This blogpost is a brief summary of the points mentioned in the FSI day at Google

Customer Data Platform with Google Cloud

Expectations have risen dramatically
Fusion of structured adn instructed data into 360 view
- amount of data
- type of data
- usability of data
Heightened data privacy and regulation
Hub and Spoke to consolidated data platform
Challenges
- Slow
  - Data silos
  - Legacy processes and technology
  - Fragmented data models
  - Complex data pipelines and batch ETLs
  - Model risk governance
- Expensive
- Inefficient
CDP
Use cases
- Events preceding account closure
- Generating sales leads from general browsing
  - Digital clickstream journey
- Propensity to buy a product
- Create digital variables and create libraries that can be used for future models
- Build behavioral data to propensity models
- CROFT - Compliance - Risk - Operations - Finance - Treasury
Traditional stats vs Behavioral Stats
- Minimal overlap
What is 3P data ?

Search, Automation and Hyper-personalization

only 17% of marketers are able to use data and AI effectively
Broadcast era - Precision Era - Predictive Era
Data
- Relevance
- Consistency and quality
- Uniqueness
Challenges
- Scaling
- Speed
- Standards
- Siloizatin
- Sharing
- Security
- Servicing
- Skills
Kakao bank in Korea - Digital bank
People are willing to share everything on a social platform but are anal about sharing data with a bank
Omni-channel activation
Re-targeting
Indonesia - many geographical regions
- Difficult for a bank in each geographical region
- Banks have managed to integrate in the mom and pop store in each of the island
Use cases
- Enhanced Customer Understanding
- Smart Consumer segmentation
- Presonalized Makreting communication
- Attribution Modeling
- Marketing spend optimization
- Consumer signals for demand sensing
- Financial product recommendation
- Consumer lifetime value optimizatoin
- Propensity to convert modeling
- Trend spotting
- Sentiment analysis
- Accelerating qualification and innovation
What is the difference between GA360 vs GA4 ?
First Party Audience activation
4 to 6 weeks to implement models
CDP - 6 to 9 months to build and implement

ZA Tech - unicorn providing platform for insurance companies
- run insurance on a digital platform
- started grab’s insurance business
- NTUC income
- sister company - leading digital bank in HK
cost saving and cost reduction - more important than revenue - shared by CEOs
Insurers don’t have a lot of data - less number of interactions - limited data - people who have the data are the agents, brokers and are not going to share with you
90% trade conversion of ideas they sent
8 billion transactions per day
Insurers problem
- Pay higher commission to the platform - partners squeeze out stuff
- Getting direct channel is crucial
Hyperpersonalization - Cautionary guardrails
- Huge mis selling
Segment of 1000 is better than segment of one,the latter is more a marketing term
PRA consultation paper - UK - 2022 June on synthetic data
- 11 million records - 90% was able to recoginizable with four additional data points
Insurers get marketing rights from the partners
Everything is long tail - ZA Tech