Recurring Revenue: The Holy Grail for Businesses
Recurring revenue is increasingly viewed as the ultimate goal for businesses. This shift is exemplified by Apple's evolution into a services-oriented company, now deriving 20% of its revenue from offerings like Apple TV+, Apple Music, iCloud, and various other subscription and non-subscription services. The cost of acquiring new customers has soared by over 200% in the last decade, underscoring the critical need for predictable revenue streams. However, subscription-based models come with their own set of challenges; companies must not only focus on attracting new subscribers but also prioritize retaining existing ones. Reducing churn is essential for success.
Danish startup Subsets aims to tackle these challenges through an innovative, AI-driven platform designed to "bridge the retention gap in the subscription industry."
Harnessing Explainable AI
Founded in Copenhagen in 2022, Subsets utilizes "explainable AI" to identify which current subscribers may be at risk of leaving, and recommends tailored "experiments" to incentivize them to stay. Explainable AI is crucial in ensuring that users can comprehend and trust AI-generated insights, addressing concerns about the “black box” nature of many AI systems.
Just six months after its launch, Subsets has attracted noteworthy clients, including the New York Times-owned Athletic and the Danish newspaper Børsen. The company is now focused on scaling its operations, having secured $1.65 million in pre-seed funding from Nordic early-stage VC Upfin and Y Combinator (YC), following its participation in YC’s Summer ’23 program.
Targeting Digital Media
Though Subsets has the potential to benefit any subscription-based business, its current concentration is on the digital media sector. "We are restricting our focus to digital media for now—covering digital publishers, streaming platforms, subscription applications, and telecom services," stated Martin Johnsen, co-founder and CEO of Subsets. "In the future, we plan to expand into other areas such as mobility, banking, and food delivery as digital subscriptions continue to grow."
Companies integrate Subsets with their internal systems—CRM (Customer Relationship Management), CMS (Content Management System), billing platforms, and data warehouses—to gather valuable subscriber data. This encompasses demographics, content consumption patterns (e.g., engagement times and content preferences), and other relevant interactions.
Subsets provides a user-friendly web app that allows commercial teams to analyze "churn audiences" through intuitive visuals and natural language insights. For instance, it may highlight thousands of subscribers who once actively engaged with various titles under a specific media brand but are now indicating a decline in interest.
Empowering Retention Experiments
Subsets empowers non-technical teams to conduct retention "experiments" on targeted subscriber segments, exploring strategies that may enhance customer loyalty. These initiatives could include a series of push notifications or an email campaign offering subscription discounts or complimentary feature upgrades. Clients can adjust these "retention flows" to align with their unique needs.
As teams identify effective strategies for reducing churn during the experimentation process, the results are documented to guide future retention efforts. This system is designed to minimize guesswork, allowing companies to "automate what works."
"Some of our experiments will significantly boost subscriber retention—those with successful outcomes become automated," Johnsen remarked. "Each audience is defined by behaviors that lead to churn, and these groups are dynamic. New subscribers continuously enter and exit the audience, thus receiving effective retention flows."
Innovative AI Algorithms
Johnsen, who brings a background in mathematical modeling and computation, revealed that Subsets has developed proprietary AI algorithms using "gradient-boosting models with temporal sequencing methods." Gradient-boosting merges multiple weaker predictive models into a singular robust model, while temporal sequencing incorporates time-dependent features into the modeling process. Furthermore, Subsets leverages frameworks from Elon Musk’s xAI to clarify "churn-driving behaviors," as Johnsen describes.
Currently, Subsets does not use any models from OpenAI's GPT suite, although they are in the process of integrating additional functionality powered by the same foundational technology that drives ChatGPT.
In a recent social media update, YC President and CEO Garry Tan noted that approximately half of YC portfolio companies utilize LLMs in some capacity, highlighting the strategic alignment with its investment in Subsets. Partnering with YC is proving beneficial for Subsets, not only for growth but also due to YC’s extensive network—especially since their client Athletic is a YC alum, having participated in the Summer ’16 program before being acquired by the New York Times.
"YC provides an incredibly strong network that has opened up exciting opportunities for us," Johnsen emphasized.
In addition to lead investors Upfin and YC, Subsets’ pre-seed funding round also saw participation from numerous institutional and angel investors, including Cuesta Labs, Sandhill Markets, and Peakon founder Phillip Chambers.
Ranking
