Tag: algorithm
Connect With Us
Stairway to (Product) Complexity (a.k.a. Why Do I have SO Much Stuff!!!)
In the last post I introduced two ideas about the sales history of any product. First, the number of unique configurations, or build combinations, depends on which features are included in its description. Second, the number of unique combinations drops whenever a feature is removed. A natural question to ask at this point is: Which feature, if it were removed, would lead to the greatest decrease in the number of unique configurations?
This is an easy question to answer, if we have a way of counting the number of unique configurations in any input set of configurations. This is really just a matter of sorting, so suppose we have such a counting algorithm. We try removing the features, one at a time. In each case we apply our counting algorithm to get a score. The score is the number of surviving unique configurations. The feature with the lowest score is the winner (like golf).
Now suppose that we permanently remove the winner, and repeat the contest again. This will determine a second winner, which can also be removed. We keep repeating until there are no features left. Now imagine a graph with the number of features removed on the x-axis and the number of surviving unique configurations on the y-axis. This is the stairway to complexity.
The stairway shown above is for a product (commercial stoves) with 23 features. The number of unique configurations starts at 1223 and drops to just 1 as the features are removed. The features are removed by looking for the biggest drop at each step.
Abstract:
Product complexity is driven by large number of options. Companies struggle to determine which feature choices are driving complexity. They typically “randomly cut choices” to streamline and rationalize SKUs. The cost of product complexity is tremendous on engineering. The current PLM systems do not have a method to measure this and provide intelligent feedback to engineers on how to standardize platforms to reduce engineering and maintenance costs. This article clearly details the metrics around product complexity and how to solve this issue.
Comments Off
The Number of Choice Combinations Depend…
The number of choice combinations depend on which product features are included. The build combinations is the product mix or the marketing mix.
Let’s consider the sales history of our product. There are two very important numbers: the number of units sold and the number of unique configurations. The number of units is well defined, but the number of unique configurations is ambiguous. The ambiguity comes from the fact that there will be more unique configurations if we use more features, especially soft features, to describe our product.
One very special soft feature is a Serial Number, or VIN (Vehicle Identification Number). The whole purpose of the Serial Number is to make each instance of the product unique. So if we look at our sales history and include Serial Number, we will see that the number of unique configurations is exactly the same as the number of units of the product (instances).
If we want to begin to understand the demand for our product we have to see which instances are actually the same. That means we have to get rid of the Serial Numbers. When we do, the instances collapse into groups of now unique configurations; that is, unique without Serial Number.
If we are interested in the tangible features of the product, then we may want to take out other soft features as well. Geographic region is important for some purposes, but may be a distraction when we are interested in the physical product. Taking out the geographic region feature will cause another reduction in the number of unique configurations. The red, V8, convertible in Florida will get combined with the red, V8, convertible in New York.
Sometimes we are interested in the variants of our product ignoring color. We know that every real variant is going to come in several colors, but we want to look at the product without the distraction of color. This is sometimes called the “body in white”. So the red, V8, convertible and the green, V8, convertible collapse into the V8, convertible.
The point I am making is that the number of unique configurations depends on which features are included, and this number drops whenever a feature is taken away. Mathematically, this is called “projecting out the feature”.
The number of unique configurations is at most the number of units sold, and at a minimum it is just one. If we take away all of the features, then every unit looks the same, which means just one configuration. There is a path from one extreme to the other that we will introduce next time.
By the way – understanding this is important as product complexity is a key driver of process complexity.
Comments Off
The Entropy of a Coin Toss.
A product is a collection of features, and each feature has mutually exclusive options. If a feature has only two options, then the choice is like a coin toss. The information contained in that choice is measured by entropy.
Entropy is a concept from classical thermodynamics that deals with the amount of disorder in a physical system (see http://en.wikipedia.org/wiki/Entropy). It was extended to information theory by Claude Shannon (see http://en.wikipedia.org/wiki/Entropy_(information_theory)). Shannon used entropy as a measure of the amount of information in a message. The simplest example is a coin toss. If we toss a fair coin, there is a 50% chance of getting tails, and a 50% chance of getting heads. Shannon defined the outcome of this experiment as having an entropy, or information content, of one bit. If I send a message (say 0 or 1) to tell you the result (tail or head), that message contains one bit of information.
Things start to get interesting when the coin is not fair. Consider a two-headed coin. The tossing experiment always results in heads, and the message will always be 1. According to Shannon, the information content of this message is zero.
If the coin is weighted so that the probability of tails is 25% and the probability of heads is 75%, then Shannon assigns an entropy of 0.811278. There is some information in knowing the outcome of the coin toss, but not as much as for a fair coin, because we already know that it will probably be heads. The graph below shows the entropy as a function of the probability of getting heads. When this probability is zero or one, the entropy is zero. The entropy reaches its maximum of one when the coin is fair (50%).
Where did the 0.811278 come from? How is the entropy actually computed?
We can’t answer this without introducing logarithms to the base two. In English, two to the third power is eight, so three is the logarithm of eight to the base two. We can write “blog” to mean log to the base 2, or binary log. If p denotes the probability of heads, then entropy is computed by the formula:
Entropy = -p*blog(p) – (1-p)*blog(1-p).
