Monday, October 22, 2007

Review: Shaffer’s How Computer Games Help Children Learn

Part critique, part proposal, David Williamson Shaffer’s How Computer Games Help Children Learn is about what kids need to learn and how they need to learn it.

Shaffer asserts that the common practice of school is already a kind of game, with rules rooted in the industrial revolution:

School is a game about thinking like a factory worker. It is a game with an epistemology of right and wrong answers in which Students are supposed to follow instructions, whether they make sense in the moment or not. Truth is whatever the teacher says is the right answer, and actions are justified based on appeal to authority. School is a game in which what it means to know something is to be able to answer specific kinds of questions on specific kinds of tests.

Back when a high-school degree led to a good manufacturing job, this version of school may have made sense. However, U.S. manufacturing jobs are somewhere between going and gone. That leaves the low-end service sector as the primary employer of the high-school-educated workforce. Thus, yesterday’s middle-class auto worker is today’s barely-getting-by burger flipper.

In Shaffer’s view, the only good jobs will increasingly be those beyond the traditional high-school education: jobs that address problems with many possible answers, that require creativity, and that reward innovation. Everything else will be automated, offshored, or marginalized to the low-wage economy.

So if society wants to leave no child behind, the education system needs to change its game. The key shift is from an emphasis on teaching facts to teaching skills. Put another way, Shaffer wants more learning by doing. He wants students to learn by making them participants in simulations of real-world challenges that engineers, urban planners, journalists, and other professionals face. And this is where computers come in.

Think about SimCity, the computer game where you manage a simulated city as it grows. It’s a simulation. It’s a computer game. It’s a learning experience:

[Players] see what happens when they make changes in urban ecosystems. For example, if you put more parks in a city, the cost of public utilities goes up because you have to keep the parks clean. If you put an industrial site next to residential housing, the residential land values fall and the crime rates rise. As a result, players must decide whether to raise taxes, decrease the green space, move the industry, or risk urban flight—or, more realistically, decide which combination of these choices and in what measure will lead to the best long-term outcomes for the city.

Immersing kids in such a world lets them learn key concepts and ways of thinking as they progress through the game. Succeeding in the game requires learning and understanding. Contrast that with reading a bunch of articles about urban planning and being tested on the facts. Will those facts ever mean anything to the students? How much more would a student want to read about urban planning after getting hooked on SimCity?

But Shaffer’s learning games (developed by him and his colleagues at the University of Wisconsin-Madison) are different than SimCity. Whereas SimCity is an imaginary place, Shaffer’s games take care to simulate real places, things, processes, and constraints from the real world. The computer simulation provides an extension of reality rather than a replacement.

For example, an urban-planning game he highlights...

...is modeled on the real world of the city players live in and the real work of planners who shape that city. Players are redesigning a city, but it is their city. They can see and touch the places they are redesigning and can see how those changes might make their lives and the lives of those around them richer and more satisfying. However, their choices are constrained by the economic, social, and physical realities of life in a city and by the norms and practices of the profession of urban planning.

Given this world, students learn key skills necessary to do urban planning in the real world. These include ways of thinking and talking about problems that apply well beyond urban planning—which is the larger point: The underlying creativity, critical thinking, and innovation are what these students will need to compete in the global marketplace, whether they become urban planners or anything else that pays more than subsistence wages.

Of course, this all sounds reasonable, but the games need to be good. SimCity was successful because of fantastic game design and execution, not because millions of consumers inherently craved a city simulator. So can a bunch of academics make something that both satisfies Shaffer’s educational vision and is compelling enough to keep kids at the screen?

Shaffer and colleagues’ games are at early stages and are themselves academic projects. Initial pilot results are promising, but the results are from tests with close participation from mentors and facilitators. While the human element is part of Shaffer’s design, how much mentoring and facilitating is necessary for success? Is that amount practical for large-scale use?

Finally, how do Shaffer’s games get integrated into schools? Here Shaffer has a surprising answer: Maybe they don’t, at least initially. Several of the pilot tests have been in after-school programs. He speculates that they could also be embedded in larger virtual worlds such as Second Life. Although some might see this as a cop-out, avoiding the main institution that Shaffer critiques, I see it as a practical path that brings change where it’s easiest to make change. If the change is good, it will spread.

As is probably apparent, I like Shaffer’s ideas. But if you want a good read along with good ideas, this book might not meet your standards. Although the prose is relatively direct, it carries two burdens. First, Shaffer often tries to use and/or explain his field’s jargon, which I’ve spared you because I found it a distraction as a general reader. Second, Shaffer spends many pages describing computer games using words and the occasional picture, yet the thing that makes the games compelling is their visual and interactive appeal. If the book’s content could be presented as a short documentary film or even a screencast, it would be far more likely to hold a general audience’s attention.

