Emcien defines product complexity as simply the ability to predict what the next order coming into the company will be.
Think about it: If you only made product configuration A, you have 100% confidence in knowing that the next order in the door will be configuration A (assuming you get an order in the door at all, not a total given in this economy). But if you have configurations A and B, it’s harder to know and with A, B and C, it’s even harder, and so on. When you have thousands of configurations, predicting the next one is very difficult.
It’s not just the number of configurations that’s important but also how they’re distributed. If I have 10 configurations but 90% of my orders are for config A, then it’s still safe to predict that the next order is config A. But having 10 configs that have each been ordered 10% of the time is extremely complex!
Each individual feature also contributes to complexity in its own way. For example, if you have a feature called “color” that has 6 options (such as red, blue, black, etc.) and another called “seats” that has 3 (deluxe, leather and cloth) and the distribution is such that most customers buy the red color but the seats are evenly distributed among the 3 choices (33.3% each), then the seat feature could be contributing just as much or more complexity than the feature color with 6 options.
Reducing complexity for a product line means looking into the complexity of the product’s features. And since each feature has complexity, it means starting with the most complex feature and working downward. If you don’t do this, you can end up spending tons of time reducing complexity to only realize months later that you still have new unique SKUs getting created each day. Those new configurations are coming from more complex feature combinations, but you focused on things that appeared complex but ultimately were more predictable than you thought.
Finally, to reduce complexity you need to simplify features that matter least to customers. Taking away the color red is much worse than taking away floor mats, for example, so ranking your features on how important they are to customers helps make sure you pull complexity out where it helps the most (reduced variability) and hurts the least (customer acceptance).
Comments Off