Logarithms to the base 2 arise naturally because one coin toss (2 outcomes) has entropy one, two coin tosses (4 outcomes) has entropy two, three coin tosses (8 outcomes) has entropy three, and so forth.
Comments Off
Variation is valuable
Advances in interconnection technologies are driving an increasingly demand-driven market. Customers are learning to expect to get what they want, when they want it, how they want it. And they tell you in each and every interaction they have with your company, or not. In a demand-driven world, increasing product variation and complexity in your business model is inevitable. Left untended, your business can become a tangled web of counterproductive business strategies with a dense portfolio of product families comprising thousands, even millions, of variants.
However, make no mistake, variation is valuable. To deny complexity or view the long tail of product variation as a management failure is to deny diversity of the world in which we make our living. Eliminate complexity in your product offer and you will find yourself competing with boatloads of product from China, India or any of a number of low-wage production markets.
The “keep it simple” principle is the root of good management. However, as Oliver Wendell Holmes, Jr. has observed, “I would not give a fig for the simplicity this side of complexity, but I would give my life for the simplicity on the other side of complexity,” it matters which form of simplicity you choose. The wrong simple answer is to try to focus on the 20% of product variants that make up 80% of your revenue, the head of the ubiquitous Pareto distribution, and find ways to minimize or eliminate the so-called unprofitable remaining 80% of product variants that lurk in the tail. Hello commodity, goodbye margins. The right simple answer is to deliver Intelligent Variation based on the voice of the customer shouting through the many interactions they have with you each and every day.
Comments Off
Help the sales team help the customer
This morning I was talking to the VP of business process improvement for a company that sells industrial machinery. Their products are highly configurable. She told me that every year they have 50% new configurations they have never seen before. The number of choices on their products has grown over time. ”A salesperson can’t know everything about the product,” she said. “Customers want a few choices, and before you know it, the quote has crept into a configuration that’s bad for the customer and bad for us. “
As the VP explained, the biggest opportunity for complexity management is at the point of taking an order. A customer wants to be guided to complete their order. This concept is called Demand Shaping. There are myriad ways a configurable product can be ordered. However, each customer cares only about a few features that are of high importance to him or her.
Understand product choices to manage complexity
A product is a collection of features, and each feature has alternative options. Understanding features helps determine strategies such as late staging. Some features are tangible, material things about the product: which engine, how much memory, Bluetooth. Other features are abstract or soft, like geographic region or sales channel. Among tangible features, distinctions can be made on the basis of the degree of postponement possible.
Pin-on features can be added to the product at the last minute, after a specific order is received. The classic example is the power cord for a printer. Hewlett-Packard avoided having different printers for different countries by attaching different power cords to a common printer.
Reconfigurable features can be changed after a real customer order is received. There are literally hundreds of different kinds of tractor tires, depending on the work a tractor will be used for. A tractor has to be built with some kind of tire just so it can be driven off the assembly line, and it’s easy to change the tires to suit the customer.
Line features, by contrast, are so basic to the product that they can’t be changed, such as the chassis or transmission for a vehicle or the motherboard for a computer.
Abstract or soft features are really attributes of the order rather than of the product itself. But they may be very valuable in understanding customer demand. The pattern of choices for tangible features may vary considerably by geographic region, which is a soft feature. For example, engine block heaters are popular in North Dakota and convertibles are popular in Florida, but convertibles with engine block heaters are almost non-existent.
Comments Off
Extending the product configuration to gain insight
One of the most important components in choice complexity is the product configuration itself, the mixture of product options that give a product its unique signature. Obviously the typical product orderable options are needed to analyze the complexity of a product, but other more abstract options can offer surprising insights into product and customer behaviors.
A typical car configuration has options such as sedan, V6 engine, automatic, blue, cloth, AM/FM/CD, sunroof. But more abstract items can be recorded along with these to offer more insight. Sales type can be recorded to analyze what types of product configurations sell better in promotional sales events as opposed to normal sales transactions. An attribute to record an extended factory warranty option may provide new ideas for packaging options together with additional warranty services that customers are moving towards.
Comments Off
How many choice combinations does your product have? That depends.
Possible combinations
This is a question with several answers. The easiest answer is the least useful. The number of possible build combinations, or unique configurations, is easily computed by multiplying the number of options for each feature. For example, if your product has feature A with 3 options, feature B with 2 options and feature C with 4 options, then there are 24 (3 x 2 x 4) possible build combinations.
These numbers grow very rapidly. If you have 5 features, each with 4 options, there are about 1,000 build combinations (exactly 1,024). With 10 such features, the number of combinations is about 1 million (1,048,576), and with 15 features it is over 1 billion (1,073,741,824).
Comments Off
8 more definitions you need to know for product complexity analysis
1. Kit
A kit is a collection of parts that are used together for some purpose — for example, all the parts needed to implement air conditioning on a particular model of a car. A kit is assigned its own part number.
2. BOM
BOM stands for bill of materials. When a customer makes a selection of choices chooses a configuration (i.e., makes a complete set of option choices), the manufacturer translates the order into a collection of parts that are needed to assemble it. The BOM is expressed in terms of part numbers. These part numbers may refer to whole kits, composite parts or specific atomic parts. A complete vehicle, or washing machine, will contain many parts that the customer has not chosen. But these parts appear in every instance, or else they are implied by the combination of choices that the customer made.
Comments Off