Media Briefing: The 'SaaS-pocalypse' is spreading to publishers

Media companies are increasingly leveraging artificial intelligence to construct custom software applications, challenging the traditional reliance on external Software-as-a-Service (SaaS) subscriptions. Known as "vibe coding," this process allows engineering teams to deploy products through natural language prompts, reducing the need to pay for specialized tools ranging from analytics dashboards to customer relationship management systems. This trend has triggered a "build versus buy" debate within publishing houses, prompting some executives to negotiate shorter contract terms with vendors. One anonymous executive noted pushing for one-year agreements instead of the standard three-year renewals, anticipating that internal builds or cheaper alternatives may emerge before the term expires. Natalie Drucker, director of AI and data strategy at Thoughtworks, confirmed seeing over 50 percent price reductions offered by vendors in response to this threat of in-house competition. Despite the appeal of cost savings, industry leaders warn against underestimating the complexities of maintenance and data compliance. Anonymous publishing executives highlighted concerns regarding the reliability of AI-generated code, citing issues with hallucinations and the difficulty of scaling internal tools to serve thousands of users. Ben Murray, a fractional SaaS CFO consultant, emphasized that vendors provide essential liability guarantees and customer support functions that internal teams may struggle to replicate. Some organizations are finding middle ground by using AI for rapid prototyping rather than full replacement. Business Insider reported utilizing the technology to build interactive quizzes and test concepts quickly, enabling editorial teams to launch products in hours rather than months. Meanwhile, other publishers are deploying AI agents for sales and marketing workflows, though critical systems like customer data platforms remain largely dependent on established vendors due to data consent and monetization uncertainties.