2601 Ocean Park Blvd #104 Santa Monica, CA 90405 | 310-598-6658 | RetentionScience.com
How to keep your holiday customers
long after the season is over
#holidaychurn
+ Present:
Matthew Michelson, Ph. D.
Director of Data Science
Retention Science
Today’s speakers
Cara Klewin
Sr. Email Marketing
& CRM Manager
ALOHA.com
Jon Fox
Client Success Manager
Retention Science
About ALOHA
2015 Holiday season sales
are predicted to be as much
as $965 billion, a 4%
increase over last year’s
shopping season.
Agenda
1. What is churn?
2. Reasons Customer Churn
3. Predicting Churn
a) Calculating Churn
b) Predictive Models
4. How to Prevent Customer Churn
5. Campaigns That Work
Deloitte Report
#holidaychurn
What is Churn?
Churn is…
#holidaychurn
When a customer unsubscribes from your service, ceases to purchase
from you, or simply stops engaging with your brand.
• Many companies, however, focus on acquisition rather than churn,
even though research shows it costs five times as much to acquire a
new customer than to keep an existing one.
To take the proactive approach of nurturing customers before they churn
to keep them engaged and purchasing, you need to look at why
customers churn in the first place.
Why are Customers Churning?
#holidaychurn
1. Poor customer service
Of customers who decide to leave a
company, 70% of them to so because of
poor service.
How to prevent it:
Though the holidays are a busy time, it’
s important to remember to take
the time to make customer service a
top priority. When customers feel a
company really cares about their
needs, they are more likely to return
the favor.
• Lifecycle Stages
Overview
#holidaychurn
2. Not enough value
When customers don’t feel they’re
getting value from your brand, they’re
quick to try another brand.
How to prevent it:
Value is a subjective term for
customers, so it’s important to find
out what each customer considers
valuable. For example, some
customers may find free shipping
valuable while others want unique
product recommendations.
#holidaychurn
3. Poor quality communications
You can no longer expect to batch and
blast your customers and get solid results.
Customers want -- actually demand --
quality communications from brands.
How to prevent it:
Give your communications a purpose, m
ake them relevant to the individual
and engage your customers in
conversations to improve your overall qu
ality.
• Lifecycle Stages
Overview
#holidaychurn
4. No brand loyalty
During the holiday season, many
customers buy from brands they don’t
usually purchase from when getting gifts.
How to prevent it:
Give new customers a reason to be
loyal to your brand. Create a
memorable experience: offer
assistance when making first
purchases or give incentives for the
user to get to know your brand more by
following you on social media.
When a customer feels you’ve let them
down, they’re quick to head to the door.
And with all the online outlets, they’re
even quicker to let their networks know
about it.
#holidaychurn
5. Unmet expectations
How to prevent it:
Underpromise and overdeliver.
Customers are delighted when you
exceed their expectations, especially
during the hectic holiday season. If
you are unable to fulfill a promise, be
transparent and remember churn
reason #1 – customer service can
make all the difference.
Predicting Churn
v
Now that you know why customers churn, it’s
time to look at how to predict which of your
customers are at risk of churning or who have
already churned.
#holidaychurn
Basic Churn Calculation
You can use Churn Rate to calculate:
• Number of customers lost
• Value of recurring business lost
• Percentage of recurring revenue lost
#holidaychurn
Churn can be tricky to calculate. This is the most basic calculation:
Using Data Science to Predict Churn
We assign a churn score to each customer that tells us their probability of
churning.
Based on scores, users are put into 3 buckets:
#holidaychurn
1. Low churn score = “Ready to Buy”
- These customers will make a purchase at some point
2. Medium churn score = “At Risk”
- These customers might make a purchase, or might churn
3. High churn score = “Churned”
- These customers are likely to not purchase again
Using Data Science to Predict Churn
#holidaychurn
Retention Science updates 100 million users per day to determine where they fall on
the churn spectrum.
Markov Churn Model
#holidaychurn
Preventing Churn
Tips for Preventing Churn
• Know who your customer is and if they change throughout the year.
