How to Configure Klaviyo AI's Send Time Optimization and Predictive Recommendations for Automated Flows

Automated email flows—such as abandoned checkout sequences, post-purchase follow-ups, and replenishment reminders—drive a disproportionate share of e-commerce revenue compared to broadcast campaigns. However, standard automation flows often suffer from two major shortcomings: rigid time delays that deliver emails when customers are asleep or distracted, and generic product grids that fail to reflect individual shopping preferences.

Klaviyo addresses both issues by combining machine learning models with native customer data. To get the maximum return on investment from these tools, merchants must understand how Klaviyo's send time algorithms operate across campaigns and flows, how dynamic product recommendation engines curate catalog items, and how to configure them inside the Klaviyo flow builder.

1. Understanding Klaviyo AI's Send Time Optimization: Campaigns vs. Flows

A common source of confusion among merchants is the difference between Klaviyo's timing optimization tools. Klaviyo offers two distinct AI timing mechanisms, and understanding how they interact with automated flows is essential for effective setup.

Smart Send Time vs. Personalized Send Time

• Smart Send Time: Discovers the single best sending hour for your audience as an aggregate whole. It runs a 24-hour exploratory phase where emails are sent randomly across an audience in each recipient's local time zone, followed by a focused send at the optimal window. This feature requires at least 12,000 active profiles and campaign recipient lists of 12,000 or more.
• Personalized Send Time: An individual-level machine learning model that predicts the specific hour each recipient is most likely to open, click, or place an order based on their unique past interaction history.

How Timing Works in Automated Flows

Neither Smart Send Time nor Personalized Send Time functions as a simple drag-and-drop node inside automated event-triggered flows. Automated flows are inherently triggered by real-time customer behavior (such as Started Checkout, Added to Cart, or Placed Order). Klaviyo manages flow timing through Time Delay components, Quiet Hours, and Predictive Analytics properties:

1. Flow Time Delays with Local Time Restrictions: In any flow, merchants can set a delay (e.g., 2 hours, 1 day) and restrict delivery to specific hours in the recipient's local time zone (e.g., "Send only between 9:00 AM and 8:00 PM in recipient's local time").
2. Applying Campaign Timing Insights to Flow Delays: Data gathered from Smart Send Time and campaign reports provides the baseline conversion hours for your store. If exploratory sends show peak purchasing between 7:00 PM and 9:00 PM local time, your second abandoned checkout reminder or winback email can be scheduled to hold until that peak window.
3. Predictive Analytics Timing: For retention and repeat purchase flows, Klaviyo AI calculates profile-level attributes:
   • Expected date of next order: Machine learning forecasts when a specific shopper is due to buy again based on historical order cadence.
   • Predicted Customer Lifetime Value (CLV): Categorizes profiles into high, medium, and low expected spenders.
   • Churn Risk: Identifies whether a customer is active, at risk, or dormant.

2. Configuring Predictive Product Recommendations in Klaviyo

Predictive product recommendations in Klaviyo rely on synced catalog data (via Shopify, BigCommerce, WooCommerce, or custom API feeds) and individual browsing and purchasing events.

Step 1: Create and Configure Dynamic Product Feeds

Before embedding recommendations into flow templates, set up dedicated feeds in the Klaviyo dashboard:

1. Navigate to Content > Products in the left-hand navigation.
2. Select the Product feeds tab and click Create Product Feed.
3. Name your feed based on its strategic purpose (e.g., Post-Purchase Cross-Sell, Cart Abandonment Alternatives, or Trending Fallback).
4. Select the recommendation algorithm:
   • What customers have purchased (Customers Also Bought): Evaluates co-purchase patterns across your catalog. If a customer bought Product A, the feed surfaces Products B and C that other shoppers frequently bought in subsequent or bundled orders. Best for post-purchase and repeat-purchase flows.
   • Recently viewed products: Dynamically surfaces items the customer browsed during recent web sessions. Best for browse abandonment and high-intent recovery flows.
   • Popular / Best-selling products: Displays top revenue-generating items across the store or within selected categories. Essential as fallback logic.
5. Set filtering and exclusion rules:
   • Filter by collections or categories to avoid irrelevant cross-sells (e.g., excluding gift cards, clearance parts, or apparel sizes).
   • Confirm that out-of-stock items are automatically excluded.
   • Enable the suppression of items already purchased by the recipient so customers are never shown items they already own (ideal for non-consumable goods).
6. Configure Fallback Logic: When a shopper has insufficient behavioral history for personalized recommendations, Klaviyo automatically falls back to your best-selling catalog items. Ensure your fallback sort order is defined.

