What are AI Recommendations?
AI Recommendations is a powerful feature that analyzes your customer conversations and provides actionable insights to improve your customer experience (CX). By identifying patterns in shopper interactions, your AI automatically generates specific suggestions to address customer pain points, enhance your product information, and optimize your website.
Accessing AI Recommendations
AI Recommendations is included as part of the Deep Research add-on. To access this feature, you'll need an active Deep Research subscription.
How to get started:
Navigate to Insights > AI Recommendations in your Rep console
If you don't have Deep Research yet, you'll see an option to activate it
Deep Research includes a free trial, so you can explore AI Recommendations risk-free
Click "Activate Deep Research" to begin using AI Recommendations immediately
For complete details about Deep Research and all its features, see Deep Research: Unlock Conversation Intelligence for Your Business.
How AI Recommendations work
Your AI continuously analyzes customer conversations, focusing on:
- Conversations where customers express dissatisfaction
- Patterns in customer questions and feedback
- Common information gaps or friction points
- Areas where customers frequently need assistance
When patterns are identified across multiple conversations, your AI creates specific, actionable recommendations and prioritizes them based on frequency and impact.
Viewing your AI Recommendations
- Navigate to Insights > AI Recommendations in the Rep Console
- See a summary of analyzed conversations (typically from the past 30 days)
- View recommendations organized in two tabs:
- Needs Action: Current recommendations awaiting your review
- Resolved: Historical log of suggestions you've already addressed
Understanding Recommendation Cards
Each recommendation card includes:
- Title: A clear description of the suggested improvement
- Detailed Description: Specific actions you can take to address the issue
- Impact Indicator: Shows how many shoppers raised this concern
- View Button: Click to see the actual conversations that generated this recommendation
- Mark as Resolved: Button to indicate when you've addressed the recommendation
- Feedback: Option to rate if the recommendation was helpful
Taking Action on Recommendations
- Review each recommendation card to understand the suggested improvement
- Click View to examine the actual customer conversations that generated the recommendation
- Implement the suggested changes on your website or store
- Click Mark it as resolved once you've addressed the recommendation
- The recommendation will move to the "Resolved" tab for your reference
Refreshing Your Recommendations
To generate fresh recommendations based on recent conversations:
- Click the Generate Recommendations button at the top of the page
- Your AI will analyze recent conversations and update your recommendations
- Check the "Last updated" timestamp to see when recommendations were last refreshed
Benefits of Using AI Recommendations
- Save Time: Automatically discover patterns that would take hours to find manually
- Customer-Driven Improvements: Make changes based on actual customer feedback
- Prioritized Actions: Focus on issues affecting the most customers first
- Measurable Impact: Track resolved recommendations and their effect on customer satisfaction
- Continuous Improvement: Regularly implement small changes that add up to a better customer experience
Troubleshooting
No recommendations appear:
- Your AI needs sufficient conversation data to generate recommendations
- Try clicking "Generate Recommendations" to refresh the analysis
- Ensure your store has been active for at least a few days with customer conversations
Recommendations seem too general:
- Click "View" to see the specific conversations that generated each recommendation
- More specific recommendations will appear as your AI gathers more conversation data
Recommendations aren't relevant:
- Use the thumbs down button to provide feedback on unhelpful recommendations
- Your AI will learn from this feedback to improve future recommendations