For example, jewelry companies may target women most of the year, but need to
change their messaging around Valentine’s Day to target men.
• Provide a smooth customer journey from start to finish.
Test each stage of the purchasing process yourself to make sure it’s intuitive and easy
to follow.
• Reward your advocates as they are the most valuable customers you’ll have.
Advocates promote your brand to their friends, family, and networks – reward them
for providing reviews and getting their friends to become customers.
• Time is of the essence when it comes to recapturing customers.
The window to recapture interest is smaller than ever, make sure you have a reaction
strategy in place to address customer actions.
How ALOHA re-engages
customers who have
churned
after the Holidays
Win-Back Campaigns
A series of marketing messages with
the core function of bringing a lapsed
customer back into the purchasing
cycle.
Two main components:
– Messaging that addresses the
reason for lapse
– An enticing offer – i.e. a
discount on a future purchase or
a free month (if dealing with a
subscriber)
Free Shipping or Percentage Off
• Win-Back offer is
dynamically inserted into
emails
• Customers with highest
CFV (Customer Future
Value) get best offer
#holidaychurn
Pair Content with Products & Offers
#holidaychurn
Give Customers a Choice
#holidaychurn
Giving customers a choice to
discover new products via free
samples is a great way to re-engage
customers for the new year.
Recap
Avoid Churn this Holiday Season:
#holidaychurn
• There are many reasons your
customers may churn like poor
customer service and not providing
enough value.
• Knowing churn rate is useful in
calculating the value of recurring
business lost
• Data science helps predict churn by
calculating churn score and
bucketing customers into three
categories: Ready to Buy, At Risk,
and Churned
• The Markov model is one model
Retention Science uses to predict
the probability a customer will make
another purchase.
• To prevent churn: Know your
customer, provide a smooth
customer journey, reward your
advocates, & timing is everything
• Win-Back campaigns are a great
way to re-engage customers who
have churned
2601 Ocean Park Blvd #104 Santa Monica, CA 90405 | 310-598-6658 | RetentionScience.com
Questions?
#holidaychurn

Predicting and Preventing Customer Churn After the Holiday Season

  • 1.
    2601 Ocean ParkBlvd #104 Santa Monica, CA 90405 | 310-598-6658 | RetentionScience.com How to keep your holiday customers long after the season is over #holidaychurn + Present:
  • 2.
    Matthew Michelson, Ph.D. Director of Data Science Retention Science Today’s speakers Cara Klewin Sr. Email Marketing & CRM Manager ALOHA.com Jon Fox Client Success Manager Retention Science
  • 3.
  • 4.
    2015 Holiday seasonsales are predicted to be as much as $965 billion, a 4% increase over last year’s shopping season. Agenda 1. What is churn? 2. Reasons Customer Churn 3. Predicting Churn a) Calculating Churn b) Predictive Models 4. How to Prevent Customer Churn 5. Campaigns That Work Deloitte Report #holidaychurn
  • 5.
  • 6.
    Churn is… #holidaychurn When acustomer unsubscribes from your service, ceases to purchase from you, or simply stops engaging with your brand. • Many companies, however, focus on acquisition rather than churn, even though research shows it costs five times as much to acquire a new customer than to keep an existing one. To take the proactive approach of nurturing customers before they churn to keep them engaged and purchasing, you need to look at why customers churn in the first place.
  • 7.
  • 8.
    #holidaychurn 1. Poor customerservice Of customers who decide to leave a company, 70% of them to so because of poor service. How to prevent it: Though the holidays are a busy time, it’ s important to remember to take the time to make customer service a top priority. When customers feel a company really cares about their needs, they are more likely to return the favor.
  • 9.
    • Lifecycle Stages Overview #holidaychurn 2.Not enough value When customers don’t feel they’re getting value from your brand, they’re quick to try another brand. How to prevent it: Value is a subjective term for customers, so it’s important to find out what each customer considers valuable. For example, some customers may find free shipping valuable while others want unique product recommendations.
  • 10.
