This could be the golden age of IT buying.
The IT buyer is blessed with commodity hardware, open source infrastructure, a surplus of vendor capacity in both hardware and software, which is in turn driving mammoth discounts, cheap offshore engineering talent, falling telecommunications costs, standardization and open protocols...The fall in the cost of bandwidth, storage capacity, computation, software all are helping IT buyers generate much more with much less. Traditional sources of vendor value - intellectual property, proprietary implementations that raise switching costs, unique product offerings - sufficiently burned IT buyers so that today they are looking for fungible, commodity solutions that reduce vendor lock-in and attendant costs.
The trends are clear - RISC to x86, Solaris to Linux, Boston to Banglore, Websphere to JBoss - mammoth costs are being stripped out of the IT value chain and IT buyers are loving every minute of it.
What does it mean for start-ups, entrepreneurs, and venture capitalists? Low cost markets respond by consolidating capacity and driving volumes, as companies look to get to minimum efficient scale. Certainly the M&A frenzy of the last 18 months supports the view that large companies are realizing that profits will increasingly be a function of market share. Consolidation, while good for exits in the near term, has troubling long-term implications. How will start-ups extract value from customers who by definition are looking to reduce proprietary solutions and lower barriers to entry/substitution? How will start-ups break into companies who are centralizing spend with IBM, MSFT, and CSCO?
Some suggest the answer is increased verticalization - build solutions that create value by solving discrete pain points for given industries. But I think that is what IBM Global Services is trying to do! Others focus on leveraging the trends in open source by funding the next Red Hat variant. This is an explicit recognition that the software model is dead - ie you are willing to trade away license revenues at 85% gross margin for services and maintenance revenues of 40% GM and take 80% of the revenue (ie license revenue) and see it go to 0%.
While both are somewhat valid, the answer has to be innovation. If we cannot innovate and create unique, proprietary products (not in the sense that they do not leverage standards but in the sense that they are unique and defensible) then we will never see healthy returns for all the risk and capital we are deploying. It is expensive to take enterprise products to market and with the current IT mindset, plethora of comparable vendors in the market it is proving to be very challenging to break through the noise and incumbent account control and to get an IT buyer to buy from a small company.
Innovation - we must create products where the value (ROI and strategic) is so compelling that the customer is willing to introduce a product with switching costs into their environment. Unless we can eliminate the pool of ready substitutes through real innovation, the Golden Age of IT buying will continue unabated and we will see feature-based M&A deals (<$100m) where the major consolidators pick up missing functions (and have a choice of 5-6 to choose from). The days of breakout companies that change the enterprise IT game will be gone.
So my sense is that the answer is not a business model (service versus license) or open source (give up license to get maintenance) or verticalization (we know more about this problem than anyone else), but a continued focus on technical innovation combined with strong product management skills that translate engineering achievement into applied value and resonance to the customer problem.