Hummer-Winblad, a two-person venture-capital firm, specializes in software start-ups. Two years ago, "software was cold." Other fund managers didn't want to invest in computer geeks managing other geeks, with no saleable assets. Now software is hot because margins in hardware and other technologies have thinned. A software company can produce an 80% gross margin and 20% after-tax profit -- and can fund future growth out of profits. It's also easier now to take a software operation global.

Unfortunately, the existing installed base of software is an albatross. Stand-alone products have limited interest. AI has failed to produce a single successful product, although it has been embedded in several applications, and now multimedia is following the same path.

High-tech alone doesn't make a good investment. Single- platform software (e.g., Unix, Mac) may not make the $10M/year in sales required for staying power. Application software (e.g., software tools) is a dangerous market if custom software is the coming trend. Small-business computing is underfinanced, unique, and hard to reach with advertising. Home computing tends to be just an extension of the office desktop. And "content" or "title" companies [like Electronic Arts?] are not yet understood by the financial community.

Areas that do appear solid (and in which Hummer-Winblad has invested) are utility software, GUI tools, client-server (i.e., distributed) data access, pen-based and remote computing, and networks. [Ann Winblad, Upside, 9/91.]