Re-engage visitors with smart push.

1Push Notification Playbook

Push notifications allow you to re-engage users even when they are not on your site. You can send targeted messages or personalized product/content recommendations directly to their browser.

How to Start (Wizard Selection)

Go to the Push Notification dashboard. You will see two primary wizards to choose from:

Quick Examples

• Back in Stock: Notify users when an item they viewed is available again.
• Cart Abandonment: Remind users to complete their purchase.
• Price Drop: Alert users when a product they liked goes on sale.

Step-by-Step Summary

1. Collect Subscribers: Use the "Subscribe" wizard to get permission.
2. Select Algorithm: Choose logic (e.g., Cross-sell, Most Popular) in the "Send" wizard.
3. Design: Create the notification (Title, Image, Icon, URL).
4. QA: Preview the notification for specific users.
5. Analyze: Track sent, viewed, clicked, and conversions.

Prerequisites (Must Have)

• Subscriber Database: You must have permission to send notifications. Use Personyze to collect subscribers (see next tab) or sync your existing subscriber list.
• Service Worker: A file (sw.js) must be placed in the root directory of your site to handle the push events.
• Tracking Code: Personyze tracking must be active to trigger messages based on behavior.

2Collect Subscribers

Before sending, you need to ask for permission using the "Web Push Subscribe" wizard.

Design the Prompt

You can customize the opt-in request to maximize acceptance rates. Options include:

• Native Browser Prompt: The default browser request (simple allow/block).
• Custom Banner/Popup: A designed overlay explaining why they should subscribe (e.g., "Get notified when items go on sale").
• Two-Step Opt-In: Show a custom soft-prompt first; if they click "Yes," then show the native browser prompt. This prevents users from permanently blocking your site if they say "No" initially.

3Design & Recommendation Logic

Use the "Web Push Send" wizard to configure your message.

1. Select Recommendation Algorithm

If sending product recommendations, choose the logic:

• Cross-Sells for Items in Cart: "You forgot these..."
• Recently Viewed: "Still interested in...?"
• Back in Stock: Triggers when a previously viewed/liked item returns.

2. Design the Message

Since browser push notifications have a fixed structure, you cannot freely design the layout like an email. However, you can customize:

• Title & Body: "Sale Alert! [Product Name] is 20% off."
• Image: The main product image (dynamically pulled from the feed).
• Icon: Your logo or a specific icon.
• Action Buttons: "View Item" or "Buy Now" (Define the destination URL dynamically).

4A/B Testing for Push

Personyze allows you to A/B test both the Subscription Process and the Push Messages themselves to optimize every stage of the funnel.

1. Testing the Subscription Prompt (Get More Subscribers)

Test different ways to ask for permission to maximize opt-in rates.

• Prompt Style: Test a "Sticky Bar" vs. a "Central Popup".
• Copy: Test "Get Updates" vs. "Don't Miss Out on Sales".
• Timing: Test asking immediately on arrival vs. after 30 seconds or 2 page views.

2. Testing the Push Message (Get More Clicks)

Test the content of the notification to improve Click-Through Rate (CTR).

• Title/Body: Test "Sale Ends Soon!" vs. "Last Chance: 50% Off".
• Algorithm: Test sending "Recently Viewed" items vs. "Most Popular" items to see which drives more sales.
• Imagery: Test showing the Product Image vs. a Lifestyle Image.

5QA & Preview

Verify that the correct products and messages are being generated for specific users.

Simulation Step

1. Go to the QA Step in the wizard.
2. Enter a Subscription ID, User Email, or CRM ID.
3. Personyze will simulate the logic and show you a preview of the push notification that user would receive based on their history.

Check: Does a user who viewed "red shoes" see the "red shoes" in the preview?

6Performance & Analytics

Track the effectiveness of your push campaigns in the dashboard.

Key Metrics

• Sent: Total number of notifications successfully dispatched.
• Viewed: Number of notifications actually delivered and rendered on user screens.
• Clicked: Users who clicked the notification to return to the site.
• Conversions: Downstream actions during the session, such as:
  • Add to Cart: Items added after clicking.
  • Purchased: Direct revenue from the push campaign.
  • Content Goals: "Full Article Read" or "Video Watched".

Tracking: You can wrap links with unique UTM parameters in the design step to track traffic sources in Google Analytics.

7Example: Back In Stock Alert

Scenario: A user views a product that is out of stock, then leaves.

Workflow

1. Trigger: The product feed updates, changing the status of "Item X" from 0 to 10.
2. Audience: Personyze identifies all users who viewed "Item X" in the last 30 days while it was out of stock.
3. Message: "Good news! [Item Name] is back in stock. Grab it before it's gone."
4. Result: User clicks, lands directly on the product page, and purchases.

8Example: Cart Abandonment Reminder

Scenario: A user adds items to the cart but closes the browser without buying.

Workflow

1. Trigger: "Cart Abandonment" event (e.g., 30 minutes of inactivity after add-to-cart).
2. Audience: Users with > $0 in cart value.
3. Message: "You left something behind! Complete your order for [Item Name] now."
4. Result: High click-through rate as the intent to purchase was already established.

9Example: Price Drop Alert

Scenario: An item a user was browsing goes on sale.

Workflow

1. Trigger: Price change detected in the product feed (e.g., $100 -> $80).
2. Audience: Users who viewed this specific item (or category) in the last 7 days.
3. Message: "Price Drop! [Item Name] is now 20% off."
4. Result: Urgent re-engagement driving traffic back to the site for a deal.