Konsensys FeatureRequest^TM

THE PROBLEM

When customers, sales, marketing and engineering make feature requests, the requests generally goes into a "black hole" (things go in, but nothing ever comes out). Until, the next version of the product is released, customers have no idea whether the request was a considered a high or low priority -- or even read. The lack of feedback has an adverse impact on satisfaction.

We care less about about whether our suggestion is actually implemented than we care about being valued enough that our suggestion is given careful consideration. Konsensys FeatureRequest^TM immediately calculates the "rough" priority of a request and shows how it ranks compared to other requests. As product management and engineering rate the requests against other factors, customers and staff can "drill down" to view the reasoning behind the prioritization.

Customers and staff see that their requests were carefully considered regardless of whether it was chosen to be implmented. With Konsensys FeatureRequest ^TM, every customer and staff person experiences their requests being heard and carefully considered.

A complex product (for example, a software product) may have 100 to 200 requested features for the next version of the product. There are typically only enough engineering resources, time-to-market constraints, cost constraints, etc. to implement about a third of the requests. Thus, every producer of a complex product must prioritize the feature requests to decide which requests get implemented, deferred or denied.

The producer of a complex product must rate each requested feature in terms of it satisfying such customer benefit categories as: product purpose, generality (usable by more people in more circumstances), usability (ease of learning/setup/use), performance, reliability, safety, economy and aesthetics. Further, the requested features must be weighed against such company internal factors as: risk to schedule, improving product positioning, improving the return-on-investment proposition, improving the infrastructure (features not visible to the customer that make it easier to fix and enhance the product in the future), etc.

Weighing 100 requested features against 10 categories of customer benefit and 6 company internal factors is calculating 1600 weights and another 1600 calculations to sum the weights. No human can estimate all those weights and make all those calculations in their head. Thus, we all naturally weigh some factors and forget about other factors.

As a consequence, you can pick experts from any two competitors and they will have very different opinions on the relative importance of product features. Worse still, experts within the same company will differ in rating a product feature's importance. In other words, the statistical inter-rater reliability is very low.

The very low inter-rater reliability implies two things:

1. There is very little consensus. The more "experts" in a company, the more contentious, political and drawn out the decisions on which features make it into the next version of a product.
2. The quality of their analysis is not statistically "reliable", i.e., you can not rely on the analysis being valid.

1. Ratings of various factors used in the decision making process are mostly subjective.
2. Different experts examine different factors.
3. Different experts use different methods of weighing the factors to reach a decision.

1. Tends to give similar ratings for the same factors, because the factors used can be objectively measured.
2. Examine a check list of factors to ensure thorough analysis.
3. Uses well known QFD-based calculations to reach a quick, high quality decision.