Skip to main content

Why the "P Principle" Applies to Modern IT

By April 16, 2012Article

You’ve surely heard that classically alliterative (and originally profane) adage, “Proper preparation prevents poor performance.” This is particularly true in matters of IT, where scale and complexity promise the undoing of any IT shop that doesn’t have its house in order.

For too long, IT practices have been more art than science, a Rube Goldberg mess of automation held together by caffeine, Aspirin and heroic efforts. But IT is being forced to change — to “industrialize.”

Why? Infrastructure and operations visionary Glenn O’Donnell ably explores this question in the Forrester report, “Prepare for the Industrial Revolution of I&O,” which suggests that traumatic events force transformational change. For example, the Great Depression led to renewed focus on innovations that drove massive and sustained gains in agricultural productivity.

When you’re brought to the brink, you’re forced to think.

If you ask me, another important event that forced new thinking was the rise of online commerce during the 1990s (remember the silly season of stock options, sock puppets and groceries at your doorstep?) Despite the irrational exuberance of the time, it led to lasting productivity gains in manufacturing and distribution. Suddenly, markets were globalized, demand was fragmented and scale was utterly explosive. Old-line manufacturers and distributors were forced to retool from an opaque and siloed value chain to deep integration and automation, from raw materials to finished goods.

The highlights of this transformation:

  • Standardization – Diverse inputs were standardized as centrally managed and reusable subassemblies that were combined to create new products. Standardization drove economies of scale, quality and manageability.
  • Mass customization – These plain-vanilla subassemblies were combined and recombined to cost-effectively deliver innovative and differentiated products to serve diverse market needs.
  • Just-in-time delivery – Warehouses of costly finished goods inventories were traded for deep supply chain integration, detailed bills of materials and blueprints and automated assembly for demand-driven product delivery.

You can apply the same ideas to the industrial transformation in IT:

  • Standardization – Hardware and software infrastructure stacks are standardized and reused for economies of scale, quality and manageability.
  • Mass customization – Standardized software components are combined and configured to create new platforms, applications and business services to serve complex and diverse business and technical requirements.
  • Just-in-time delivery – Deeply managed blueprints that are integrated back through the software supply chain replace static templates and opaque images as the unit of management. Need a new image? Generate one on demand. Need to change a system? Drive the change through the blueprint.

If you think about it, enterprise IT has a strong genetic connection to manufacturing — and IT organizations today look a whole lot like manufacturers circa 1998. Today, the expectation is for flexibility, speed and scale, while also fundamentally altering the economics of what it takes to achieve these ends.

What’s at stake? O’Donnell puts it well when he says: “Combining the traits of quality at high speed and flexibility … yields benefits that are not just incremental, but dramatic. Such shifts can change entire industries and produce juggernauts at the expense of established leaders.”

But first IT needs to get its house in order, making the foundational investments in automation and engineered IT practices for the age of speed and scale.

IT organizations that get this right will make history as the new disruptors — the next Google, perhaps. Those that don’t will be the disrupted — a footnote in the dustbin of industrial history. (And, really … who wants to walk among sock puppets when you have the chance to walk among the titans?)

Jake Sorofman is founding partner of Marketlever, a boutique strategy and communications consultancy. Prior to founding Marketlever, Jake was CMO of rPath, a leader in cloud automation tools.

Copy link
Powered by Social Snap