Millions of patents were filed across the world in 2020, according to recent publications from global patent offices. Patents have high predictive power and are used by data scientists and investment firms to carry out key functions like portfolio analysis , research and development, and competitive intelligence. But the sheer number of filings last year and in the first quarter of 2021 shows how difficult it is for these organizations to keep on top of patent developments to generate these insights.
Although developing products and inventions in a laboratory was made more difficult by the pandemic, the pace of patent applications and grants in the last year has been staggering. For example:
Over 680,000 patent applications were filed in China in the first half of 2020 alone.
Patent data is highly sought after by organizations of all stripes because of its high predictive power. This is because the average time from patent to product is between 2 and 5 years. Patent data offers invaluable insights for companies seeking to:
Make investment decisions
Predict business and product trends
Understand the strategies and future products of competitors
Carry out research and development
Alternative data and AI technologies have allowed data scientists and companies to find meaning from patent data in new and transformative ways. A recent article in the Journal of Big Data analyzed more than 13,000 patent applications over nearly three decades to understand the way in which big data analytics of patents has developed and will continue to increase in the coming years. The researchers identified many benefits of applying technology to patent data, calling it “a valuable resource for understanding the dynamics and activities of the invention ecosystem”.
Technology unpicks the meaning of patent data
Patent data clearly presents significant opportunities for companies, but the high volumes of filings last year also presents a challenge: How do you cut through the noise of millions of patent applications to find patterns, trends and predictions? The answer lies in acquiring the best possible data and using technology to reveal meaningful insights.
Companies in the market for patent data should consider the following:
Does the data give you global coverage? There are more than 100 patent authorities across the world at both national and international level so patent data is spread across a multitude of locations.
Does the data let you track trends over time? Being able to access patent data over a period of a decade or more allows easier identification of trends in innovation in different sectors. Some providers offer point-in-time patent data, allowing you to follow how a product or idea was developed over months and years.
What is the source of the data? The most authoritative source is patent offices in different jurisdictions, so patent data should preferably be gathered from them directly. Going to 100 authorities is very time-consuming, whereas data providers can offer global data gathered in one place.
Is the data regularly updated? Patent datasets should be as current as possible, or the user will miss the very latest trends.
Is the data in a format that reduces time-consuming data wrangling? Look for enriched patent data that can be usefully compared to and combined with other sources, such as news coverage, company data and legal data, and optimized for data analytics technologies.