Experian Information Solutions v. Nationwide Marketing Services
893 F.3d 1176 (2018)
Premium Feature
Subscribe to Lexplug to listen to the Case Podcast.
Rule of Law:
A factual compilation, such as a database of names and addresses, is copyrightable if its selection and arrangement involve a minimal degree of creativity. However, the copyright protection afforded is 'thin,' requiring a plaintiff to prove 'bodily appropriation' or that the infringing work is 'virtually identical' to establish infringement.
Facts:
- Experian Information Solutions, Inc. (Experian) creates and maintains the ConsumerView Database (CVD), containing over 250 million consumer records.
- Experian's process involves selecting data from over 2,200 sources, using thousands of 'business rules' and algorithms to resolve conflicting information, and determining which name-address pairing is most accurate.
- Experian also curates its database by excluding certain records it deems not valuable to its marketing clients, such as business addresses and addresses of individuals in prison or the very elderly.
- Nationwide Marketing Services Inc. (Natimark), a competitor, acquired a database called the National Consumer List (NCL) with approximately 200 million consumer records.
- In 2012, a data broker for Natimark attempted to sell a data sample to Experian.
- Experian tested the sample and found a match rate of over 97% with its own highly curated CVD, leading it to suspect the data was stolen.
- Experian's expert later determined that Natimark's entire NCL database had approximately a 94% match rate with Experian's CVD.
- Natimark had acquired its database for an unusually low price without a customary written agreement containing industry-standard restrictions.
Procedural Posture:
- Experian filed suit against Natimark in the U.S. District Court for the District of Arizona, alleging copyright infringement.
- After the district court ruled that the data pairings were not copyrightable, Experian amended its complaint to add a claim for trade secret misappropriation.
- The district court granted summary judgment in favor of Natimark on both the copyright and trade secret claims.
- Experian, as the appellant, appealed the district court's grant of summary judgment to the U.S. Court of Appeals for the Ninth Circuit.
Premium Content
Subscribe to Lexplug to view the complete brief
You're viewing a preview with Rule of Law, Facts, and Procedural Posture
Issue:
Does a defendant's database infringe the copyright on a protected factual compilation when it contains a substantial majority, but not the entirety, of the protected work's data?
Opinions:
Majority - Judge Schroeder
No. Copying a substantial majority of a factual compilation does not constitute copyright infringement because the 'thin' protection for such works requires a showing of bodily appropriation or virtual identity, which is not met when the defendant's work is materially smaller than the plaintiff's. The court first held that Experian's database was copyrightable as a compilation. Unlike the telephone directory in Feist, Experian's process involved a minimal degree of creativity through its selection of sources, use of algorithms to resolve data conflicts, and exclusion of certain categories of data to enhance the database's value. However, the court affirmed summary judgment for Natimark on the infringement claim. It reasoned that copyright protection for factual compilations is 'thin' and infringement requires 'bodily appropriation' or that the works be 'virtually identical.' Since Natimark's database (200 million records) was significantly smaller than Experian's (250 million records), the maximum possible overlap was 80%. The court held that an 80% match rate is insufficient to establish bodily appropriation of Experian's work. The court reversed the summary judgment on the state law trade secret claim, finding triable issues of fact as to whether Experian's database qualified as a trade secret and whether Natimark knew or should have known it was misappropriated.
Analysis:
This decision solidifies the two-step framework for analyzing copyright claims for factual compilations. It confirms that the bar for copyrightability is low, allowing sophisticated data curation methods to qualify as sufficiently creative. However, it establishes a very high bar for infringement, reinforcing the 'bodily appropriation' or 'virtual identity' standard for these 'thinly' protected works. This ruling makes it very difficult for database compilers to succeed on copyright infringement claims unless a competitor copies their work nearly in its entirety. Consequently, the case highlights the importance of trade secret law as an alternative, and perhaps more effective, means of protecting valuable compiled data.
