Story
On August 10th, 1628, the Swedish warship Vasa embarked on its maiden voyage. It was, according to many nautical experts, one of the most spectacular warships ever built. Designed to be the flagship of the Swedish navy, the Vasa carried 64 heavy bronze cannon and a crew of 300. It also came equipped with hundreds of sculptures and other artwork by some of the finest wood carvers and craftsmen of their day. The ship was said to be a magnificent sight, so magnificent a sight that it cost more than five percent of Sweden’s gross national product. The problem was that the ship was only designed to carry 34 cannon, half the amount it launched with. This doubling of cannon required the addition of a second deck, and the original design certainly did not account for the sheer volume and weight of all the lavish ornamentation it ultimately received. The result, the Vasa, on its maiden voyage, sank in the Stockholm harbor after traveling less than one mile. The cause, extra guns, decks, and carvings added during construction compromised its stability. Feature creep literally sank the ship.
Takeaway
What is feature creep? Feature creep is a continuous expansion or addition of new product features beyond the original scope. It’s one of the most common causes of cost and schedule overruns. It usually occurs because features are easy or convenient to add, this is particularly common in software development, features accumulate over multiple generations of a product like this example of Microsoft Word with all of its toolbars visible, or features are added to appease internal project stakeholders. This is often due to what’s called the internal audience problem. Designers or marketers think they know what’s best for the customer, but they’re often wrong. The key driver that underlies all feature creep is the belief that more is better, and therefore, features are continuously added and rarely taken away. But this is flat out wrong. Adding features adds complexity, and complexity costs. For example, a study commissioned by Phillips Electronics found that at least half of the returned products had nothing wrong with them. Consumers just couldn’t figure out how to use them. Best case, feature creep changes the scope of a project, increasing time and cost with nominal impact on performance and the customer experience. Worst case, it has unintended performance or usability consequences. This is what sank the Vasa.
So what to do about all of this? As they say, recognition is the first step to recovery. So step one is to be on the lookout for feature creep in design and development, and educate your peers about the trap. Feature additions and changes typically come in little bits and pieces, and so it’s easy to get ensnared. This is why it’s called feature creep and not feature explosion. Step two, ensure features are linked to customer needs and are not added out of convenience or appeasement. Whenever someone proposes a change just because it’s easy or they know what’s best for the customer, give them a link to this movie. Step three, when doing updated versions of a design, ask not just what can be added, but what can be subtracted. Every good product release should subtract as well as add. And step four, create a project milestone to formally freeze the product specification. Freeze means no more changes. Shout feature creeper at anybody attempting to add features beyond this point. So whether you use your knowledge of feature creep to help keep your projects on time and on budget, to incorporate feature subtraction as a formal part of your product update cycle, or to avoid the unintended and sometimes catastrophic consequences of scope changes, remember the next feature you add could well be the feature that breaks the project’s back.