Using Theory of Constraint to manage large Software Project

“The Goal” – a book written by Eliyahu M. Goldratt & Jeff Cox is reaching closer to 8 million copy sale. This is incredible. I am not sure how many business books have reached such a level. The sale clearly indicates something inherently great about this book. I found this book written in a story style. The concept of Theory Of Constraint ( TOC) is explained in a very simple language taking simple situations. I was thinking there could not have been better way to explain TOC concepts than the way explained by the authors. The examples are taken from manufacturing industry.

I am reproducing a short extract from the Preface of the book.

“Companies, even the relatively small ones, are incredibly complex. When facing complex systems, it is so easy to fall into the trap of looking for complicated solutions. Still, we all know that complicated solutions do not work! Instead, what we should do is realize that every complex system is based on inherent simplicity. The way to rapidly improve our systems is to search out that inherent simplicity and capitalize on it.”

I am sold on principle of simplicity. When I finished reading this book, I was struggling with a project of 50 persons with an estimated effort of 20 man years. The project was delayed. The completion date appeared like a moving goal post. It was challenging. Although TOC principles are primarily applicable to Manufacturing type of systems, I found it very relevant to apply the principle to solve the complex situation we were in.

We in software business talk about life cycles like Waterfall, Agile, Iterative etc. The principle of TOC can work in conjunction with any one of them. Let us look at the brief case study.

On 1st October 2009, we saw this project getting out of our hand. We brainstormed and brought major changes the way we managed the project. We looked at each element of project team. The team structure can be depicted as sequential block of related activities. Each block represented a processing station.

We noticed a lot of backlog, late working and low motivation at Integration station. UAT team, which was a Client team also, had a similar problem. The natural outcome was to define Goals for the project and structure the team around the defined goals. The goals were agreed as:

1. Improving throughput through the system.

2. Ensuring high quality of deliverables. It was thought to reduce UAT defects at least by 67%.

Besides the above goals, a working discipline was agreed to support the stated goals. Every station will work towards ensuring the utilization of critical stations ( Integration & Acceptance in this case) and the work allocation will be done to assist faster throughput and not ensuring higher utilization of the members in different stations. The work was structured in batches and batch movement across various stations was monitored and managed on above principles.

Implementing this discipline in the team faced resistance from both internal team members and the Client. A series of discussion were held with stakeholders and finally a concurrence from the customer was obtained for trying this out. We were confident to demonstrate the benefits within a week or so. This was used to make Client agree to watch the process for a fortnight.

After agreeing with the basic rules of the game, all team worked to support the goals. We carried through system of daily stand up call concentrating on three questions - what got done in last 24 hours, what are the plans for next 24 hours and what are the bottlenecks. We worked to solve the bottleneck.

The first batch with the new process was submitted on schedule. The result of the UAT was very encouraging and changed the morale of the team. People had renewed level of confidence. They were upbeat with the results. This changed the whole game. The theory was followed in word and spirit and in three months time the project got completed and accepted by the customer. Although we had a bad start, the excellent finish gave us excellent respectability from the client and great sense of achievement for all of us.

Software industry is gradually maturing. Time is getting apt for applying proven principles of manufacturing ( Factory Management Principles) in this maturing industry. TOC is one of them and I am upbeat with my experience on solving large complex project issues using this theory. It has worked well and can work with any life cycle selected for the project execution.