Control Data Tightens Reins At Micrognosis; Growing Pains Felt At 60 Wall Street Project

No one believed Micrognosis Inc. could swallow Morgan Guaranty Trust Co.'s massive 60 Wall St. project whole without experiencing some signs of indigestion. Plop-plop fizz-fizz.

The formidable Morgan Guaranty project, estimated to be worth over $25 million, has triggered warp-speed growth, just as it has forced Micrognosis into the unaccustomed role of full-bore custom software development. For Micrognosis Inc., with 1987 revenues of about $57 million, 60 Wall St. is a bellyful.

But changes at Micrognosis occasioned by the project are matters of structure as well as scale. The development of specifications for Morgan Guaranty coincided with a sea change in the architecture of trading systems. The turmoil created by that change has led to untimely turnover among top engineering management.

Majority shareholder Control Data Corp. has responded by tightening management controls, calling in senior engineering resources, and cultivating a direct interface with top project managers at Morgan Guaranty.

In this context, Control Data's recent buyout of Micrognosis minority shareholders has raised questions concerning the Connecticut-based systems integrator's future and focused critical attention on its engineering management function.

Unlike the Micrognosis International subsidiary of Control Data, Micrognosis Inc. was never wholly-owned by CDC. As of March 31, CDC raised its stake in the company to 96% from about 74% by purchasing approximately 270,000 non-voting shares at $15 a share. By virtue of the purchase, CDC's ownership crosses the 80% threshold qualifying Micrognosis Inc. for tax treatment as a corporate subsidiary.

The buyout is characterized in sharply contrasting terms by former shareholders and by Micrognosis management. Outspoken former shareholders claimed they bailed out, while Micrognosis officials read the move as an active expression of CDC's strategic interest in their operation.

One disgruntled shareholder offered the following explanation of the purchase:

"First, as shareholders, we're a bit of an irritant. They'd like to do away with stockholder meetings and our embarrassing questions. "Second, they may be positioning the company for sale. No one wants to buy a company with outstanding minority stockholders.

"Finally, there's the tax aspect. Micrognosis Inc. made a profit in '87 and unless Micrognosis can be treated as a subsidiary, the money has to stay in Micrognosis and be taxed as a profit. If it's accounted for as part of CDC, it can be applied against CDC's losses."

Officials at Micrognosis, however, suggest that the stock purchase has long been a feature of CDC's strategic plan for the company. This is confirmed by David White, president of Control Data's business services group, of which Micrognosis is a unit.

The shares were purchased from a group of 24 individual shareholders who had organized a negotiating group. Most shareholders are former Micrognosis employees. Others are still employed at the firm. All but one of the Micrognosis shareholders have accepted the offer.

If Micrognosis is, as it appears to be, a profitable company with excellent growth prospects, why did 24 of 25 shareholders elect to sell out? Some former shareholders point to management troubles.

"The management was atrocious in the last year or so. It's a house of cards there, and if the right card is pulled the whole thing could come crashing down. I think that was the general feeling of everybody in the shareholder group," says one former shareholder.

One card recently pulled was Stephen Sutton, former vice president of engineering at Micrognosis. Sutton, who resigned his post Feb. 1, will be taking a position with CP Technology in London. Sutton served as senior technical liaison with Morgan Guaranty during much of the 60 Wall St. project, and was replaced by CDC's Ed Craddock.

Shareholders charge that the 60 Wall Street project has been jeopardized by turnover among top engineering staff at Micrognosis.

"There are a lot of sour grapes going around," says Chris Manderson, vice president of sales at Micrognosis. "Morgan Guaranty is delighted with the progress we're making."

Ingrid Howe, Micrognosis project manager for 60 Wall St., says there has been no undue turnover among Micrognosis engineering staff.

Morgan Guaranty concedes some engineering-related difficulties, but minimize their effect on the project as a whole.

"Frankly, although we've had some concerns with various details of this very large project, they're resolved now and we're moving ahead very nicely," says Paul A. J. Hurley, senior vice president of systems at Morgan Guaranty.

