How Microsoft used a Common Business Practice to Turn Around their IT Failures

Microsoft manages over 40 thousand data center servers worldwide that run thousands of business applications, processing 40 billion dollars of sales annually. These servers provide an essential service and need to be available and reliable or else can cost revenue. Microsoft wanted to reduce the financial impact of server failures by reducing the instance of failures. They succeeded and significantly reduced failures by using one of the most celebrated efficiency-boosting methodologies in business. Here’s how.

The Strategy

Microsoft set out three phases to identify and eradicate defects that can cause server failures.

  1. Establish standards
  2. Measure standards and make defects visible to all levels of IT
  3. Eradicate defects

These aims seem too simple to drive the technical advancements needed to reduce server failures, but they managed to do so by using a system developed by Motorola in 1986. This system, called Six Sigma, reduced Motorola’s production defects to no more than 3.4 instances in 1 million opportunities. It had a deep impact on the business world, and earned Motorola international acclaim and awards.

The Six Sigma System

Six-Sigma uses data reviews to improve quality control and generate incredible results. Traditionally, it was composed of five steps:

  1. Define. This step is executed by a team of people with a Six Sigma expert leader. They define the faulty process on which they need to focus, as well as the goals and quantifiable result.
  2. Measure. The team then measures the initial process performance and develops a list of variables that could cause the problem.
  3. Analyze. The team isolates each variable and tests it as the root of the problem. This identifies the reason(s) for the problem.
  4. Improve. The team works to improve the problem(s) once they know the processes that cause it/them.
  5. Control. The team then adds controls to the problem processes to stop them from causing problems again. These can be alerts to engineers or even self-fixing controls.

Six Sigma has a great success rate, but it does require learned experts. Experts in Six Sigma are graded on a belt system. Master black belts are the highest grade, but even white belts know their stuff. Many companies are insisting employees obtain at least a white belt certification, as the impacts on company efficiency can be very valuable. Motorola’s Six Sigma University is not the only provider of expertise; sites like www.6sigma.us/six-sigma-white-belt.php are starting to offer online webinars.

The Results

Microsoft used Six Sigma across several teams. They prioritized the risk of failures on three dimensions: impact, likelihood and detectability. The Microsoft IT project team applied a special Six Sigma framework, called the DPMO framework, to continuously detect and eradicate the defects that were causing their server failures. Within a year, the business and IT teams worked together and reduced their defects by 60%.

The Microsoft Case Study is different to many in that the process they used can be executed not just by large corporations with deep pockets, but can be utilized by small to medium enterprises as well. With the efficiency benefits they boast, Six Sigma implementations can be well worth it.