturers, their activities have been submerged by other considerations, says Lutz. Under the new regime, CSG also has the autonomy to direct its own product strategy, formulate its competitive approaches and to choose its own partners. In the latter respect it appears to have already chosen well. Much of the technology being channeled into the company’s large servers comes from a five year-old relationship with Intel. The relationship appears to have given the company a headstart in harnessing Intel’s P series products, which it is using to build its next generation of multi-processor Intel- based servers. The 10-way 6100, for instance, is considerably ahead of anything yet announced by other Intel-based vendors such as Compaq and positions CSG to go head-to- head with server companies like Sequent. The choice of Windows NT, rather than Unix, for the 10- way 6100 machine is also significant. Unisys has been a fully paid-up member of the Unix community since the mid-1980s. However, Unisys’ strategic drift is towards Microsoft. To be successful in Unix you have to have the application vendors porting to your operating system very quickly or you don’t have anything to sell, says Lutz. Market share determines that and we don’t have [sufficient] market share. The answer is to move that development focus over to NT. If you can’t be anything more than number five in a marketplace, then don’t bother playing; move over to a place where you can be number one two or three in the market place. Working hand in hand with Intel and Microsoft is paying dividends. But that approach suggests that Unisys is joining the growing generation of companies who have gone from being computer manufacturers to become machine integrators, assemblers of commodity products and software. The question remains though, can Unisys do a better job of bringing these products to market. Perhaps, the first important test of whether it understands this itself will be how it handles the re-branding of its PC product range. Even though Unisys has made an impressive showing in the PC market in the last couple of years, by Lutz’s own admission we are not even a blip on Compaq’s radar screen. Its initial response to this problem, typically, has been to invest in technology and, in particular, manufacturing capacity. In its San Jose manufacturing plant Unisys now has the capability to take standard basic configuration PCs (something it terms as Z-boxes) and configure them precisely to customer requirements for delivery in volume within 14 days. This has helped win friends at larger corporations, and has had the benefit of substantially reducing inventory and other manufacturing costs. The overall effect on Unisys’ PC business has been dramatic. A year and half ago, says Lutz, the PC business was a loss leader. The only reason we sold PCs was to keep our name in front of the customer. Last year though, the company increased PC revenues by 73% to $623 million and it expects to see a further 40% growth in shipments this year, taking PC revenues passed $1 billion. The UnisysPC operation is also one of a handful of large PC vendors which made money on PCs last year – and it expects to do so again in 1996. According to Peter Zwetchkenbaum, director of branding research for IDC, Unisys must now really go for broke. In terms of its manufacturing capability, he says, Unisys truly has an advantage of the competition but they need to find a unique value proposition to brand that end of the business.
Challenge to return to profit
The positive changes to the PC operation, aside, the company’s overall weak performance in recent years has put a tremendous stress on the management at all levels of Unisys. Nevertheless, the new structure seems to have genuinely revitalized morale and given new drive to staff in all areas of the company. The challenge for Unruh and his management team now is to replicate this enthusiasm in the marketplace and return Unisys to profit. He says that profitability will come this year and continue into next. Some analysts say that is wishful thinking – but for Unruh and for Unisys there will be no margin for error.
