Data storytelling in the era of generative AI
summary
Claims a working service that uses Large Language Models over clustered consumer surveys to create evidence-informed synthetic personas that answer marketers in natural language and export targeting to Meta Google and TikTok. Pipeline uses zero‑party surveys, segmentation, and model checks with cosine similarity 0.85 and an F1 0.94 threshold; an MVP ships in six-month MVP and runs beta from the marketing site with token limits and a QABS staging/production sandbox. Early tests show faster profiling, higher campaign ROI and occasional audience‑size caps when criteria overfilter.
AI-written summary · about
chapters (15)
- 0:00Data storytelling with AI
- 1:06Kudo’s mission focus
- 2:56Kudo’s customer data
- 4:43New product opportunities
- 7:31Marketer’s customer journey
- 9:39Shift to conversational storytelling
- 10:43Chat to Persona for marketers
- 12:30Balancing customer and business value
- 15:01Chat to Persona’s business potential
- 16:49From POC to MVP
- 19:16Testing chat to Persona
- 21:45Future plans for chat to Persona
- 22:07Q&A: Zero-party data
- 24:51Q&A: Chatbot and dashboards
- 26:59Closing remarks
related talks
33mThe Next Frontier of Audience Intel
41mMarketing Data Science
22mFrom the Trenches: Automated Testing of Large Language Models
28mCrafting Hyper-Relevant User Experiences
07mBackstage interview
36mArchitecting Data: A Deep Dive Into the world of Synthetic Data
15mThe Subjective World Model
29mLies, damned lies and large language models