‘Game’ theory: Perfecting in-app purchasing through analytics

When you think about buying and selling, it’s usually about a tangible product or service. But what if the job is to sell new forms of reality and value itself? If you’re an online gaming company, that’s exactly your goal: to create and sell virtual worlds of escape and adventure, with the “value” increasingly coming through analytics-fueled “in-app” sales of virtual accoutrements — like weapons upgrades or special outfits for player avatars.

Analytics to navigate change and competition

It turns out the rise of in-game, or in-app, purchasing is fueling a massive business; online gaming companies generate $109 billion in revenue annually in a low-overhead, high-profit industry. What’s more, some companies are now looking to analytics to further optimize the in-app marketplace — specifically, they’re positioning analytics to automate many of the routine decisions about when and where to put in-app purchase offerings, and at what price.

This is worth exploring as a powerful example how big data and automated decisioning are helping an entire industry adapt to a changing business model. The past two decades have seen a massive shift in the gaming industry from subscription-based pricing to in-app purchases. World of Warcraft, for instance, became the industry’s first billion-dollar game back when it was enjoying wild success from subscriptions a decade or so ago.

As consumer preferences and demands have changed, WoW has since adopted a hybrid model of subscription and in-app purchases to generate income from its millions of players worldwide. Other games, like Hearthstone, now rely primarily on in-app purchasing. The challenge is: an enterprise level in-app marketplace requires countless low-level decisions that are just crying out for automation.

If you have only human analysts busy with these routine, tactical decisions about what to sell and where, it becomes to keep up when you are operating with millions of players. The right analytics architecture, on the other hand, can parse and curate high volumes of behavioral game play data to help decide the best product, placement and price for an in-game purchase. This frees your people up from minute tactical decisions — so they can focus on bigger and more strategic priorities.

Staying true to a mission

I have a contemporary who is the chief technology officer of a major online gaming business. When I visited him recently on the company’s campus, larger-than-life statues of game characters loomed on the grounds outside the main building. It reminded me how seriously he and his colleagues take their work. To developers and players alike, these characters and worlds are real — at least in terms of the time, effort and money that’s invested.

For these worlds to stay credible and compelling, they need to be fair; we must make sure in-app purchases generate revenue without tipping the balance of power. Analytics can help sift through rich behavioral data to see who’s winning, who’s losing and where an in-app purchase might help —- but in ways that don’t tip the balance of power so fully that people can buy their way to what he referred to as “guaranteed wins.”

“All these issues of fairness, marketing and functionality need to happen with high concurrency at a scale of more than 220 terabytes of game-play data per day for a typical game,” he explained. “So, the more we can automate that flood of tactical decisions, the better those decisions will be and the more we can free up our people to think more strategically, make more games and come out with new iterations and better features for existing games.”

Broader implications

Given how in-app purchasing is an increasingly common tool in many industries, especially retail — the benefits I’m describing can be applied more broadly than just the gaming industry. More and more companies are realizing analytics is the only way to manage the amount of data, processing power and automated decision-making needed to position and price in-app products at scale.

And there’s a bigger lesson as well: The way analytics can automate the in-app marketplace is just one example of the ways analytics can be used in all sorts of fields to take over routine decisions. In many kinds of industries — not just gaming, but also manufacturing, marketing, telecommunications, retail and any number of other sectors — autonomous decisioning that’s driven by sophisticated algorithms can free up people’s time for important strategic decisions versus countless tactical ones.

Adapting to changing circumstances and autonomous decisioning, it turns out, are two critical elements of an analytics approach I co-developed with the Northwestern Kellogg School of Management’s Mohan Sawhney. I’ve written a lot about our “Sentient Enterprise” analytics capability maturity model, and there’s even a full-length book that’s coming out soon. It’s just one way of looking at the challenges of big data, but whatever your approach, it’s hopefully designed to embrace agility and position data to help with, and even make, decisions at the speed of business.