Smarter Algorithms Stop Factory Robots from Colliding
How researchers helped JD.com improve the efficiency of its fulfillment warehouses.
Smarter Algorithms Stop Factory Robots from CollidingPaying for apps such as HotSchedules, Quick Appointments, or Eavesdrop is more useful to you if your friends also buy them. That makes you likely to tell your social media friends about such apps, prompting more people to purchase them.
Companies can capitalize on this network effect to sell their products—be they apps, games, toys, or operating systems—with mathematical precision, according to University of North Carolina’s Nur Sunar, Chicago Booth’s John R. Birge, and Sinit Vitavasiri of Ericsson, the phone maker. It’s even possible for a company to determine whether it shouldlaunch a product and to figure out who the optimal buyers are on the basis of their social media friends, the researchers find.
They looked at what managers call a “promote-and-exploit product-launch” strategy. Here’s how it works: suppose a product is more valuable to you if your friends also have it because it allows you to collaborate with them (think Google Docs) or track their locations (think Find My Friends). Betting on this increase in demand, thecompany can launch the product by selling first to a select group of consumers at a lower price, and raising the price once demand increases.
The study is among the first to consider together all three stages of a product launch— development, launch, and postlaunch—and to determine mathematically a company’s optimal course of action as it moves through these phases, the researchers write.
The Equation: The fastest way to promote your product
Sunar, Birge, and Vitavasiri developed a mathematical model to represent a set of customers, capture how they are connected to one another on social media, and track whether they buy the product when it’s launched. The model was designed to evolve randomly as a business develops its product before launch, much as it would in the real world, where external forces such as bad weather or a competitor’s advertising campaign might play a role in decision-making.
First, a company has to decide whether to launch a product. If it does, it identifies a launchdate, and the model incorporates a range of possible prices. The company promotes the product selectively based on the model’s insights into which customers will be most likely to buy the product and tell their social media network about it.
The model, says Birge, “gives you a process to pick out those customers whom you would really like to target, and also gives you suggestions about how to price a product and what customers to focus on.” For instance, a product developer could home in on the most likely customers and send them a coupon.
Nur Sunar, John R. Birge, and Sinit Vitavasiri, “Optimal Dynamic Product Development and Launch for a Network of Customers,” Working paper, July 2018.
How researchers helped JD.com improve the efficiency of its fulfillment warehouses.
Smarter Algorithms Stop Factory Robots from CollidingChicago Booth’s Ram Shivakumar on making the leap from start-up to superstar.
Strategies for Scaling UpRecent years have seen a decline in GE’s market capitalization and an uptick in uncertainty over its future. How did the once-mighty company find its way into trouble?
Line of Inquiry: James E. Schrager on GE and ‘The End of History’Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.