PortfolioAbout This Topic
When customers are registered separately in the online store (Shopify) and physical stores (POS), there is a need to identify them as the same person and integrate the data.
This is a challenge called "customer matching" or "identity resolution," and it is key to achieving omnichannel experiences.
Why Customer Matching is Necessary
Common Situations
| Case | Situation | Problem |
|---|---|---|
| Offline to Online | Created membership card at store (2 years ago) -> Registered online today | Same person but exists as separate records |
| Online to Offline | Registered online (last month) -> Asked to register at store today | Already an online member but risk of duplicate registration |
Example: Taro Yamada (yamada@example.com) is registered in POS with member number 00001. When registering online today, a separate record is created as Shopify customer ID 12345.
Problems Without Matching
| Problem | Impact on Customer | Impact on Operations |
|---|---|---|
| Split points | Points separate between store and online | Increased complaint handling |
| Fragmented purchase history | Cannot see overall purchase trends | Cannot provide appropriate recommendations |
| Duplicate communications | Same DM arrives twice | Cost increase, poor impression |
| Inconsistent member experience | Treatment differs by channel | Brand value decline |
Basic Matching Approach
This Project's Approach
| Item | Content |
|---|---|
| Primary key | Email address |
| Configuration | Set as required field in both systems |
| Matching timing | Automatic matching at registration |
| Judgment criteria | Treated as same person if matched |
Fallback:
- Email doesn't match -> Treat as new customer
- Manual merge later is also possible
Reasons for adoption: Simple and reliable, easy to automate, no additional burden on customers
For Those Who Want to Learn More
Customer matching is explained in detail in three articles.
1. Matching Key Selection
Which information should be used as a key to match customers between two systems. Comparison and selection rationale for email addresses, phone numbers, names, etc.
2. Address Data Normalization
Address formats differ between Shopify and POS, causing problems if stored as-is. Explanation of address normalization, particularly for Japanese addresses.
3. Sync Error Recovery
How to handle cases where sync fails due to network issues or API rate limits.
Overall Matching Flow
Email address: yamada@example.com
Search Shopify -> Not found / Search POS -> Found!
Email addresses match -> Identified as same person
Shopify: Create as new customer / POS: Update existing customer's member number with Shopify ID
Recognized as same customer in both systems
Benefits of This Mechanism
For Customers
- Recognized as "the same person" across all channels
- Points are integrated, usable at both stores and online
- Receive appropriate recommendations based on purchase history
For Operations
- No duplicate customer data, easier management
- Ability to analyze purchase history across all channels
- Ability to execute effective marketing campaigns