In the past I managed a lot of projects in ecommerce. Most of the time it was about building or optimizing online shops. This blog post summarizes some of the techniques that I am using for making it possible to deliver the requested quality on time and on budget.
Further I will bust some myths and explain how you should behave.
I do project management according to IPMA. Even if many companies are currently investing a lot in changing their methodology to Scrum I still believe in the good old waterfall model. You need scrum if you are supposed to deliver an undefined thing. If you have to deliver something that is quite well defined like an online shop, it does make sense to use a waterfall approach.
The magic triangle
In project management a lot of variables influence each other. The most important variables are time, costs, and quality.
|Time||Time means the ending event of the project is on time like it was planned for in the beginning. This means that the critical path is on track|
|Costs||Costs means that you need not more people nor more resources like materials or travel expenses than you expected. Especially unpredictable risks like sick participants, weather or the stock market might create some trouble here|
|Quality||Quality means that your deliverables are in a good shape and will not cause issues when they are used|
They are interconnected. If change requests flow in during your project at least on of these variables has to change. If the project sponsor does not want changes for these variable e.g. make the ending stable then he or she should not allow change requests.
Myth 1: Training does not pay off during the project
Especially Tom DeMarco wrote in his book The Deadline that there are no fast ways to boost developer productivity. I disagree, there are some easy techniques like daily deployments or unit tests. If you teach your team these techniques, and they use them, it will pay off within 48 hours.
In the following competition of web development platforms https://www.plat-forms.org/ the researchers of the FU Berlin figured out that teams that do a lot of unit testing were already more productive after 48 hours based on a quality assessment than their competitors.
Myth 2: Agile will solve all problems
Everybody things that an agile project methodology is less complicated compared to the waterfall model but if you look closely then you will figured out that e.g. scrum just means to go through the whole waterfall process as quickly as possible.
If you look at the chaos report of the standish group you will see that the main reasons why projects overrun their budgets is not that the project management methodology sucks but that the end users and the requirements are not well enough defined.
Another big problem is the running time of a project. It is true that you have an exponential growing rate of project failures if the projects run longer then 6 months. This was proven by some research of IBM. Further, in average the project targets are changing around ~8% every month. It is very important that you get your project done in time or cancel the project in time. A project that is already late is very likely to overrun it’s budget by more then 2x – 3x times without any results.
Myth 3: Overtime helps to keep the project timeline
I recognized that a lot of project managers use overtime as a project management tool to keep a deadline. In all 20 projects I was involved so far it was never the case that overtime helped to keep the deadline. The deadline was always postponed despite overtime.
One big reason why this is the case is that people who do overtime e.g. working 60 hours a week are less productive then people working only 40 hours. The same is happening for holidays. Germans have a lot more holidays then americans. Nevertheless they are still more productive.
Here is a small table comparing vacation days to productivity per hour:
|Country||Avg. vacation days||Avg. hour productivity|
As you can see even with a lot of vacations you can be very productive.