When David Birch says that "start-up is a terrible phase," he sounds more like a fledgling entrepreneur than a respected Massachusetts Institute of Technology researcher. There is a good reason for his new outlook. Birch, well-known for his landmark study of small business job creation, is now running a start-up himself, as president of Cognetics Inc., a market research firm he founded last year with four MIT colleagues.

Birch continues to serve as director of The MIT Program on Neighborhood and Regional Change. During his years as its director, he has acquired and refined information on 5.5 million U.S. businesses. At Cognetics, he uses that database in conjunction with some 80 others that cover such diverse subjects as aircraft ownership and water utilization.

Cognetics helps its clients -- virtually anyone selling anything to business -- by searching those databases and pinpointing the most promising prospective customers. "Rather than giving our clients a huge laundry list of names," says Birch, "we're using a tremendous amount of data and knowledge to zero in on a few companies that should be drooling to use that particular product or service."

About a dozen clients have come calling on Cognetics so far, including accounting firms, banks, several municipal agencies, and real estate developers and brokers. One landlord, for example, recently commissioned the firm to drum up a list of potential parties likely to be interested in leasing or buying the commercial building he was renovating in Bedford, Mass. Cognetics drew up a ranked list of 200 candidates. Birch proudly reports that 4 out of the first 20 prospects said they would take the building if it could be made available immediately, and a press time, the number two prospect was negotiating a lease, not only for the building but for the surrounding land as well.

Does Birch find that entrepreneurship in-action differs from the phenomenon he has spent much of his professional life quantifying? "I would be dishonest it I said it was a piece of rosy cake," he remarks, adding that a substantial chunk of his budget has been devoted to keeping down the mouse population in his new quarters, an abandoned-looking old furniture factory in East Cambridge, not far from his MIT office.

But, ever the academic, he points out that "Cognetics is the ultimate realization of our research. If we really understand what's going on, we ought to be able to make the economy work more efficiently." Then his voice takes on an entrepreneurial tone. "Also," he adds, "It sure helps put the kids through college."