This document discusses strategies for retaining customers beyond basic loyalty programs. It emphasizes using data science and precision marketing techniques to foster emotional loyalty. Key points include segmenting customers using migration patterns, building predictive models tailored to specific needs, designing experiments to test individual messages and communication streams, and catering marketing to each customer's unique journey over time. The goal is to achieve the "ultimate" in customer marketing through speed, emotional intelligence, and a dynamic, data-driven approach.
2. https://www.linkedin.com/in/yoavsusz/
Yoav_s@Optimove.com
Yoav Susz leads Optimove’s New York-based business development team,
where he is playing a major role in Optimove’s successful expansion into the
US market. Yoav specializes in helping retail brands build emotionally-
intelligent retention marketing strategies that focus on leveraging each
brand’s core identity. In his position, Yoav focuses on building relationships
with partners across the Americas. Prior to joining Optimove, Yoav served as
a corporate litigator for Israel’s leading law firm, where he represented some
of the world’s leading technology companies. Yoav holds an LLB from The
Interdisciplinary Center and an MBA from the Stern Business School at NYU.
3. 03
More information is available at
www.optimove.com
Science-first Customer Marketing Hub
13. 13
Data Science / First level of acquaintance – divide and conquer
Migration is actually the biggest part of segmentation
From / To
Registered
Only
New Active One Timers Churn
Churn One
Timers
Dormant
Registered Only
New
Active
One Timers
Churn
Churn One Timers
Dormant
80.5%
0%
0%
0%
0%
0%
0%
11.5%
60.0%
0%
0%
0%
0%
0%
0%
8.2%
89.0%
3.5%
5.6%
1.5%
1.5%
0%
31.8%
0%
89.0%
0%
0%
0%
0%
0%
11.0%
0%
92.0%
0%
0%
0%
0%
0%
7.5%
0%
91.5%
0%
8.0%
0%
0%
0%
2.4%
7.0%
98.5%
14. 14
Data Science / Predictive models
As humans, in our day to day, we do ‘predictive’ all the time by factoring in our past experiences
In many cases, simple heuristics will do the trick unless you are sensitive to performance
PPC example - Optimize on first purchase Vs. an LTV model
When you do build a model, things to remember:
• They are as good as the raw data they seat upon
• They require customization
• One needs a ton of talent and experience to build
a proper model
15. 15
One interaction with one message
One interaction with a few messages
Communication streams
Experiments Design
20. 20
Competing strategies
Communication Streams
• The problem of hidden cannibalization (what if I only got
revenue sooner, but did not impact LTV?)
• How can I prove that I’m really getting an uplift? (Maybe
what I was doing before Optimove was just as good…)
• The concept of “forever control” – the only way to
measure this scientifically