    #holidaychurn 3. Poor qualitycommunications You can no longer expect to batch and blast your customers and get solid results. Customers want -- actually demand -- quality communications from brands. How to prevent it: Give your communications a purpose, m ake them relevant to the individual and engage your customers in conversations to improve your overall qu ality.
  • 11.
    • Lifecycle Stages Overview #holidaychurn 4.No brand loyalty During the holiday season, many customers buy from brands they don’t usually purchase from when getting gifts. How to prevent it: Give new customers a reason to be loyal to your brand. Create a memorable experience: offer assistance when making first purchases or give incentives for the user to get to know your brand more by following you on social media.
  • 12.
    When a customerfeels you’ve let them down, they’re quick to head to the door. And with all the online outlets, they’re even quicker to let their networks know about it. #holidaychurn 5. Unmet expectations How to prevent it: Underpromise and overdeliver. Customers are delighted when you exceed their expectations, especially during the hectic holiday season. If you are unable to fulfill a promise, be transparent and remember churn reason #1 – customer service can make all the difference.
  • 13.
  • 14.
    v Now that youknow why customers churn, it’s time to look at how to predict which of your customers are at risk of churning or who have already churned. #holidaychurn
  • 15.
    Basic Churn Calculation Youcan use Churn Rate to calculate: • Number of customers lost • Value of recurring business lost • Percentage of recurring revenue lost #holidaychurn Churn can be tricky to calculate. This is the most basic calculation:
  • 16.
    Using Data Scienceto Predict Churn We assign a churn score to each customer that tells us their probability of churning. Based on scores, users are put into 3 buckets: #holidaychurn 1. Low churn score = “Ready to Buy” - These customers will make a purchase at some point 2. Medium churn score = “At Risk” - These customers might make a purchase, or might churn 3. High churn score = “Churned” - These customers are likely to not purchase again
  • 17.
    Using Data Scienceto Predict Churn #holidaychurn Retention Science updates 100 million users per day to determine where they fall on the churn spectrum.
  • 18.
  • 19.
  • 20.
    Tips for PreventingChurn • Know who your customer is and if they change throughout the year. For example, jewelry companies may target women most of the year, but need to change their messaging around Valentine’s Day to target men. • Provide a smooth customer journey from start to finish. Test each stage of the purchasing process yourself to make sure it’s intuitive and easy to follow. • Reward your advocates as they are the most valuable customers you’ll have. Advocates promote your brand to their friends, family, and networks – reward them for providing reviews and getting their friends to become customers. • Time is of the essence when it comes to recapturing customers. The window to recapture interest is smaller than ever, make sure you have a reaction strategy in place to address customer actions.
  • 21.
    How ALOHA re-engages customerswho have churned after the Holidays
  • 22.
    Win-Back Campaigns A seriesof marketing messages with the core function of bringing a lapsed customer back into the purchasing cycle. Two main components: – Messaging that addresses the reason for lapse – An enticing offer – i.e. a discount on a future purchase or a free month (if dealing with a subscriber)
  • 23.
    Free Shipping orPercentage Off • Win-Back offer is dynamically inserted into emails • Customers with highest CFV (Customer Future Value) get best offer #holidaychurn
  • 24.
    Pair Content withProducts & Offers #holidaychurn
  • 25.
    Give Customers aChoice #holidaychurn Giving customers a choice to discover new products via free samples is a great way to re-engage customers for the new year.
  • 26.
  • 27.
    Avoid Churn thisHoliday Season: #holidaychurn • There are many reasons your customers may churn like poor customer service and not providing enough value. • Knowing churn rate is useful in calculating the value of recurring business lost • Data science helps predict churn by calculating churn score and bucketing customers into three categories: Ready to Buy, At Risk, and Churned • The Markov model is one model Retention Science uses to predict the probability a customer will make another purchase. • To prevent churn: Know your customer, provide a smooth customer journey, reward your advocates, & timing is everything • Win-Back campaigns are a great way to re-engage customers who have churned
  • 28.
    2601 Ocean ParkBlvd #104 Santa Monica, CA 90405 | 310-598-6658 | RetentionScience.com Questions? #holidaychurn