Acast has announced the launch of ‘Smart Recommendations’—an AI search engine that allows advertisers to find their perfect podcast audiences in seconds, simply by describing who they want to reach. Its aim is to supercharge how podcast advertising is bought and sold, making campaign planning faster, smarter and more effective.
Powered by a decade of Acast’s own data and experience, and using OpenAI’s LLM technology combined with cutting-edge retrieval-augmented generation (RAG), ‘Smart Recommendations’ acts like an AI media planner—turning a simple prompt like “I want to reach women in Australia interested in investing” into a curated list of high-fit podcasts. It’s designed to surface the right shows, faster, using natural language and advanced search to deliver unmatched precision in audience targeting.
‘Smart Recommendations’ is the first release from Acast Intelligence, a new core capability from Acast dedicated to using AI to enhance its product suite for both creators and advertisers. By blending human expertise with intelligent data processing, Acast Intelligence unlocks the potential of Acast’s expansive dataset to drive smarter discovery, planning and performance in podcast advertising.
“The podcast landscape is vast, and finding the perfect audience can be time-consuming and challenging,” said Matt MacDonald, chief product officer for Acast.
“Smart Recommendations solves this by harnessing the power of Acast’s decade of proprietary data combined with Podchaser’s insights, and providing a ‘second brain’ for advertisers. This empowers them to discover hidden gems and connect with ideal audiences with unprecedented speed and precision. It’s about moving beyond guesswork to truly understand and reach the right audiences.”
‘Smart Recommendations’ is available to advertisers launching campaigns via Acast’s ad platform, and used by Acast’s internal sales teams around the world.
Key features include:
- Natural language prompts: Advertisers can simply describe their target audience in plain language.
- Advanced semantic search: The system understands the intent behind a query, including context and related concepts. It can extrapolate from a prompt and apply a deeper level of nuance.
- Rich show insights: Recommendations are based on deep analysis of podcast content, audience demographics, engagement, tone, and style.
- Transparent recommendations: Every suggestion is accompanied by detailed explanations of why it was chosen, helping advertisers make informed decisions.
Early testing of ‘Smart Recommendations’ with Acast’s internal sales teams and select customers directly has yielded very promising results. Over 200 campaign briefs have utilised the tool in testing, leading to a significant reduction in planning time—by up to as much as 92 per cent. Notably, 80 per cent of ad buyers in testing discovered additional, previously unconsidered podcasts for their campaigns. Early indicators from shows with under 50,000 weekly listens show a significant uplift, including a 14 per cent increase in median purchase rate and a 35 per cent increase in median visit rate, according to Podscribe data.
“This is more than just a search engine; it’s a powerful tool that transforms data into campaign success, ultimately helping advertisers maximize their podcast ad ROI with AI,” added MacDonald. “Smart Recommendations is the next step in our vision to become the intelligence layer powering podcast advertising globally, offering AI-powered podcast recommendations that truly deliver.”
“The launch of Smart Recommendations is a perfect example of our relentless pursuit of valuable podcast audiences in Australia and the use cases are nearly endless,” said Henrik Isaksson, regional managing director for Acast ANZ.
“Simplifying the buying process for advertisers has long been our mission and we’re uniquely positioned to offer this at scale across our entire content slate.”
