3.2 — AI-Powered Customer Segmentation & Retargeting
AI-Powered Customer Segmentation & Retargeting
You have 1,000 customers on WhatsApp. But they're not all the same. Some are bargain hunters. Some are premium buyers. Some bought once and disappeared. This lesson teaches you to segment customers by behavior and AI-target them with different messages.
Why Segmentation Matters
No Segmentation: You send same message to everyone. "Limited offer: 50% off everything!"
Result:
- Premium customers feel disrespected (they don't buy discounted goods)
- Budget-conscious customers skip it (not deep enough discount)
- First-time leads don't understand what you do
- Conversion: 2–5%
With Segmentation: You send different messages to different groups.
Premium segment: "Exclusive early access: New Premium collection. VIP-only, 15% off" Budget segment: "Flash sale: Everything 50% off. Today only" Leads segment: "Popular items this week. Most bought: [product]"
Result:
- Premium customers feel valued (buy full price or exclusive items)
- Budget customers go crazy (deeper discount for them)
- Leads learn what sells (social proof)
- Conversion: 15–25%
Improvement: 3–5x higher conversion just by segmenting.
Customer Segments You Should Track
Segment 1: VIP / Repeat Customers
- Who: Bought 3+ times OR total spent >PKR 20,000
- Behavior: Loyal, price-insensitive, want early access
- Messaging: Exclusive offers, early launches, VIP perks
- Example: "Reserved for you: New collection launches Friday. Click early access: [link]"
- Conversion rate: 25–40%
Segment 2: Warm Leads / One-Time Buyers
- Who: Bought once OR showed interest but didn't buy
- Behavior: Curious, need reassurance, respond to social proof
- Messaging: Case studies, reviews, complementary products, small discounts
- Example: "Other customers loved this item. See reviews: [link]. 20% off for first-time buyers: [code]"
- Conversion rate: 10–20%
Segment 3: Cold Leads / Browsers
- Who: Inquired but never bought
- Behavior: Price-sensitive, need education, high churn
- Messaging: Value props, testimonials, rock-bottom offers
- Example: "Quick question: What stopped you last time? Budget? Shipping time? Let's solve it: [reply]"
- Conversion rate: 3–8%
Segment 4: At-Risk / Inactive
- Who: Haven't purchased in 60+ days
- Behavior: Likely moved to competitor, need reactivation
- Messaging: Win-back offers, new features, reminder of benefits
- Example: "We miss you! New products you'll love. Come back? 30% off: [code]"
- Conversion rate: 5–15%
Segment 5: Industry / Interest-Based
- Who: Customers interested in specific categories
- Behavior: Predictable, targetable, high-intent
- Messaging: Specific to their interest
- Example: If fashion: "New winter collection" | If electronics: "Latest laptop deals"
- Conversion rate: 20–35%
How to Tag Customers in WATI
WATI has a "Labels" feature. Use it to segment:
Setup:
- WATI Dashboard → Settings → Labels
- Create labels:
VIP_RepeatWarm_LeadCold_LeadAt_RiskProduct_FashionProduct_Electronics
- When customer messages, add label with one tap
Automation via WATI Rules:
IF customer total spent > PKR 20,000
THEN auto-add label: VIP_Repeat
IF customer bought in last 30 days
THEN auto-add label: Warm_Lead
IF customer last message > 90 days ago
THEN auto-add label: At_Risk
AI Segmentation — Smart Grouping
Manual labeling works, but AI does it automatically.
Example: AI Rule "If customer asks about [womens clothing], auto-label: Interest_Fashion"
Your AI system learns:
- Customer asks about jeans → Label: Interest_Fashion
- Customer asks about budget → Label: Budget_Sensitive
- Customer buys expensive items → Label: Premium_Buyer
- Customer waits to see reviews → Label: Risk_Averse
No manual work. The bot learns and tags automatically.
WATI Setup:
- Integrations → AI > Configure Smart Labels
- Set rules based on keywords/behavior
- AI auto-applies labels going forward
Creating Segment-Specific Messages
Once segmented, send different messages to each group:
Message Template for VIP Segment:
Subject: VIP Early Access — New Collection
Audience: VIP_Repeat customers
Send time: Friday 6 PM (before general release)
"Assalam-o-alaikum [Name]! 💚
You're our best customer. Here's a gift:
🎁 EXCLUSIVE VIP Early Access
New Winter Collection launches Friday.
You get 48 hours early + 15% VIP discount.
