Friday, July 29, 2011

A Universal Starting Point for Problem-Solving

By popular request, I'm going to hold my next Change Artist Challenge until next week to give some of my readers a little more time to catch up.

This essay should also be helpful to change artists, who often have to start their work by defining or redefining a problem that's presented to them. It's adapted from Chapter 5 of Exploring Requirements 1: Quality Before Design.

How can we reduce the great variety of potential starting points to a single solid platform for exploring requirements? A possible solution is to regard every design project as an attempt to solve some problem, then reduce each starting point to a common form of problem statement.

A problem can be defined as
a difference between things as perceived
and things as desired.

[For a full discussion of problem definition, see Donald C. Gause and Gerald M. Weinberg, Are Your Lights On? How to Know What the Problem Really Is.

Figure 5-1. A problem is best defined as a difference between things as perceived and things as desired.

This definition can serve as a template measuring each idea for starting a development project. If the idea doesn't fit this definition, we can work with the originator to universalize the idea until it does.

Universalizing a Variety of Starting Points
Let's see how this universalization process can be used to reduce six different starting points to a common form of problem definition.

Solution idea
Perhaps the most common starting point is thinking of a solution without stating the problem the solution is supposed to solve. In other words, the idea doesn't say what is perceived (and by whom) and what is desired, so it doesn't fit our definition of a problem. Here are a few examples we've experienced.

1. A marketing manager told a systems analyst, "We need sharper carbon copies of our sales productivity report." Rather than immediately begin a search for a way to produce sharper carbons, the analyst asked, "What problem will sharper carbons solve for you?" The manager explained that the carbons didn't make very good photocopies, so the salespeople had trouble reading them. "So," the analyst confirmed, "you need one clear copy for each salesperson, and you're now making multiple copies of the report we give you?" Eventually, through such give and take, the problem was redefined as a need to provide timely and clear comparative information to a sales force of four hundred—something readily accomplished by slightly modifying an existing on-line query system. The final design didn't have sharper carbons. It didn't have carbons at all. Not even paper.

2. In another case, a university dean said, "We need a way to attract more students." The dean never said why they needed more students, and each faculty member hearing the statement formed a different idea. Some thought "more students" meant getting more outstanding students. Some thought "more students" meant being able to support more teaching assistants in certain departments. Still others thought "more students" meant the dean wanting to fill the vacant dormitory space.

After arguing for months about the best way to get more students, the faculty finally learned what the dean really wanted: to create the impression in the state legislature that the school was doing a higher quality job by increasing the rejection rate of applicants, so the university appropriation would increase. Once this goal was understood, the faculty approached a solution in several ways, none of which involved an increase in student enrollment.

Technology idea
Sometimes we don't have a problem in mind, at all, but literally have a solution in hand: a solution looking for a problem. When tearing off those perforated strips on computer paper, have you ever felt there ought to be something useful to do with them? The perforated strips are the solution, and the problem is "What can we use them for?" After thirty years of searching, Jerry bought Honey, a German Shepherd puppy, and suddenly he discovered the problem his solution was looking for. Computer paper edges, crumpled up, make perfect litter for puppy nests!

When a new technology comes along, it's often a solution looking for a problem. The Post-It™ note developed by 3M is a conspicuous example. The semi-stickiness was originally just a failed attempt to produce an entirely different kind of adhesive. Instead of simply discarding it as another failed project, the 3M people thought of problems for which such semi-adhesive properties would provide a solution. They created Post-It™ notes, but the solution-to-problem process didn't stop there. As soon as Post-It™ notes appeared in offices, thousands of people began seeing problems they would solve.

Some of the high-technology companies we work with are dominated by this kind of solution-to-problem starting point. In effect, their problem takes the form of the following perception and desire:

Perception: We own a unique bit of technology, but others don't want to give us money for it. For example, a chalk company buys rights to a new vein of chalk that has exceptional purity and strength. To most people, however, chalk is just chalk.

Desire: Others will pay us a great deal of money for the use of this technology in some form. For instance, if the company can create the idea of Superchalk in the public mind, the unique purity and strength become an asset of increased value.

Such a problem statement allows the technology to become a kernel around which many designs can be built. Without it, technology firms often make the mistake of believing that "technology sells itself." Although this slogan may be true in certain cases, usually it's an after-the-fact conclusion. Want to turn a solution into a problem requiring it? Ordinarily, you'll need an enormous amount of requirements and design work. For example, how will you make teachers believe they can't really teach well without Superchalk?

Many product development cycles start with a variety of metaphorical thinking—a simile, or comparison, as when someone says, "Build something like this." Although the customer may emphasize "this," the job of the requirements process is to define "like."

