Snowman Practice Overview

Snowman Value Proposition

The Snowman Practice focuses on the challenges of building large complex IT systems. No other existing methodology addresses these challenges. The result is that most large IT systems are delivered late, over budget, or not at all. Those that are delivered have functional, security, performance, and reliability problems.

The Snowman Practice breaks these large systems down into small self-contained pieces that are well aligned to the business, highly secure, very reliable, and much less expensive to deliver. These self-contained pieces are called Snowmen because they are strongly vertically partitioned and because each consist of three stacked packages of related closely related functionality: business functionality (the head); technical functionality (the torso); and data functionality (the bottom).

Snowman Philosophy

We need to make large IT systems as simple as possible. Simpler is cheaper, more flexible, and has more robust features than complex. The simplest possible large IT systems are those in which the subsystems are designed as Snowmen, that is, autonomous subsystems that have synergistic collections of functionality, well defined points of interoperability, and deep partitions that extend through the technical and data layers.

Snowman Analytics

Mathematical formulas are used to compare two or more proposed designs with respect to their deviation from the ideal Snowman design. The measurements can be used to predict cost, risk, and overall robustness of future designs very early in the design life cycle, when changes are still easy and inexpensive to make.

The Snowman Identification Process (SIP)

Optimal partitioning of a large IT project into Snowmen results in dramatic reductions in organizational complexity. That is the good news. The bad news is that achieving this optimal partitioning is very difficult. Using standard non-directed approaches, it is impossible. And even a small error in the partitioning process result in a large complexity penalty. 

The Snowman Identification Process (SIP) gets around this problem by applying a directed approach that is grounded in probability, partitioning, and complexity theory. It allows one to dissect a business problem, identify a representative sample of business functions, and partition those functions into a highly optimized Snowman design. The result is not only optimal, but can be verified and validated. SIP is in many ways the heart of the entire Snowman Practice. 

The Snowman Architecture

Once we have optimized the partitioning through SIP, the Snowman Architecture dictates how that partitioning is driven deeply into the technical and data architecture. The overall result: a design in which simple and autonomous IT systems closely reflect the business systems they support.

The Snowman Design Engine

The Snowman Design Engine is a software tool that guides the use of SIP, automating many of the algorithms and mathematical formulae used in the partitioning process.

In summary,the Snowman Practice drives the next generation of IT systems that are simpler, more efficient, and highly effective at solving business problems.

Page last updated 6/6/2013