Preview: [link]
Reserve now (VIP only): [link-vip]
Code: VIP_WINTER_15
Valid: Friday 6 PM – Sunday 6 PM
Your loyalty means everything. Thanks for being part of the family! 🙏"
Message Template for Cold Lead Segment:
Subject: We've Missed You — 50% Off Everything
Audience: Cold_Lead who inquired but didn't buy
Send time: Tuesday 7 PM
"Assalam-o-alaikum [Name],
We noticed you were interested a while back. Here's why you should come back:
✅ Quality: 1,000+ 5-star reviews
✅ Speed: Delivery to your city in 2–3 days
✅ Trust: PayPal Seller Protection + 30-day returns
✅ Price: Today only — 50% off everything
Don't miss out: [link]
Use code: COMEBACK50
Questions? Reply here. We're listening 💚"
Message Template for At-Risk Segment:
Subject: Come Back — Special Win-Back Offer
Audience: At_Risk (no purchase in 90+ days)
Send time: Wednesday 5 PM
"Hi [Name],
It's been [X] months since we last chatted. A lot has changed:
🆕 New products in your favorite categories
💰 New payment plans available (installments)
🚀 Faster shipping now (2 days vs 3)
See what's new: [link]
Special offer for you: [code]
We'd love to work with you again. Any concerns from last time? Reply — we'll fix them 💚"
Broadcast Campaign Strategy per Segment
Weekly Calendar:
| Day | Segment | Campaign | Expected Conversion |
|---|---|---|---|
| Monday | Cold_Lead | Educational tip | 2% |
| Tuesday | At_Risk | Win-back offer | 10% |
| Wednesday | Warm_Lead | Social proof + discount | 15% |
| Thursday | Interest-based | Product category focus | 20% |
| Friday | VIP_Repeat | Exclusive early access | 30% |
| Saturday | All (broad) | Flash sale | 8% |
| Sunday | All (broad) | Newsletter/value | 3% |
This prevents message fatigue (each segment gets 1–2 messages/week max) while maximizing conversions.
Retargeting — The Second Chance
Customers who abandoned their carts or browsed but didn't buy need retargeting.
Retargeting Sequence (48-hour window):
[Immediately after browsing] Bot: "Saw you looking at [product]. Still interested? Price: PKR 2,500 In stock: Yes Reserved for you: 1 hour Book now: [link]"
[After 12 hours if no action] Bot: "Quick reminder: [product] is almost sold out. In stock: Only 3 left Reserved for you: Expired (but can re-reserve) Secure yours: [link]"
[After 24 hours] Bot: "Last chance notification: [Product] sold out for this color. Want alternative color? [link] OR get notified when back in stock? Reply: NOTIFY"
[After 48 hours] Bot: "Closed: Your reservation for [product]. Similar items in stock: [link]"
Result: 10–20% of abandoned browsers convert on retargeting.
Measuring Segment Performance
Track metrics per segment:
| Segment | Members | Last Broadcast | Open Rate | Click Rate | Conversion Rate | Avg Order Value |
|---|---|---|---|---|---|---|
| VIP_Repeat | 45 | "New collection" | 95% | 45% | 32% | PKR 8,500 |
| Warm_Lead | 120 | "Review + discount" | 65% | 28% | 14% | PKR 4,200 |
| Cold_Lead | 280 | "50% sale" | 40% | 12% | 5% | PKR 2,800 |
| At_Risk | 95 | "Win-back" | 55% | 18% | 10% | PKR 3,500 |
Key insight: VIP segment gets 6x higher conversion than Cold segment. Spend more effort acquiring VIP-quality customers, fewer on cold leads.
Practice Lab
Task 1: Define Your Segments For your business, create 5 segments:
- Segment name: ___________ Who: ___________ How to identify: ___________
[Repeat for segments 2-5]
Task 2: Create Segment-Specific Messages Write a 50–100 word message for your highest-value segment (VIP):
___________
Write a 50–100 word message for your riskiest segment (Cold Lead):
___________
Task 3: Plan Your Weekly Broadcast Calendar For 7 days, plan which segment gets which message:
- Monday: [Segment] | [Message type]
- Tuesday: [Segment] | [Message type] [...etc]
Pakistan Example: Fashion Brand AI Segmentation
Meet Amber — runs a sustainable fashion brand in Islamabad. 500 WhatsApp contacts but no strategy.
She implemented AI segmentation:
Segments Created:
-
VIP_Repeat (40 people): Bought 3+ times, avg order PKR 8,000+
- Message: Exclusive preview of new collections
- Frequency: 1–2x/week
- Discount: 15% VIP code
-
Warm_Lead (150 people): Bought once or twice
- Message: "Complete your collection" + complementary items
- Frequency: 1x/week
- Discount: 10% loyalty code
-
Browsers (200 people): Viewed catalog but didn't buy
- Message: Social proof + educational tips
- Frequency: 2x/month (less frequent)
- Discount: 20% first-buy incentive
-
At_Risk (110 people): Haven't purchased in 120+ days
- Message: "We miss you" + new arrivals
- Frequency: 1x/month
- Discount: 30% win-back offer
Results After 2 Months:
| Segment | Before | After | Growth |
|---|---|---|---|
| VIP_Repeat | PKR 12,000/mo | PKR 45,000/mo | +275% |
| Warm_Lead | PKR 8,000/mo | PKR 22,000/mo | +175% |
| Browsers | PKR 2,000/mo | PKR 8,000/mo | +300% |
| At_Risk | PKR 0/mo | PKR 6,000/mo | ∞ |
| Total | PKR 22,000/mo | PKR 81,000/mo | +268% |
The secret? She stopped treating all customers the same. VIP got luxury messaging. Budget customers got discounts. Browsers got education. Each felt personally targeted.
Key Takeaway: Segmentation is not complex. It's just labeling customers by who they are, then sending messages that match their stage. AI makes it automatic. Result: 2–3x higher conversion with the same message volume.
Next lesson: Capstone — Build your complete WhatsApp commerce empire combining everything you've learned.
Lesson Summary
AI-Powered Customer Segmentation
4 questions to test your understanding. Score 60% or higher to pass.