List view
Understanding Nobi
Understanding Nobi
Getting Started
Getting Started
Implementing Nobi On Your Site
Implementing Nobi On Your Site
Ecommerce Merchandising
Ecommerce Merchandising
Reporting & Analytics
Reporting & Analytics
Beta Products
Beta Products
Developers Guide
Developers Guide
References
References
Insights
Understand what customers search for, how their needs evolve, and which topics drive engagement. Use insights to identify merchandising opportunities and stay ahead of customer demand.
Overview
The Insights page helps you understand customer behavior through:
- Query trends - What customers search for and how volume changes over time
- Topic analysis - Common themes and patterns grouped into categories
- Search attributes - Product characteristics customers care about (color, size, material, price)
- Query intents - What customers want to accomplish (find products, get recommendations, ask questions)
Use trending queries to:
- Spot emerging customer interests early
- Adjust merchandising for growing demand
- Phase out products for declining searches
Change in Conversion Rate
Track which queries convert better or worse over time. Shows how effective search results are at driving purchases for specific terms.
High-volume queries with low or declining conversion rates are prime candidates for merchandising optimization (boosting, slotting, or improving product selection).
Topics
Related queries grouped into common themes. Topics help you see the bigger picture beyond individual search terms.
Example topics:
- Waterproof gear - All searches related to waterproof products (boots, jackets, bags)
- Gift ideas - Searches about gifts for different occasions or recipients
- Size and fit - Questions about sizing, measurements, and fit
- Sale items - Searches for deals, discounts, and promotions
Each topic shows:
- Total query count
- Trending indicator
- Sample queries within the topic
Search Attributes
Product characteristics customers search for most frequently. Shows what attributes matter to your customers.
Common attributes:
- Color - "black dress", "red shoes"
- Size - "size 10", "large", "oversized"
- Material - "leather", "cotton", "waterproof"
- Price - "under $100", "affordable", "budget"
- Style - "casual", "formal", "minimalist"
- Fit - "slim fit", "wide width", "true to size"
Query Intent
What customers want to accomplish when they use Nobi. Categorized into common intent types.
Intent categories:
Search for products - Looking for specific items or browsing categories
Get recommendations - Asking for suggestions or personalized advice
Ask questions - Inquiring about products, policies, or information
Browse collections - Exploring product groups or categories
Compare products - Evaluating differences between items
Filtering and Time Ranges
Time Range Selection
View insights for different periods:
- Last 7 days
- Last 30 days
- Last 90 days
- Custom date ranges
Longer time ranges help identify sustained trends rather than short-term fluctuations.
Sorting and Filtering
Query sorting:
- By volume (most searched)
- By trending status (increasing/decreasing)
- By conversion rate (best/worst performing)
Use sorting to focus on different priorities: immediate opportunities (trending), sustained volume (top queries), or performance issues (low conversion).
Common Questions
How often do topics update?
Topics are recalculated regularly as new queries come in. Check back weekly to see how topics evolve.
Why are some queries marked as spam?
Spam queries are test searches, bot activity, or gibberish. You can filter these out in Performance Metrics.
Can I create custom topics?
Topics are automatically generated by Nobi's AI. Custom topic creation is not currently available.
What's the difference between attributes and topics?
Attributes are specific product characteristics (color, size). Topics are broader themes that group related queries together.
How far back does historical data go?
Insights show data from when Nobi was installed on your store. Longer time ranges provide more context for trends.
ℹ️ LLM-based Categorization
The insights feature uses LLMs to automatically categorize queries. You might see inaccuracies or inconsistencies.