Product Management: When to add a new feature

When do you add a new feature as opposed to improving the existing feature?

Past week I had a chance to spend time with a bunch of customers.  It always gives me a new perspective, a perspective that really matters to the company, i.e. the reality.  

Customers care about their unique problem.  They might not be at the stage where they can make use of all  the product features.  Their users may not have been trained on all the features, they may not have the browser version that supports the latest feature, or their internal roles may not have been set up the way product was designed to support.  Or they may not even know about the capability of the product.

Each customer is at different adoption curve.  They have a different set of problems they are dealing with.  When they are having user adoption problem, they are not interested in hearing all the great features that the product has to make it easier to scale the roll out.

Because of the unique customer perspective, many customers have their wishlist.  It is a list of enhancements to make their jobs easier.

In addition there are other new features that you are working on.  These are bets that you are making to solve customer's problem in a new way.  By definition these are features that no customers have asked for.  You are not even sure whether it can be used by customers.  (If you are sure it is going to be a success, either you are smoking something or a real beginner PM.)

Then the question arises.  As product manager, how do you decide when to work on improving existing features versus working on a brand new feature?

It depends.  There are times when improving existing features makes total sense and there are also times when investing on a new feature makes sense.  That's because expected return is different based on the product maturity and competitive pressure.

If you have a relatively new product that is looking for customer validation, it makes all the sense to invest resources in improving the existing features because they are not yet proven to be useful to customers.  If you have a mature product that is getting commoditized by competitive pressure, it makes sense to add a new feature to differentiate your product from the rest.

Key to which strategy to pursue doesn't end with a decision to invest in existing features or building a new one.  It must also include how to measure the success once the strategy is implemented.  If the feature is enhanced per customer's request, there must be a follow up with the customer to ensure that they see the value.  It's a great opportunity to build relationship with the customer who requested the enhancement and earn their trust.  You would want to create metrics to ensure that the enhancement can also be applicable for the rest of customers, and measure them.

Same is true with a brand new feature.  If decision was to pursue a new feature, then its adoption and usefulness must be tracked so that it can be improved with each iteration.  Make sure that you include a few customers in your study so that you get a view of the reality.

Note that introducing a new feature is a lot riskier than improving existing ones.  I recommend that you consider the following points before making the decision:

• Keeping customer happy and not losing them is 10 times easier than winning them over once they leave to competitor's solution.
• Adding a new feature and getting it adopted tend to get underestimated.  It is a factor of magnitude more expensive than improving existing features.
• Pursue *only* the enhancements that can be widely applicable; pursue *all* enhancements that can be widely applicable.
• Do cheap enhancements first; not all enhancements yield the promised user experience benefit.
• Understand what enhancement the customer is looking for and why; there may be other ways to solve the problem.
• One easy test to validate the viability of enhancement is to sell the idea to other customers.

Hope this has been helpful to you.  If you have comments or feedback, please let me know at jaeho9kim (at) gmail (dot) com.  If you are on Twitter, please follow me: @jaeho9kim.