Onboarding Recommendation
This guide provides a step-by-step process for onboarding new clients to Netcore Unbxd's Recommendation engine (Recs) and integrating it with Netcore's CE panel.
Sales Coordination
Team | Actions |
---|---|
Netcore Support | Contact Unbxd Sales team to initiate the Monthly Recurring Revenue (MRR) purposes. |
Email to reach | [email protected] |
Customer Onboarding Process
1. Client Access to CE Panel
Netcore Team grants client access to the CE (Customer Engagement) panel. Contact Netcore Support team to get your access.
2. Request to Unbxd Onboarding Team
Netcore team need to raise a request to Unbxd Onboarding Team to onboard client to Recommendations panel. Please contact:
3. Feed Integration by Netcore Unbxd
Important
Netcore ID data triggers model training, which takes two weeks; previous client data is explored for training.
The two-week duration is essential for gathering sufficient user data and ensuring the accuracy of the trained model.
Unbxd onboarding team performs the following actions:
- Feed Uploading: Upload the product feed to Unbxd system.
- Dimension Mapping: Map necessary dimensions post-upload.
- Event Setup: Configure analytics events: Impression, Click, Add to Cart, Orders.
- Netcore ID Check: Ensure checkbox for Netcore ID data tracking is enabled (manual task if required).
- Model Training Process
- The recommendation model is trained using the Netcore Customer ID.
- Once the event tracking and feed integration are completed, the TPM (Technical Project Manager) will trigger the model training pipeline.
- Model training typically takes up to 2 weeks, as it requires sufficient user interaction data.
- To reduce training time, we should explore the possibility of using historical data provided by the client.
- The ML team will share updates once the model has been successfully trained.
- API & Key Delivery: After training the TPM team shares the API Key and Site ID with the Netcore team.
TPM also conducts a demo session on setting up experiences via CE, Using email pages, Custom algorithm configuration if needed.
- Final Testing: After the keys are integrated, the Netcore team will conduct final tests to ensure the recommendation engine is working seamlessly and that the system is ready to go live.
Refer here how to set up Site ID and API keys from Unbxd Recs platform.
4. Enabling Journey/Broadcast in Unbxd Platform
Toggle Activation: Enable Journey/Broadcast toggle in Unbxd platform. CSM needs to create the required experience for Customers.
Note
Toggle must be activated before Journey/Broadcast pages are visible.
5. Experience Creation Guidelines
The experience names for both Journey and Broadcast pages must match. For the list of Supported Algorithms & Widgets, refer here.
Important Point to Remember
While creating algorithms on the Unbxd panel, avoid using special characters like $, @, #, and so on, because using these can negatively impact email communications.
For example: Most Viewed Price 20 DASH 60$, the $ symbol can lead to the email communication being dropped.
The below table provides information for different experiences, actions, and necessary widgets for both the broadcast and journey pages.
Experience | User Story | Action | Widget | Notes |
---|---|---|---|---|
Best Selling Products | User can call for best-selling products on both pages. | Create experience with Top Seller Algorithm. | Widget 1 | |
Most Viewed Products | User can call for most viewed products on both pages. | Create experience with Most Viewed Algorithm. | Widget 2 | Not supported right now. |
Random Products | User can call for random products on both pages. | Create experience with Top Sellers + Most Viewed Algorithms. | Widget 3 | |
Top Sellers (Category-Based) | User can get top-selling category recommendations. | Create experience with Category Top Seller Algorithm. | Widget 4 | |
Trending Products | User can call for trending products on both pages. | Create experience with Top Seller Algorithm (Look-by period = 1). | Widget 5 | |
Recently Viewed Products | User can call for recently viewed products on journey page. | Create experience with Recently Viewed Algorithm. | Widget 6 | |
Frequently Bought Together (BAB) | User can call for frequently bought together products. | Create experience with BAB Algorithm. | Widget 7 | Product ID required in the API by the Netcore team. |
Similar Products | User can call for similar products on journey page. | Create experience with Similar Product Algorithm. | Widget 8 | Product ID required in the API by the Netcore team. |
Cart-Specific Recommendations | User can call for cart-specific recommendations. | Create experience with Cross Sell Algorithm. | Widget 8 | Product ID required in the API by the Netcore team. |
Recommendations Based on Last Purchase | User can call for recommendations based on last purchase. | Create experience with Purchase-Based Recommendations Algorithm. | Widget 8 | Product ID & Netcore ID required in the API. Use Recommended for You. |
Recommendations Based on Last Viewed | User can call for recommendations based on last viewed. | Create experience with Viewed-Based Recommendations Algorithm. | Widget 8 | Product ID & Netcore ID required in the API. Enable VAV-V2. |
6. Resources
Updated 6 days ago