Cost Per Click and Marketing Spend have increased 2-3x over the year. Brands can not control the digital war zone between Google and Facebook.
However, brands are sitting on a big trove of their customer data that they can leverage to create competitive advantage with powerful machine learning to increase efficiency, return on ad spend and lift conversion rate.
Long gone the day of manual segmentation!
Instant Digital Issuance: An Overview With Critical First Touch Best Practices
Marketing + AI - How to optimise conversion rate by leveraging Machine Learning
1. Marketer + AI:
How to optimise conversion rate by
leveraging machine learning?
Jack Nguyen | Country Manager
Vietnam Mobile
Day 2018
2. 16 Country, 250+ Employees, 5 Continents
Trusted by
Singapore
Seoul
Dubai
Tokyo
London
Sydney
Istanbul
Taipei
Kuala Lumpur
Jakarta
Taiwan Ho Chi Minh City
All-in-One Platform
Increase User Engagement
Across All Touchpoints
6. It’s a Disconnected Digital World
“Hey, I want to buy
something
One click to an
eCommerce,
I'm on heaven…”
“Hey, we have amazing
things to sell…”
Hello … Can You Hear Me?
8. Low Conversion Rate High Marketing Cost
Huge Challenges
● Too many choices ● Who to best target?
● Not relevant experience
● Instant Gratification
● Low attention
● Who to best to NOT target?
● How to do it automatically?
● How to do it in real-time?
9. LTV
Lifetime values are
relatively low
because of the
competition &
winner takes it all
Retention
Many online businesses are failing because of low LTV/CAC ratios.
Marketing is Expensive!
Acquisition
CPC & CAC
CPCs increased
288% from 2010 –
2017
17. Machine Learning (ML)
Advanced area of AI
Absorb huge amounts of data and create predictive algorithms that improve over time.
Marketers can offer more personalized content and product suggestions than ever before.
Makes machines do things that would require intelligence if done by a human.
E.g: learning, seeing, talking, socializing, reasoning, or problem solving.
Artificial Intelligence (AI)
18. With customers creating more valuable data at every stage of their journey, AI and Machine Learning
brings endless opportunities for the brands.
A Discount Chaser’s Shopping Journey
1. She comes from a search for “sales” on
Google
2. Then quickly looks for Seasonal Offers on
Homepage
19. A Discount Chaser’s Shopping Journey
3. Filters Sales & Price
Low on Listing Page
4. Ony View
Discounted
Items on
Product Page
5. Use voucher code to make purchase
20. Understand Your Users Accurately
“I always look for great
discount offers and try to
find the best-valued
products.”
“I generally buy expensive
products but buy them very
occasionally.”
“I do a solid research about
the product before buying,
so that I can buy from the
best supplier”
“I shop very much and very
frequently.”
DISCOUNT CHASER PREMIUM SHOPPER INSPIRATION SEEKER RESEARCHER
22. Insider’s AI
Algorithms Across
All Users Touch
Points
Predictive Segmentation
1. Likelihood to Purchase
2. Discount Affinity
3. Interest Clustering
4. Predicted Conversion Date
5. Likelihood to Churn
6. Customer Lifetime Value
Advertising
Desktop Web
Mobile Web
Messaging Mobile App
ACQUISITION
ACTIVATION
RETENTION
REVENUE
Predictive
Audience
DATA without ACTION is a daydream!
23. 1. Likelihood to Purchase
Continuous update on the Predictive Model
Anonymous users are also segmented
Real-Time
Predict the chance of users buying in the next 7 days
24. Target users who are likely to buy and searching for competitor keywords on Google with higher bidding
Use Case on Ad Channels
1. Likelihood to Purchase
25. Understand your users’ affinity toward discount rates to distinguish “discount” and “premium” shoppers.
Use Case on Ad Channels
2. Discount Affinity
26. Use Case on Desktop Web
Exit Intent Recovery
For Premium Shopper:
Sending Survey to their get feedback
For Discount Chaser:
Sending voucher with countdown timer
2. Discount Affinity
27. Use Case on Email Channel
Cart Abandonment Emails
2. Discount Affinity
For Premium Shopper:
Warm reminder of items in cart
For Discount Chaser:
Sending voucher with countdown timer
30. 3. Interest Clustering
Rearrange navigation bar based on Most
Interested Category to increase engagement &
reduce bounce rate
Use Case on Mobile Web
31. 4. Predicted Conversion
Date Range
Monitor and analyze each
visitor’s purchase history
to predict their conversion date range.
1-3
days10-15 days
3-7 days
Use Case on Web Push
32. 5. Likelihood to Churn
Identify visitors who are more likely to churn and create retention strategies before it’s too late.
Use Case on Desktop Web
33. High (VIP)
Low
6. Customer Lifetime Value (CLV)
Identify visitors who are more likely to return and
buy more from you vs those who never come back
34. Users who are predicted as high LTV users and did
not make purchase in last 30 days
Use Case on Web Push
Segment
Automatically send a price-drop web push to
note the item they viewed has been discounted
Action
6. Customer Lifetime Value (CLV)
36. ROASCONVERSION RATE
Success Story - Top 3 Biggest Fashion Retailer in Europe & MENA
Apply high likelihood to purchase audience for dynamic remarketing campaigns.
UPLIFT
160.71%
UPLIFT
629.73%
37. ROASCONVERSION RATE
Success Story - Top 3 Biggest Online Travel Agency in Taiwan
Apply high likelihood to purchase audience for Google AdWords
campaign of hot items.
UPLIFT
407%
UPLIFT
350%
39. 3 Key Things To Remember
Build a Machine Learning that:
● Understand different things about each user
● Update real-time
● Strong action layer at all users touchpoints