Perhaps the most cogent analysis of the situation comes from another Morgan Guaranty executive close to the 60 Wall St. project:

"Here's what's happening. You're taking an entrepreneurial organization -- maybe 250 people -- a traditional engineering firm, and you want to make the transition into the big leagues with organization, structure, and solid management reporting.

You need a solid infrastructure to do that. You need to get out of the mom and pop garage. I think this whole Sutton activity sped up that transition."

Another key point is made by a former Micrognosis engineering employee:

"Their [Micrognosis engineering management's] attitude was that this is not an engineering project, this is a software project -- they changed us from a systems house to a software house."

Systems integrators are having a bumpy ride of late. Stepped up competition, rapid evolution of systems architecture, and increased emphasis on custom software development are forcing them to learn a new game.

The same thing that happened at Rich Inc. (see TST, Dec. 21, 1987) is happening at Micrognosis, and more. The shift in emphasis from systems engineering to software development is occurring in parallel with the evolution from an entrepreneurial to a corporate structure.

These two changes, either of which alone would be enough to wrench the guts of any company, have been sufficient to generate unusual friction.

Drew Henderson, ex-Micrognosis engineering executive and organizer of the shareholder negotiating group, says, "Control Data manages everything at a distance. They manage from paper, from the bottom line. They're not involved in the business. They don't understand the business. So they make bad management choices."

CDC may be getting more involved. An executive with Morgan Guaranty's 60 Wall St. project says, "CDC is very interested in Micrognosis as a long term investment and as a vehicle for their entry into this marketplace. Consequently I think their interest in purchasing the outstanding shares is a positive one. I've met on several occasions with top-level executives from CDC and that's been the message conveyed to the house of Morgan."

But why should Morgan Guaranty have to interface with Control Data officials at all? "The Morgan contract represents a very large opportunity for us," says White, "and it's not incongruous for us to take a direct interest in an important client."

That depends on who you listen to. One former Micrognosis employee says, "Control Data never touched Micrognosis [Inc.] when things were going well." Another claims the reason for CDC's intervention was a technical problem:

"The problem was designing with a database management mentality, not having a real appreciation for the real time aspects. Rather than doing upfront systems design, they immediately sat down and started writing code.

"Micro convinced Morgan -- which Morgan really loved to hear -- that you could run the whole thing not on multi-bus chassis types of products, but on Microvaxs. The idea was to rewrite all the servers, port it to Microvaxs and hang them all on a LAN and voila ... DEC hardware and software under VMS written in C.

"They didn't have the slightest idea what real time was all about for a system of this magnitude.

"The fact is, you can't field a system with that much performance with the huge overhead of a big operating system.

The problems of feed servers and embedded display drivers are more suited for real time kernel than a full-fledged, disk-based operating system. The virtues of the operating system are wasted when you're doing a threaded, multi-tasking real time system.

"When they started to bring up these Vaxs they didn't get anywhere near the performance they needed, so now you've got a crisis, a real crisis."

The problem at 60 Wall St. was an architectural one, having to do with the distribution of processing power among servers and workstations within the context of a very high-throughput messaging environment. A source at 60 Wall St. comments:

"The issue is, as you rack forward on digital sources -- versus video -- the traffic rates on the underlying digital networks are going to go up exponentially.

"As digital proliferates, the need to use data from internal trading systems as well as market feed systems becomes greater, but the traffic rates that an internal system produces can be far in excess of an average collection of market feeds.

"In general terms that's where the problem was and we have a very solid game plan for dealing with it.

"We don't buy off the shelf like some banks do. We're asking for things that no vendor has in their product set today, despite what they might tell you.

"When we're through this thing, I actually think we're going to have a world-beater system at the hands of Micrognosis. I don't think we would have gotten it from Rich because they would have been too inflexible, and Data Logic, although more adept at the software end of the business, have no presence for this scale of project in New York."

The Morgan Guaranty 60 Wall Street project is projected for completion in the third quarter of 1989.