Step 2: Insert Dynamic Product Blocks into Flow Emails

1. Open your target flow in the Flow Builder (e.g., Abandoned Cart or Post-Purchase Series).
2. Click the email action node and select Edit.
3. In the email template editor, drag the Product block from the left panel into your layout.
4. In the block configuration panel, change the mode from Static to Dynamic.
5. Select your newly created Product Feed from the dropdown menu.
6. Customize the display attributes:
   • Set the layout (1 to 4 columns).
   • Toggle product title, image, price, and "Shop Now" buttons.
   • Ensure links dynamically append tracking parameters to measure flow-driven revenue.

3. Implementation Playbook for Core Customer Flows

4. Detailed Configuration for Key Flows

A. Abandoned Checkout Recovery

• Primary Conversion Goal: Remind the customer of what they left behind while offering related options if price or sizing caused friction.
• Timing Structure:
  • Email 1 (Hour 1–2): Quick transactional reminder. The shopper's intent is still warm. Include dynamic cart tags ({{ event.extra.line_items }}).
  • Email 2 (Hour 24): Schedule with a Time Delay set to 24 hours, restricted to local delivery between 10:00 AM and 7:00 PM.
• Product Feed Setup:
  • Below the abandoned cart items, insert a Dynamic Product Block configured with a category-filtered "Best-Sellers" or "Customers Also Bought" feed.
  • If the shopper hesitated because the original item was too expensive or out of style, the recommendation block provides alternative purchase paths.

B. Post-Purchase and Repeat Purchase Retention

• Primary Conversion Goal: Increase second-order rates and maximize Customer Lifetime Value.
• Timing Structure:
  • Instead of arbitrary 30-day delays, analyze Klaviyo's Repeat purchase timing analytics cards located in your Analytics dashboard.
  • If your average second order occurs at 42 days, set up your primary cross-sell touchpoint at Day 14 (immediate complement) and your second touchpoint at Day 35 (anticipating replenishment).
• Product Feed Setup:
  • Configure a feed using the algorithm "What customers have purchased".
  • Klaviyo checks the customer's initial purchase and ranks products by cross-order probability.
  • For consumables (skincare, supplements, specialty food), create separate replenishment flows triggered by predictable usage windows, using "Refill / Reorder" feeds.

C. Predictive Winback & Churn Prevention

• Primary Conversion Goal: Re-engage high-value customers before they permanently lapse.
• Segment Trigger: Create a dynamic segment defined by:
  • Predictive Analytics > Churn Risk is High
  • AND Predictive Analytics > Historic CLV > $0
• Timing Structure: Deliver messages during daytime hours in the recipient's local time zone. Avoid overnight sends that degrade deliverability.
• Product Feed Setup:
  • Do not show obsolete products from their distant browsing history.
  • Use a "New Arrivals & Best Sellers" feed to show fresh catalog additions that the customer has not yet seen.

5. Best Practices, Pitfalls, and Deliverability Safeguards

1. Verify Minimum Data Thresholds: Klaviyo's predictive models (CLV, Churn Risk, Expected Order Date) require sufficient transaction volume. If your store has fewer than several hundred completed orders, predictive properties may remain unpopulated. Allow historical data to sync thoroughly after initial store integration.
2. Always Configure Fallbacks: Never assume every visitor has a populated browsing profile. If a new subscriber enters a flow, an unconfigured recommendation feed can show blank spaces. Always set a fallback rule to display best-selling catalog items.
3. Respect Quiet Hours: Ensure your flow delays do not release automated emails or SMS during late-night or early-morning windows. Use Klaviyo's "Quiet Hours" and "Only send during specific hours" settings to align delivery with waking hours.
4. Avoid Duplicate Purchases on Non-Consumable Catalogs: In furniture, electronics, and apparel, selling a customer an identical item rarely works. Ensure the "Exclude already purchased products" setting is toggled on in your dynamic product feed settings.
5. Monitor Revenue Lift via Flow A/B Testing: Use Klaviyo's Flow Branching A/B test feature to compare an email containing only static copy against one with a Dynamic Product Recommendations block. Track placed order rate, click rate, and revenue per recipient to validate performance gains.

References

• AI Workflow Automation Tools for Better Marketing & Customer Service - Klaviyo
• Omnichannel Personalization Software: Build Better Marketing & CX - Klaviyo
• Email AI: How to Use AI to Write an Email for Marketing - Klaviyo
• Predictive Analytics for Ecommerce: Forecast Who Buys Next - Klaviyo
• The Marketing Automation Platform for Email, SMS, & WhatsApp - Klaviyo