Churney

Churney is a Copenhagen, Denmark-based company that builds predictive customer lifetime value (pLTV) models for performance marketing. Its core argument is that ad platforms optimize toward conversions occurring inside a short attribution window, while a customer's true value often materializes over subsequent months. Churney connects to a customer's data warehouse, trains causal machine learning models on that first-party historical data, and predicts each user's future revenue.

Those predictions are transmitted back to advertising platforms — Meta, Google and TikTok — as real-time, user-level conversion events via server-side integrations such as Meta CAPI and the Google Ads API, so that platform bidding algorithms optimize toward long-term value rather than early proxies such as installs, trials or first purchases. The company states its processing is anonymized and privacy-safe, stripping personally identifiable information while retaining predictive signal. Beyond the conversion-event feed, Churney also describes dashboard outputs from a "decision engine" used by client management and marketing teams.

The company's own materials describe onboarding in roughly 10 to 14 days, followed by an A/B experiment between the Churney pLTV conversion event and the client's business-as-usual conversion event, typically run for 60 to 120 days to measure incremental return on ad spend.

Founding story

Churney was co-founded by Noy Rotbart (Co-founder & CEO), Brian Brost (Co-founder & CPTO), Casper Hansen and Christian Hansen, all of whom hold PhDs in Computer Science from the University of Copenhagen. Rotbart previously was a data science manager at LiveIntent, an architect in the innovation department at Nykredit, and lead data engineer at Save the Children. Brost's doctoral research covered applied machine learning and causal inference, and he was previously a research scientist at Spotify, where he initiated work on using causal machine learning to maximize the impact of promotional campaigns. Casper Hansen's research focused on user modeling and recommender systems, with more than 25 publications and two US patents; Christian Hansen worked on machine learning for sequential data applied to search and recommender systems.

Business model

Churney sells a B2B software/service to advertisers, integrating with the client's data warehouse and ad accounts. It states a performance guarantee: it works with a client only if long-term ROAS improves by a double-digit percentage, and does not charge if that improvement is not delivered. Specific pricing terms are not disclosed on the company site.

Client-paid, with fees contingent on delivering a double-digit percentage improvement in long-term ROAS according to the company's stated guarantee.

Traction

The website publishes client case studies across the USA, UK, Denmark, Israel, Turkey, Ukraine and China in retail, apps and SaaS categories, citing ROAS improvements ranging from +12% to +190%, including Los Angeles Apparel (+29%). The company states advertisers typically see a 10-50% ROAS improvement within 60 to 90 days, with a 10% uplift as its minimum success benchmark.