Stand Back and Deliver: Accelerating Business Agility (Roni Dover's Library) by Pollyanna Pixton & Niel Nickolaisen & Todd Little & Kent McDonald & Jim Lewis

Author:Pollyanna Pixton & Niel Nickolaisen & Todd Little & Kent McDonald & Jim Lewis
Language: eng
Format: mobi, epub
Publisher: Addison-Wesley
Published: 2009-03-12T16:00:00+00:00

Case Study: The Swiss Stock Exchange

The Swiss stock exchange was preparing to enter electronic trading. There had been two prior attempts to develop the software solutions. The first attempt was an internal development effort, and it failed. In the second attempt, the project was outsourced to a prominent consulting firm; it failed as well. In both of these prior failures, everyone thought that they were doing things by the book using software development best practices.

The third attempt was getting under way and there were real concerns about the prospects for yet another failure. Pollyanna got involved on the team and eventually ended up leading the project. The challenges were significant. The team consisted of more than 120 people, and 60% of those were contractors. There were multiple customers, as every Swiss bank (approximately 50 in total) had financed the project, and each bank had different requirements and desires. The existing systems were 20 years old and upgrades were drastically needed. Clearly, this was a very high-profile project with a lot of money at stake.

On the positive side, the project attracted the best talent and was able to specify its delivery date within a reasonable time frame. The 2-second throughput requirement, total software and hardware redundancy, and “lights out” maintenance model all seemed doable. A retired trader was heading up the functional specification team—a sort of on-site “customer.” Things looked better than before.

But there were other ominous signs that history would repeat itself. The team in charge of building one of the key server-side components had developed a two-year plan detailing tasks down to the level of every 15 minutes. What were they thinking? Creation of such a schedule is over-planning even for a project with almost no uncertainty. The team leaders had fallen into the trap of the illusion of control—even though they actually had almost no control. Instead, they spent an inordinate amount of time updating the detailed plan and trying to justify how they would get back on schedule.

Pollyanna had the team members move to a system that would allow them to update and reprioritize the features and the work tasks on a weekly basis. The team leaders reviewed progress at the end of each week and then, based on this information, reprioritized the next week’s work. Every six weeks, the teams re-estimated and re-planned tasks for the next three-month iterative release. Change was inevitable and was managed through the reprioritization process. All team leaders made the decisions together.

Next, Pollyanna worked with the trading system team to build the workstation where the traders would make their trades and gather information. The trading system was associated with a lot more uncertainty because it was brand new and would be more intimately connected to the end users. To get it right, the developers formed a user group of traders and built prototypes to obtain useful feedback. It took four prototypes to get a system that the users liked—but what really mattered is that the users were happy with the final delivery.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.