How many Wiliam employees does it take to change a line of code?

“Can’t you just implement this? It’ll only be 5 minutes worth of work.”

Is a famous line that is said periodically throughout the lifecycle of a project. Where it be adding in a simple tracking script or changing the checkout process so it doesn’t purely rely on Ajax, sometimes 5 minute features actually are quite time consuming, not to mention risky – especially when the project is due to launch!

Let’s walk through an example: Adding in a simple tracking script.

• It will take one producer to write clear, idiot-proof instructions on where to place the tracking code on the page and what metrics the developer will need to gather in order to generate the statistical data that the client would like.
• It will take one developer to quickly read the documentation around implementing the tracking code, ensuring that it is stable, fast and suitable for the site they are applying it on.
• It will take one developer five minutes to implement the tracking script, assuming he or she has complete knowledge of the project and technical architecture surrounding it, including all debts, liabilities and ‘gotchas’.
• It will take the same developer or a product tester to test the main (if not all) functions of the site to know that they have not caused any repercussions.
• It will take one code reviewer to review the code that was implemented to ensure that it is implemented in a way that is not harmful to the overall performance of the application.
• It will take one producer to run a third check to ensure that it meets the clients’ requirements.
• It will then take the client to test, sit and wait for the tracking script to collect data before analysing the results.

Not to mention if something goes wrong, then the original developer must be pulled and perform an investigation to check what the root of the issue is!

So through no fault of your own, the simple 5 minute feature turns into an hour-long process (if you’re efficient). The really important feature that you were hoping to get implemented before launch will have less of a chance of getting done on time!

But what if the developer was scheduled on another project? And now that project is slightly overdue?