Scrum Principle : The Drum Beat of Regular Incremental Deliverables : Credibility from Customers, Confidence within Teams. When projects constantly start and restart both customers and teams lose confidence in each other.

Once upon a time there was a software company that hadn’t been able to put out a new release of their product for over 12 months. The company specialized in real time news services, and this product was an online version of the news that was accessible to anyone through the Internet.

We initiated daily Scrum meetings to get insight on what the team was doing with their time. Every team member was doing at least four or five different tasks and working on at least two features. After one week of Scrums, most of the team was working on different features, the priority having been changed.

We investigated how work was allocated. It turned out that anyone senior to the engineer on the team was able to change what that engineer should be working on … many people in the company had differing visions on what should be in the next product release, and what those features should look like. The meeting point for these disagreements were the engineers, who would take direction from the most senior person that gave them direction last.

The engineers were demoralized because they never got to work on anything long enough to complete it, and they knew that they were falling behind in the marketplace. All they had to do was open up their browser and compare their product to those of their competition.

A backlog list was established. The backlog consisted of all features and facilities that were planned and desired for the product. We got marketing to assign one person, the product manager, as being the only person that would manage the backlog. Anyone could ask him to make changes, but only he could actually change the list.

We then held a Sprint planning session. In the session, everyone talked about the product vision, where the competition was, and what needed to be done. There were disagreements, of course. However, after the session, the product manager organized the backlog list based. This established and confirmed product ownership … one person was responsible for the product vision and had the authority to employ engineering resources to deliver it.

A second Sprint planning session was then held. The engineering team took as many features from the top of the backlog as they felt they could deliver in one 30 day Sprint. They defined the design that would deliver these features in 30 days, and came up with what each one of them would have to do over the next 30 days.

We started the Sprint. Daily Scrum meetings confirmed that the team was making good progress on their work. The primary impediments that we had to remove were requests that they do work outside of what we had defined for the Sprint. This work included making presentations, flying to customer sites, adding new features, redefining current features … things that hadn’t been defined as top priority by the product manager.

When protests were raised, we asked which was more important, side assignments or the Sprint’s release. Every person felt that their request was just a small diversion that, of course, wouldn’t affect the overall progress. What they didn’t see was that they were a diversion, and that engineers need to focus on a problem to finish it. If engineers are constantly diverted, the refocusing time kills commitments. They also didn’t see that they were only one of many sources of diversions. Scrum got rid of the diversions and provided discipline.

At the end of the Sprint, the completed work was demonstrated. The engineers had not only completed the selected features, but they had been able to implement them using a new design that was much more user friendly and used advanced Internet features. They attributed this to their being able to work closely without interruption for the 30 days.

The entire marketing department, several executives, a salesman, engineering management, and the team were at the demonstration. They discussed how to build on the work. In the process they realized that, since the work didn’t have to be delivered to anyone, but only put up on the production server to be made available, that the results of each Sprint could be deemed a new release. So they put the Sprint demonstration up on the production server and completed their first new release in 13 months.

Scrum provided a necessary discipline. Engineering is a valuable resource. To the extent that their productivity can be maximized, interruptions minimized, and results produced, the organization is more competitive. To the extent that marketing can focus its directions to engineering based on competitive analysis and product positioning, the enhanced productivity can be harnessed to beat the competition and gain market share.