Last year we celebrated our birthday by sharing data on how much content we’ve processed and identified over the years. We’re proud of our scale and how much we’ve grown, so we’re back to share updated stats on our fingerprinting and matching. At Pex, we’ve been identifying content to help rightsholders and platforms for 10 years now, and our 2023 data shows we’re still getting bigger and better. Check out our recap below and reach out to learn more about our technology or products.
We made nearly 2 billion more matches for our customers in 2023
What does a match mean? With Pex Discovery, our customers can track the use of an audio or video file across 20+ platforms, including YouTube, TikTok, Instagram, and Facebook. When Pex identifies a use of a tracked asset, that’s a match! These matches can be opportunities to claim ad revenue, to pursue a license, or to just learn how content is being reuploaded and shared online. In total, we’ve made 12,942,898,680 matches for our customers, roughly 1.9 billion more than we had made at the end of 2022. How do we make these matches? It starts with digital fingerprinting.
Our matches were made using 27,772,761,427 fingerprints
Digital fingerprinting is a method of compressing content so it can be processed and identified quickly, even in large quantities, like when dealing with the millions of uploads made to social media platforms. We continuously index public content from these platforms and turn it into fingerprints, so our customers can see where their tracked assets are used. In 2023, our database grew by 4 billion to a total of more than 27 billion fingerprints.
At the platform level, TikTok took up a larger percent of our total fingerprints last year, increasing from 2.7% of all fingerprinted content in 2022 to 8.7% in 2023, as the platform has become an increasingly important place for music rightsholders to track their content.
It takes 393 TB of storage to house our fingerprints
Even after compressing content, it still takes a tremendous amount of storage to contain 27 billion fingerprints. Since we increased our fingerprints by 4 billion, we also increased our storage needs by 56 TB from 337 TB in 2022. This amount of storage is nothing to scoff at, but it’s trivial compared to what we’d need if we were dealing with uncompressed files.
Combining fingerprints with our state-of-the-art matching technology doesn’t just save us space, it also allows us to identify content in real time. We can even identify content during the upload process to prevent unlicensed or harmful material from going online. We believe as digital content continues to evolve, processing and identifying it will only become more important. So we’re always working to improve our systems to be faster without compromising accuracy. Fingerprinting, along with our architecture, helps us accomplish this.
What’s in store for 2024
As we’ve continued to innovate over the years, we’ve increased our fingerprint types from two to five: audio, video, melody, phonetic, and now voice. These fingerprint types all work differently, but together, to find the most matches of content, especially recordings and compositions. We’ll share more data throughout 2024 on how we deploy these fingerprint types to match more content, including AI-generated voice clones.
Want to track your content with us this year? Let us know.
