I stumbled upon an interesting article from earlier this year while surfing the web for venture capital links. There were a few points that I found very interesting.

• “Hardly a week goes by without a reminder of the ongoing vendor consolidation. Larger software companies have emerged as a new tier of active consolidators, including IBM, Oracle, Microsoft, Computer Associates International and Sun Microsystems. As these larger companies bulk up–either by acquisition or internal development–smaller software companies are finding fewer niches to fill.”
  

HP Software is a prime example of this enterprise consolidation. The acquisition of Mercury Interactive back in late last year truly made HP Software a serious player. The internal product managers touted the breath of offering that only a few others could compete with.

• “A critical requirement for start-ups is that they’ve identified an area where pain is extreme, so extreme that a company is willing to deal with a start-up and willing to pay money for it,” said David Skok, a partner at venture capital firm Matrix Partners. “No question, it’s tough.”

This is probably one of the reasons its so hard to enter the enterprise software space. Enterprises won’t talk to you unless you fill a burning hole so big that they will deal with a startup. Plus, once you get in, you better bend over backwards to make things work. This seems so different compared to opening a consumer focused Web 2.0 company. What kind of burning pain did MySpace fill?

• “The amount of money required to start a hosted-software venture may be higher than a traditional software company. Salesforce.com, for example, required about $65 million in venture funding before it went public, according to Trinity Ventures. Rangaswami noted that payments to hosted-software companies tend to be spread out over time, rather than pay out upfront. Matrix Partners’ Skok said companies that sell on-premise software–rather than hosted software–need about $30 million to $35 million, in general, to get started, which is consistent with years past.”

Really? I guess I can see the additional cost in hardware, bandwidth and delayed revenue stream. However, aren’t coding cycles longer, support costs high and implementation costs higher with “Shrink Wrap” products? Plus, isn’t the distinction between “hosted software” verses “shrink wrap” software blurring? I know that many of the Mercury products were being offered in both models. Why not offer both? I do believe that much of software is going towards software-as-a-service but large enterprise software may be the last to go that way (despite taunts by Google to move into this direction).