Spending eighteen months on a software project doomed to fail (either gets shut down or market irrelevance) is a waste of the important asset – your time. If you find yourself stuck on one of these you will pick up bad software development habits and find the innovation sucked out of you. Here are five clues to help you sniff out software project that should be avoided.
1.) Projects with Lots of Mass
My general rule of thumb is to avoid any project with significant amounts of mass in its early stages. By mass I mean lots of user research, lots of PowerPoint slides flying around, lots of executive attention, lots of developers. This chaos will generate even more chaos and usually failure. Some of the most successful projects at Yahoo! began as very small side projects: Yahoo! Answers and Yahoo! Sketch-a-Search. Find small projects with people (two three developers, a product manager, and designer) you respect with a track record of building good stuff.
2.) The Project with Three Names
If a project is not going well there is a trick to help it regain stature – rename it. It used to be called “Phoenix” now its “Marin” – problem solved! Before joining a project be sure to ask about the naming lineage.
3.) The Replacement Project
These are easy to detect. Somebody decides to follow a technical trend and in the name of flexibility, scalability, blah-blahbility decides to replace outdated an outdated technology platform with a new shiny platform (i.e the latest technology trend).
4.) The Project that Declares War
These are not hard to detect but hard to ignore, as their pitch inspires high emotional excitment. Declaring war on your archenemy naturally excites us humans. We like conflict. These projects are prone to failure because there is usually no clear angle of attack (see my post here). Rather, its lets build or sell a slightly better or cheaper version of the competitive offering. For example, if you are Monster.com trying to fend off Craigslist offering job listings for $50 is not going to do much other then erode your margins. The reason Monster is failing is because of poor applicant quality. Address that issue rather than chasing Craigslist.
5.) The Sacrificial Lamb Project
Projects that SVPs keep in their portfolio just in case somebody decides its time to start whacking some projects in the name of cost cutting. Because these are relatively rare they are also hard to detect.