The biggest problem is typically overly ambitious expectations combined with undercapitalization.

The single greatest business opportunity that is now emerging in the global marketplace is the ability to analyze digital log data to trace digital actions and from those traces to develop algorithms that can predict future outcomes with greater accuracy.

Systems are corporate funded mechanisms for increasing efficiency; programs are user funded mechanisms for increasing effectiveness. Programs should generally be charged back to users, systems should never be. Allocating corporate overhead to the operating units is simply a mistake.

Smart mobile phones connect you with 1 billion users worldwide, basically for free - you don't pay for the phone, you don't pay for the Internet, you don't pay for the wireless connectivity. Social networks let you add a new customer or a new agent, again for free.

To fix the business, to bring it back to health, you must assimilate enough of the disruptive innovation to modernize the operating model without jettisoning your business model. This typically requires new leaders and definitely requires new (if temporary) rules. The CEO is the only person who can dictate the correct terms in a timely manner and maintain the enterprise's commitment to those terms for the duration of the rehabilitation effort.

The danger is to cling to comfort and custom at a time when events demand breaking away from both. But it is also foolish to jump at every startling moment. Darwin selects primarily for prudent fast-following.

When you are incubating new ideas, "Don't put all your eggs in one basket" is very good advice. But when you are seeking to transform your enterprise's portfolio by scaling a fledgling business to material size - say ten percent of total enterprise revenue - then it is imperative that you make that the singular focus of everyone in the enterprise for the two to three year period it is likely to require to reach its tipping point. Expecting to do two such scaling efforts in parallel is simply folly, yet that is what the "eggs/basket" idea is often used to justify.