This document discusses measuring the business value of using Kafka to power event-driven applications. It begins by explaining why measuring value is important for ROI, stakeholder commitment, and benefits realization. It then outlines three real-world examples of using Kafka: resolving ATM disputes faster resulted in 50% less agent time and 75% fewer avoidable payments; a customer 360 application improved targeted offers for increased revenue and better inventory management; and a fraud prevention system enabled real-time detection and prevention, decreasing insurance premiums. The document concludes by recommending establishing credibility through sound assumptions, defining what is actually being measured, and accepting that value is subjective and changing over time.
10. 1010
Innovate or be disrupted...
1. Create New
Business Models
2. Deliver new,
real-time customer
experiences
3. Deliver
massive internal
efficiencies
● Retail: Rent rather than
buy clothing
● Recommendations on
Netflix
● Seeing real-time ETA of
Uber driver
● Text alert - credit-card
fraud
● Global bank consolidating
data silos
● Regulatory reporting
● Corporate office seeing
worldwide sales in real-time
Making More
Money
Saving or Protecting
Money
New Ways of Making Money
16. It’s not just the data that is important, it is how it is
moved around - and how timely it is
→ The ability to move the oil becomes
almost as valuable as the oil itself
→ “Data is only as valuable as your ability to quickly
act on it” - Jay Kreps, co-founder and CEO, Confluent.
19. 19Confidential
3. What does
value mean?
- The Subjective
Element of valuing
something
- Value changes...
$940 k
1999
$65 M
2018
$91 M
2019
97x in 20 yrs
20. 3. What does
value mean?
- The Subjective
Element of valuing a
(contextual) piece of
technology
- Value changes...
1. Cost: The iWatch? $83
2. Price: $350
3. With fitness information?
Medically accurate electrocardiogram?
21. 2121
So, valuing is hard
1. We have to predict the future - with assumptions
and estimates
2. We have to agree on what we’re actually measuring
- the engine or the car? The oil or the pipeline?
Perhaps they’re the same thing?
3. Value is subjective and it changes - we have to be
‘situationally aware’ and be comfortable with that.
27. 27
Universal Event Pipeline
Data Stores Logs 3rd Party Apps Custom Apps/Microservices
from ATM’sContextual Information - Bank Accounts Analysis / Historic info - feeds from other ATM / banks
Universal event pipeline
--- high throughput, persistent,
ordered, and has low latency ---
All events and systems are
connected
ATM Disputes Benefits:
1. From batch to real-time
2. Single Enterprise wide source
of events
3. Built for scale
4. Store events while maintaining
real-time replay
5. ...
28. 28
Universal Event Pipeline
Data Stores Logs 3rd Party Apps Custom Apps/Microservices
ATM
Disputes
App
Real-Time
Fraud
Detection
Real-Time
Customer 360
Machine
Learning
Models
Real-Time
Data
Transformation
...
Contextual Event-Driven Apps
● 50% reduction in agent time
● 75% reduction in avoidable
payments
ATM Disputes
49. 4949
Recap
1. Measuring Value
2. Three Real-World Examples
- ATM Disputes - save money
- Customer 360 - make money
- Fraud - save money
3. Recommendations...
1. Why?
2. Why Kafka?
3. Problems!
1. Credibility?
2. What we’re measuring?
3. Be comfortable with
subjective elements and
change