For instance, Maureen, the leader of a software project, told her team she wanted a new user interface "like a puppy." "First of all," she elaborated, "people see a puppy and are immediately attracted to it. They want to pet it, to play with it. And they aren't afraid of the puppy, because even though it might nip them, or even pee on them, puppy bites aren't serious injuries, and puppy pee never killed anyone. Also, you can't really hurt a puppy by playing with it."

Although the team couldn't yet build a system with this requirement, playing with the simile did inspire them to ask probing questions. "How about housebreaking and obedience training for the puppy?" a teammate asked.

Maureen thought a bit, and said, "Yes, the interface should be trainable, to obey your commands, so it becomes your own personal dog."

"Okay," asked someone else, "will it grow up to be a dog, or remain a puppy?"

"That's easy," said Maureen. "It will stay a puppy if you want it to be a puppy, but if you prefer, it will grow up to be a real working dog doing exactly what you say."

"What kind of working dog?"

"A watchdog, for one thing. It should warn you of dangerous things that might happen when you're not paying attention."

Someone else got into the spirit by asking, "What about a sheepdog? It could round up the 'sheep' for you, and put them safely in the pen. And guard them from anyone stealing any."

By this time everyone was involved, and the requirements process was running like a greyhound, though not necessarily in a straight line, as when someone asked, "How about fur? Should it be a longhair or a shorthair?" Nobody could figure out what fur meant for an interface, though the question did lead to an extensive discussion of touch screens and other interface hardware they had never previously used. Eventually this tangent was clipped by someone observing, "Our tail is starting to wag the dog."

The simile is excellent as an idea-generation tool, but eventually the requirements group has to groom their ideas into prize-winning form, which requires some idea-reduction tools. You know when the simile has become a bit dog-eared when you can no longer make fruitful connections between it and your product. It's important, though, to keep it going a bit past the point where it becomes ridiculous, just to be sure you've generated enough ideas.

Many people do not consider themselves metaphorical thinkers, believing they think more concretely. In the requirements process, they would more likely say, "Here is a chair. Design a better chair." or "Here is ordinary chalk. Design a superchalk." In fact, the norm is also a metaphor, seeming literally "close" to the thing desired. The great danger of using a norm is the constriction on our thinking once we identify what would almost satisfy the customer.

Another great danger is making one big leap in logic to the end result. Instead, starting with a norm and working by increments tends to protect us from the colossal blunder. The Wright brothers, for instance, were bicycle builders, and they used many of the norms from bicycle construction to create their success at Kitty Hawk.

A third danger is starting with the wrong norm, which could prevent us from making a great leap forward when one is possible. Orville and Wilber Wright did use a rail to launch their plane, but they didn't become the first heavier-than-air fliers by putting wings on a locomotive (Figure 5-2).

Figure 5-2. Don't let the norm dictate the form. If the Wright brothers had been train builders, they might have specified a plane that looked like this hybrid, which might have been on the right track, but would have had a hard time getting off the ground.

Suppose we agree to use a chair for a norm. Unfortunately, your mental picture of a chair may be very different from mine. A mockup is a way to protect against this ambiguity by providing an actual scale model of a product. Moreover, we can benefit by using the mockup to demonstrate, study, or test the product long before the product is actually built.
A mockup serves as a norm, when no norm exists, or when none is available. As such, it has all the advantages and disadvantages of a norm. It also has the advantage and disadvantage of being a fantasy product. When we use a mockup, we aren't restricted to what exists, but on the other hand, we can easily mock up a product that could never actually be built.

In printing, and in computing, for example, the mockup is often in the form of a layout of printed matter, or material on a screen. The customer and users can point to the layout and say, "Yes, that's what I want," or "No, what's this doing here?" What we are actually testing with the mockup is the customers' emotional responses—their desires. In effect, a mockup says, "This is what we think the product's face will look like. Let's see how you react to this!"

Many ideas for design projects simply begin with a name: Create Superchalk. Build me a table, chair, pencil, clock, elevator, steering wheel, speedometer, or bicycle. Although the name provides a quick and common connection for all participants to grasp, names also come with a large baggage of connotations. As we've seen, each word is worth a thousand pictures, and each connotation of a name may introduce implicit assumptions.

For instance, Jerry spent thirty years searching unsuccessfully for a use for computer paper edges largely because the name itself narrowed his thinking unnecessarily. Our colleague Jim Wessel observed that his four-year-old daughter isn't so limited. She cuts these strips into smaller pieces and calls them "tickets." We would do well to emulate the four-year-olds who have little trouble making up names for objects lacking conventional ones.

 Further Reading

The Exploring Requirements books can be obtained from a variety of retailers. Go to my website and chose your favorite source of books or eBooks.

No comments: