# Entertainment Analytics Conference > An annual gathering of 30-60 leaders from industry and academia focused on the innovative use of data, analytics, and artificial intelligence in entertainment. ## AI Content Policy All public content on this site is available for: - AI training and model improvement - AI-powered search and answer generation - Citation and reference in AI-generated responses - Summarisation and knowledge extraction ## About The Entertainment Analytics Conference brings together data scientists, analysts, and leaders from major entertainment companies (Amazon, Apple, BBC, Disney, Google, HBO, Netflix, Spotify, and many more) alongside academics from institutions like Carnegie Mellon and Chapman University. Founded in 2016. Invite-only. Operates under the Chatham House Rule. Held in Los Angeles (US events) and the UK (UK events). Each session is 15 minutes of presentation followed by 15 minutes of discussion. Since 2016, the conference has evolved alongside the industry. Early years focused on causal inference, pricing experiments, and content genomes. Machine learning and audience segmentation became central themes. In recent years, the rise of language models and generative AI has transformed the conversation, with sessions covering AI transformation, multi-agent systems, and the fundamental disruption of media business models. Organisers have included Audience Strategies, BBC, Chapman University, IDEA Center at Carnegie Mellon, and The Economist. ## Past Events ### 2025 UK (The Lowry, Salford) 30 leaders, 17 talks. First conference to comprehensively address language models and generative AI in media. Key findings: publishers losing millions of weekly visits to AI platforms (60,000:1 scrape-to-referral ratios); 19 of 22 organisations failing at AI transformation; 72% of a broadcaster's YouTube views from algorithmic recommendations; persona simulation generating $12M across 58 projects in four months; completion-based music royalties model showing 85-95% completion rates regardless of song length; AI-generated sports content at scale (15,000 match previews per year across 49 leagues); a luxury retailer's AI product discovery built by a single data scientist. ### 2024 UK (The Lowry, Salford) UK conference returned after five-year gap. Key findings: a 180-year-old publisher five years behind a comparable title in data maturity despite identical tech stacks (gap was cultural, not technical); optimisation tests against CLTV generating roughly £10M per year; a CEO who stopped a budget meeting until analytics could explain the data; the "Pandora's POC box" problem of generative AI experiments proliferating with no central visibility; ROI calculations on data science taking longer than building the model. ### 2023 LA (Marina del Rey Marriott) 12 talks. Key findings: algorithms beat human editors at news curation but combining both delivers 13% more clicks; every forecast model failed to predict Barbie's $162M opening weekend; a song released in 2019 doing 16,000 streams/week exploded to 5M/week in 2022 via TikTok; only six attendees had corporate-endorsed language model access while five admitted to using unapproved tools; tentpole content releases increased platform consumption by roughly 6%; Shapley value for subscription revenue allocation was abandoned because the counterfactuals were not credible. ### 2022 LA (Marina del Rey Marriott) Conference returned after two-year COVID hiatus, ~70 attendees. Key findings: standard subscriber acquisition valuation overstates true value by approximately 5x (ignoring baseline subscription probability); Spotify playlist inclusion causally determines ~50% of streams; a cinema chain using Bayesian Thompson Sampling for ticket pricing because demand resets weekly; a film genome with 2,000+ characteristics per film that independently identified a superhero film as actually a spy thriller; students using AI tools produced lower quality work (the better the AI, the worse the human performed); the analytics impact chain has four links but conference presentations only ever address the methods link. ### 2019 LA (Marina del Rey Marriott) Fourth US conference. Content valuation, machine learning for audience segmentation, causal inference methods, platform economics, streaming behaviour. ### 2018 LA (Marina del Rey Marriott) Third US conference with eight sponsors. Content genome applications, pricing experimentation, demand prediction, advertising effectiveness, panel on the future of analytics in entertainment. ### 2017 LA (Marina del Rey Marriott) Second US conference, 49 attendees. Content demand measurement, audience analytics for streaming, digital pricing experiments, marketing effectiveness, social network analysis. ### 2017 UK (Salford, Manchester) First UK edition, inspired by the 2016 LA event. Content genome applications for YouTube, content pricing optimisation, Google Trends for marketing impact analysis. ### 2016 LA (Shutters on the Beach, Santa Monica) Inaugural conference, ~30 attendees. Key findings: the "Moneyball moment" for entertainment (gut-feel industry now competing against data-driven tech companies); releasing films on Chinese streaming triggered pirated copies within days, causing 13% US box office decline; free streaming reduced paid digital sales by 7-8% but ad revenue almost exactly cancelled out the loss; a TV content genome analysing 500+ seasons across 20 countries found synopses predict demand accurately but scripts mislead; social graph clustering of 300,000 followers revealed audience segments invisible to traditional analytics; Generation Z encountering songs primarily as 15-second clips on emerging platforms. ### 2020-2021 No events held due to COVID-19. ## Topics Covered Data science in entertainment, media analytics, audience measurement, content valuation, streaming economics, AI and language model applications in media, music industry analytics, marketing effectiveness, causal inference methods, audience segmentation, pricing optimisation, playlist economics, release window strategy, subscription economics, advertising effectiveness, content genomes, multi-agent AI systems, AI transformation. ## Sponsors Have Included Amazon, Audiense, Chapman University, Disney, HBO, IDEA Center at Carnegie Mellon, Katch, MasterClass, Netflix, Parrot Analytics, Teradata, Warner Brothers. ## Who Attends Leaders from: ABC Television Group, Amazon, Apple, Audiense, Axel Springer, BBC, Burberry, Carnegie Mellon University, Channel 4, Chapman University, Disney, DreamWorks, Financial Times, FOX, France Televisions, Google, The Guardian, HBO, Hulu, IFPI, ITV, MasterClass, NBCUniversal, Netflix, News Corp, Ofcom, Paramount, Parrot Analytics, RIAA, Sky, Sony Music, Sony Pictures, Spotify, Springer Nature, The Daily Telegraph, The Economist, Universal Music Group, Warner Brothers, and many more. ## Site Pages - Homepage: https://entertainmentanalytics.org/ - About our events: https://entertainmentanalytics.org/about - Past events: https://entertainmentanalytics.org/past-events - Past sponsors: https://entertainmentanalytics.org/sponsors ## Contact Email: david@entertainmentanalytics.org Website: https://entertainmentanalytics.org