Fireside chat: Digital IQ and marketing big data
summary
Social listening surfaces that 2 million posts around Hurricane Sandy reveal not what is said but where silence signals outages and need, and volume often outperforms sentiment for predicting market moves. Practical tools include image recognition finds organic brand users for micro‑influencer outreach, personalization at scale from recommendation models, chatbots for simple transactions, and recurrent neural networks with attention to attribute multi‑channel funnels; generative models create paraphrases and synthetic training scenes, while dark social, privacy controls and biased training data remain unresolved risks.
AI-written summary · about
chapters (13)
- 0:00Intro
- 1:25Questions
- 2:05Volume of conversation
- 5:32How social media has changed
- 10:26Behavioral segmentation
- 10:50Chatbots
- 13:56Different data
- 15:35Multichannel attribution
- 19:31Can you go viral
- 23:03Generative models
- 26:46Text sentiment analysis
- 28:15Data privacy ethics
- 31:09Final thoughts
related talks
43mConsumer Intelligence and the Age of Predictive Marketing Analytics
23mBig data and big insights
04mHarriet Fryman interviewed at Web 2.0 Summit 2011
28mBig Data & AI in Marketing
28mML & GenAI for Vibe Marketing: Future of Audience Connection
10mFireside Chat: AI-Powered Brand Safety
36mML: Predictive analytics in Ad Engineering
25mDeep learning for disaster management and response