For decades vendors have offered one-size-fits-all infrastructure technology that gets "tuned" by in-house or outside consultants after it's deployed. Applications vendors have offered specific solutions to several industries for years. Manufacturing, financial services and healthcare all have their technology leaders but they tend to be on the applications -- not the infrastructure -- side.
Let's challenge the vendors to not only develop customized infrastructure solutions, but let's also get them to develop simple infrastructure technologies that satisfy specific industry requirements -- without the overkill that characterizes just about every infrastructure technology out there.
What am I talking about? Three things:
Vertical Industry Requirements
Vertical industries have mature relationships with information technology. It's not like we're still trying to figure out how banks should use IT to solve their problems. Vanguard -- the giant financial services company outside of Philadelphia -- knows exactly how to leverage IT on to old and new business models and processes. But it does almost all of it in-house. Like many companies, Vanguard has served itself with wide and deep vertical expertise. It has customized off-the-shelf software to satisfy its financial services requirements.
But why couldn't the vendors customize it for them? I realize, of course, that one can always hire an army of consultants to do whatever is necessary, but why bloat the process with more overhead than necessary? Besides, consultants are really, really expensive and once they've become part of the furniture, well, they've become part of the furniture. Aren't the financial services, pharmaceutical and the chemical industries, among others, big enough to justify their own infrastructure requirements -- and solutions?
Just about all of the major vertical industries have sets of requirements that are generalizable across the companies within that industry. Are Fidelity's requirements all that different from Vanguard's? Are Dupont's all that different from Rohm and Haas'? What happens when RFID really takes off? How many flavors will there be? Why wouldn't there by one primary flavor for each vertical industry?
I'd love to see the primary vendors of hardware, software and communications infrastructure develop full vertical suites complete with all of the bells, whistles and hooks that make it possible to transact business across any number of vertical industries. Issues like privacy, compliance, reporting, business-to-business (B2B) transaction processing, data representation, and security, among others, are approached differently by different vertical industries. Why can't we get some help here to at least create some intra-industry infrastructure standards?
The chemical industry is a case in point. Just about all of the major companies have standardized on SAP as their ERP application. This means that infrastructure providers (without the consulting middleman) can optimize their solutions to back-office and front-office SAP application modules. How easy is that? Vendors who provide communications, workflow, groupware, knowledge management, messaging, content management, security and data base management, among other capabilities can tailor their offerings to SAP and the chemical industry.
Are you tired of buying, deploying and supporting capabilities that you seldom, if ever, use? Human factors experts have told us for years that we use about 10% of the capabilities of Microsoft Office, but, of course, pay for 100%. Every time I use the phrase "bloatware" in a speech, the audience loves it. CFOs really hate the whole concept of "shelfware," where bloated applications never even get installed -- even though they were purchased as part of the enterprise licensing agreement.
One of the most interesting opportunities for open source providers is the possibility of simplifying hardware and (especially) software architectures. It's unlikely that the major proprietary vendors will travel down this path, but the Linux desktop crowd could move in this direction pretty easily.
But the real leverage still lies with the optimization of proprietary legacy infrastructure technology. While the open source community wants to simplify its offerings, real progress could be made if the proprietary vendors would lighten and verticalize their applications: we really don't need all the bells and whistles they shrink wrap; instead, we need applications stripped down -- and juiced up -- to optimally support specific vertical industries.
Let's push these people to customize and simplify our vertical worlds. Vertical industry infrastructure requirements are far from mysterious. Why can't they make our vertical lives easier?