Wizards of the Coast is a great software developer, and you’ve probably never thought of them that way because they don’t make software. They make a card game - Magic: the Gathering was a 1990s phenomenon and is one of the only card games of its time to survive. At this point, it’s not competing with other card games: it’s really competing with World of Warcraft and tabletop roleplaying games. They are successful in a way that software developers can learn from: they gather huge amounts of testing data, they release regularly, and they make sure that their developers are in frequent contact with their users.
Skip this paragraph if you already know how Magic works. Magic is a collectible card game. Wizards of the Coast prints new Magic cards four times a year. Players buy cards in cheap randomized bunches or as expensive singles. From their cards, players construct decks and play against one another. They play casually and in tournaments. Tournaments have prizes, but you can only participate in tournaments with a deck of the right format - tournaments may require decks from the last two-or-so years of cards printed, the last seven-or-so years of cards printed, or a very large subset of all cards ever printed. Wizards of the Coast runs a series of high-profile high-prize events, and entering these events requires registering the list of cards that you used to build your deck.
Wizards of the Coast has been making Magic for seventeen years, and they have gathered more and more testing data as time goes on. They gather a ridiculous amount of data about how their game is used through their Organized Play initiative, representing the usage patterns of millions of worldwide users. All of the most intensive users of their product are sending big streams of feedback data - and not just opinion, they’re sending behavioral data, speaking through action. Every Magic user has a goal in mind when they use the “software,” and Wizards has set up a huge dragnet to gather data about how users tried to get to that goal and how successful they were. This isn’t all the data, either. As any given version of Magic ships, internally, the company is testing the next two versions. This gives them a great way of testing the accuracy of their internal testing: they have a lot of external usage data to compare to the predictions that internal testing made when they were developing a given version. If you were developing a web app, this is the kind of performance and usage data you’d really want: a huge corpus of data about what people wanted, what they tried doing in order to get it, and whether or not they got it.
The data-gathering that Magic’s creators do synergizes with their release schedule: they release a new version four times a year without fail. Their business model requires this - it’s a model that should sound familiar. They have an entertainment product that doesn’t depreciate in value when it gets into consumers' hands, that people form emotional attachments to, and that can be easily transferred from one consumer to another. Wizards of the Coast has demonstrated a lesson that the RIAA, MPAA, and publishing houses could learn from. Instead of evilly and stupidly punishing their consumers for liking their product, they are constantly adding value. They have gathered into their company the best damn card-game designers around: because of the economics of collectible card games, in the long run a competent CCG designer is overwhelmingly likely to be either employed by Wizards of the Coast or not employed as a CCG designer at all. They have hired smart, they realize the limitations of their core product, and they are constantly innovating. They sell accessories of all kinds. They sell ancillary goods of all kinds. They translate their product into other media with abandon, seeing what works and what doesn’t. They do not, it is important to note, sue their customers. Magic isn’t a perfect parallel to music, movies, or books, but I believe it demonstrates that it’s possible to make money without also making your customers hate you. If your customers hate you, they will flee as soon as an appealing alternative presents itself - a phenomenon we’ve seen numerous times in the last decade and change. By behaving evilly, the previous generations of media companies are ensuring that newer media companies will out-compete them. Wizards, instead, is choosing to actually remain competitive.
Of course, you have an elevated risk of being distanced from your customers and alienated from them if you never talk to them. Wizards has avoided that very deftly. They publish lots of material for their community, and the designers of the game solicit feedback directly. If you write to Mark Rosewater, the man who currently has more power over the direction of Magic than any other human, you will get a reply. If you go to a major Magic tournament, you will find delegates from Wizards' development team. Sometimes they will be observing, but quite often they’ll be “gunslinging,” playing the game against anyone who wants to take them on. Magic’s creators eat their own dogfood and are highly exposed to the public. They are not losing touch with their customers or regarding them as just numbers. They are exposed, directly, to issues with their product - and they respond to those issues.
Wizards of the Coast gathers enormous amounts of data about their users' experience with their product, they are constantly adding value to their product, and they don’t hide from their users. When you look at software companies that are doing well, and especially at the ones that have become universally recognized names, I believe that you’ll find them doing the same things. These strategies work everywhere: there’s no excuse for screwing them up.