More from these Authors

• Case | HBS Case Collection | September 2019 (Revised September 2020)

  SOFWERX: Innovation at U.S. Special Operations Command (Abridged)

  Herman Leonard, Mitch Weiss, Jin Hyun Paik and Kerry Herman

  Citation:

  Leonard, Herman, Mitch Weiss, Jin Hyun Paik, and Kerry Herman. "SOFWERX: Innovation at U.S. Special Operations Command (Abridged)." Harvard Business School Case 820-047, September 2019. (Revised September 2020.)  View Details

  CiteView DetailsEducatorsPurchase

  Related

• Case | HBS Case Collection | September 2020

  PDS: Ring-Fencing the Ranch

  Dennis Campbell, Tarun Khanna and Kerry Herman

  Pallak Seth, Group CEO of PDS Multinational Fashions, is contemplating options to bring better collaboration across his global apparel supply chain platform. PDS, a group of 50-plus subsidiary companies, each led by its own CEO and with different apparel industry specialties, has grown rapidly over the past decade, yet the industry is increasingly competitive and challenges remain constant. Looking for ways to reach the Group’s 2023 goals of $2 billion in revenue, Seth and his executive board are considering two ways to increase collaboration and reward performance: a joint P&L approach to drive partnerships across the subsidiaries and an employee stock option plan, to unlock value across the Group.

  Citation:

  Campbell, Dennis, Tarun Khanna, and Kerry Herman. "PDS: Ring-Fencing the Ranch." Harvard Business School Case 721-361, September 2020.  View Details

  CiteView DetailsEducatorsPurchase

  Related

• Working Paper | HBS Working Paper Series | 2020

  Time and the Value of Data

  Ehsan Valavi, Joel Hestness, Newsha Ardalani and Marco Iansiti

  This paper investigates the effectiveness of time-dependent data in improving the quality of AI-based products and services. Time-dependency means that data loses its relevance to problems over time. This loss causes deterioration in the algorithm's performance and, thereby, a decline in created business value. We model time-dependency as a shift in the probability distribution and derive several counter-intuitive results.
  We, theoretically, prove that even an infinite amount of data collected over time may have limited substance for predicting the future, and an algorithm that is trained on a current dataset of bounded size can attain a similar performance. Moreover, we prove that increasing data volume by including older datasets may put a company in a disadvantageous position.
  Having these results, we answer questions on how data volume creates a competitive advantage. We argue that time-dependency weakens the barrier to entry that data volume creates for a business. So much that competing firms equipped with a limited, but sufficient, amount of current data can attain better performance. This result, together with the fact that older datasets may deteriorate algorithms' performance, casts doubt on the significance of first-mover advantage in AI-based markets.
  We complement our theoretical results with an experiment. In the experiment, we empirically measure the value loss in text data for the next word prediction task. The empirical measurements confirm the significance of time-dependency and value depreciation in AI-based businesses. For example, after seven years, 100MB of text data becomes as useful as 50MB of current data for the next word prediction task.

  Keywords: economics of AI; Machine learning; non-stationarity; perishability; value depreciation;

  Citation:

  Valavi, Ehsan, Joel Hestness, Newsha Ardalani, and Marco Iansiti. "Time and the Value of Data." Harvard Business School Working Paper, No. 21-016, August 2020.  View Details

  CiteView DetailsSSRN Read Now

  Related