Early Identification of Emerging Technologies Using the Number of New Companies

• Sung Min Park,
• Sodam Kim,
• Hee-Woong Kim,

Vol. 19, No. 9, September 30, 2025
10.3837/tiis.2025.09.011, Download Paper (Free):

• Early identification
• Emerging Technology
• forecasting
• New company
• Patent
• Search traffic

Abstract

Recent studies have utilized patent or search trend data to identify emerging technologies, but the inherent limitations of each data source can delay or misidentify emerging technologies. This study aims to show that a new data source, the Startup DB, is useful for early detection of emerging technologies. To do this, we collected 121,938 patents filed with the USPTO over 23 years and 39,714 new companies registered in Crunchbase for 30 technologies cited by Gartner/WEF as emerging technologies since 2000. Then, we evaluated how quickly the S-shape pattern of the cumulative time series data of new companies by year appears compared to the cumulative time series data of patent applications by year through time-lagged cross-correlation analysis. As a result, we find that new company data captures emerging technologies 2.7 and 3.0 years earlier than patent application data in terms of growth and impact characteristics. We also found that new company data is superior to search trends in capturing emerging technologies early without misjudgment. Therefore, we suggest that academia and industry utilize data on the number of new companies, growth, and impact indicators to proactively detect emerging technologies in technology planning, and we look forward to further research based on this.

Statistics

Cite this article