(This suggests an accompanying book: How Videos of Children Learning from Computer Games Help Adults Learn the Value of Children Learning from Computer Games. Or should it be a video?)

Bottom line: How Computer Games Help Children Learn gave me a fresh angle on today’s educational challenges while detailing the first steps of a promising way forward. For people already interested in the issues, the book will be well worthwhile. For anyone else that wants to put a toe in the water, check Shaffer’s Epistemic Games Web site, and if that gets you going, then get the book.

Saturday, October 13, 2007

Customer Service Everywhere

“In most corporate cultures, customer service is regarded as an afterthought and a cost center,” said Craig Newmark, explaining why great customer service is the exception, not the norm. Craig is the founder of craigslist, the world’s most popular forum for classified advertising. It handles 8 billion page requests per month across 450 craigslist sites, each covering a specific geographic region in all 50 U.S. states and more than 50 countries (more info at craigslist’s fact sheet). By his choice, Craig’s full-time job is customer service representative, addressing craigslist users’ questions, complaints, and problems.

Although almost any executive would agree with Craig’s advocacy for better customer service, I see few following Craig’s method of doing something about it, even part-time. To be clear, I am not counting executives who “talk to customers” by talking to executives at their biggest accounts. While that’s good and appropriate, it’s rarely the same as talking to the people who actually use your product.

An executive should talk to users directly because whether the executive’s product delivers value is determined at the point of use, not at a power lunch with another exec. It sounds obvious, yet so much of corporate and product strategy is based on assumptions about users by people who do not regularly engage with said users. In theory, “customer intelligence” percolates up the corporate hierarchy and/or is collected independently from customer surveys. In practice there’s no substitute for executives having ongoing, direct interactions with a representative sample of actual customers.

But let’s not restrict this to executives.

Engineers should talk to users directly because it makes problems real. For example, if the people in your organization who regularly talk to users can’t convince the engineer responsible that a problem matters, have the engineer talk to a couple users who have been burned by the problem. Your organization has an issue if you constantly need to invoke this, but on an occasional basis it can be just the right medicine.

Sales people should talk to users—after the sale has been made—to really understand how the product is used. That way, the salesperson can tell a story to prospects that is not just compelling but also realistic. “Realistic” is important because if expectations are set right, a whole class of customer-service problems disappears: those where Sales promised X, but after the product is bought and installed, Customer Service must now answer for the fact the product actually does Y.

Maybe it’s too easy for me to take this position, because most of my career has involved a founder-level role in start-ups and business units within larger companies. In the early days of any effort, founders tend to work directly with customers—the execs, the users, and anyone else that might matter—because no one is more qualified to do so (and/or because no one else is available to do so ;). Along the way, it’s natural to use this customer proximity to learn first-hand how to improve the product and the human processes that support it. These elements all fit together because a founder is often responsible for it all.

For bigger companies with established products, it’s different. There are whole organizations for customer service versus product development versus market/customer research. At one of my start-ups, Personify, we were well down that path when we reached the 100-employee mark. Although I still spent a significant amount of time directly with customers—and often with users specifically—it tended to be with the more challenging ones: the biggest/highest-stakes customers, the most creative customers that were exploring the product’s boundaries, and the most messed-up customers that needed turning around.

This sample was not representative. In retrospect, I think it biased my later product-design decisions in a way that favored the edges of the distribution, not the great middle.

Looking back, I had the advantage of having started from a position of being close to all the early customers. For an executive hired into an already established company, I suspect that the institutional barriers would be even higher to getting customer visibility that is first-hand and representative—short of taking Craig’s path and living on the support desk.

But that’s not a reason to avoid the issue. It’s just a warning to be deliberate rather than having your customer contact occur as a byproduct of something else you’re doing—at least if you’re planning on using what you learn to inform other decisions.

Finally, having so far highlighted the benefits of people outside the customer-service org doing customer service, it’s worth noting the reverse. People in Customer Service are great exports to other parts of the company, assuming you haven’t offshored all of Customer Service and thus permanently siloed those people. At Personify, people who started in Customer Service and then went onto other organizations (such as Business Development, Product Marketing, and Presales) were often better performers than their peers. I believe it was because they had a tangible sense for what the company and its product actually did—and did not do—from the customer’s perspective. We tried to explain it via presentations, documentation, training, and tag-alongs in meetings—all of which were no substitute for actually being there and doing it.

The moral of the story: When it comes to serving customers and particularly users, be there and do it. You don’t have to be Craig Newmark, but be more than the person who only knows customers as concepts.

[Update, 3/30/2009: I randomly ran across this BusinessWeek article about Amazon.com, which says: “To make sure that everyone at Amazon understands how customer service works, each employee, even [CEO Jeff] Bezos, spends two days on the service desk every two years. ‘It’s both fun and useful,’ says Bezos. ‘One call I took many years ago was from a customer who had bought 11 things from 11 sellers—and typed in the wrong shipping address.’”]