Insight Cards Guide
Discover how Insight Cards help you spot pricing opportunities and competitive threats automatically. Learn to read, filter, and act on AI-powered pricing intelligence.
On This Page
What are Insight Cards?
Insight Cards are AI-powered pricing alerts that automatically analyze your scan results and surface the information that matters most. After every scan, Nesika AI compares your prices against competitor data and generates actionable insights so you can respond quickly to market changes.
Each Insight Card tells you:
- What happened — a competitor undercut your price or an opportunity to increase your price exists
- How severe it is — from informational to critical
- How confident the system is in the finding
- What you can do about it — with actionable recommendations
Automatic Generation
How Insights are Generated
Insights are generated automatically through a multi-step pipeline that starts when a price scan completes. Here is an overview of the process:
Generation Pipeline
Scan Completes
All vendor searches finish and job results are collected.
Event Published
A "scan completed" event is published to the event bus.
Batches Queued
Jobs are divided into batches and queued for parallel processing.
Analysis
Each job is analyzed: prices compared, Weber's Law applied, consensus checked.
Cards Created
Insight Cards are created with severity, confidence, and recommendations.
Available to You
Cards appear on your Insights dashboard, ready to review and act on.
Eligibility Requirements
What happens during analysis?
For each product in your scan, the system performs these checks:
- Price comparison — Your price is compared to each competitor's price
- Significance test — Weber's Law determines if the gap is meaningful (not just noise)
- Consensus check — Are all competitors pointing in the same direction?
- Confidence scoring — How confident is the system based on history and match quality?
- Deduplication — Has this exact insight already been generated recently?
Types of Insights
Nesika AI currently generates two types of Insight Cards, with a third type coming soon. Each is designed to highlight a different pricing scenario:
🔻 Competitor Undercut
A competitor is selling the same (or closely matched) product at a significantly lower price than yours. These insights help you stay competitive and avoid losing sales to cheaper alternatives.
When is it generated? When all competitors with matched products are pricing below you, and the gap exceeds your configured threshold. The system uses 100% consensus — if even one competitor is priced above you, the undercut insight is not generated, ensuring only genuine competitive threats are flagged.
Example: "A competitor is selling Outdoor Garden Hose 30m at $45.99, which is 15% less than your price of $53.99"
📈 Price Increase Opportunity
Your price is well below the market average, suggesting room to increase your price without losing competitiveness. These insights can directly improve your margins.
When is it generated? When all competitors with matched products are pricing above you, and the gap is significant. Like undercut detection, this requires 100% consensus — every competitor must be higher for the opportunity to be flagged, giving you confidence the market supports a price increase.
Example: "Your product Premium Pool Cleaner is priced at $29.99 while competitors average $42.50 — a potential 41% increase opportunity"
🏷️ Promo Detected
A promotional or temporary price has been detected on a competitor's product (or your own). This will help you understand whether a price change is permanent or part of a limited-time campaign.
When will it be generated? When the system detects a price that is significantly lower than the product's usual price based on historical data, suggesting a promotional or sale price rather than a permanent change.
Example: "A competitor has a promotional price of $19.95 on Silicone Sealant 300ml, down from the usual $27.50"
In Development
Multiple Insights per Product
Understanding Severity Levels
Every Insight Card is assigned a severity level based on the size of the price gap. This helps you prioritize which insights to act on first.
🔴 Critical — 20%+ price gap
A major price difference that requires immediate attention. A competitor may be significantly undercutting you, or you may be leaving substantial margin on the table.
🟠 High — 10–20% price gap
A notable price difference that should be reviewed soon. These insights often represent meaningful competitive or margin opportunities.
🟡 Medium — 5–10% price gap
A moderate difference worth monitoring. These may not require immediate action, but should be checked during your regular pricing review.
🔵 Low — Under 5% price gap
A small price difference provided for your awareness. These are informational and typically don't require action unless part of a broader trend.
Focus on What Matters
Confidence Scores
Each Insight Card includes a confidence score (displayed as a percentage) that indicates how reliable the insight is. A higher score means Nesika AI is more certain the finding is accurate and actionable.
What affects confidence?
- Historical data — more past scans provide a stronger baseline for comparison
- Competitor agreement — when multiple competitors show similar pricing patterns, confidence increases
- Product match quality — well-matched products (validated or fine-tuned) produce higher confidence insights
- Price stability — stable pricing trends are more reliable than volatile, one-time price changes
How confidence grows over time
The confidence score starts at a base value (default 50%) when a pattern is first observed. Each subsequent scan that confirms the same pattern increases the score:
| Confirming Scans | Approximate Confidence | Level |
|---|---|---|
| 1 scan (first time) | 50% | Medium |
| 2 scans | ~58% | Medium |
| 5 scans | ~77% | Medium |
| 8+ scans | 80%+ | High |
Confidence Improves Over Time
Low Confidence Insights
Weber's Law — How Price Gaps are Evaluated
Nesika AI uses a principle from psychophysics called Weber's Law to determine whether a price difference is meaningful to buyers. The core idea is simple: a $1 price difference matters much more for a $5 item than for a $500 item.
📐 The Two-Tier Threshold
Instead of using a single percentage for all products, Nesika AI splits products into two categories and applies different rules:
💰 Low-Ticket Items
Products priced below $20 (default)
Uses an absolute dollar difference. A price gap must be at least $1.00 (default) to be considered significant.
💎 High-Ticket Items
Products priced at $20 or above (default)
Uses a percentage difference. A price gap must be at least 10% (default) to be considered significant.
🧠 Why does this matter?
Without Weber's Law, a flat 10% threshold would ignore a meaningful $1.50 difference on a $5 item (that's 30% — huge for a buyer!) while flagging a trivial $5 difference on a $500 item (just 1% — nobody notices). The two-tier approach ensures:
- Cheap items are evaluated by dollar amount (because buyers feel the actual price)
- Expensive items are evaluated by percentage (because buyers think in proportions)
- You only get notified about price gaps that actually influence purchasing decisions
Customizable Thresholds
Working with Insights
Insight Cards are designed to be actionable. Here are the actions you can take on each card:
👁️ View Details
Click the Insight Card
Review the recommendation
Auto-Mark as Viewed
🚫 Dismiss
If an insight isn't relevant or you don't want to act on it, you can dismiss it to remove it from your active view.
Click the dismiss button
Optionally provide a reason
⏰ Snooze
Temporarily hide an insight if you want to revisit it later. The card will reappear after the snooze period ends.
Click the snooze button
Select a duration
✅ Take Action
When you've acted on an insight (e.g., adjusted your price), mark it as actioned to keep your dashboard clean and track your response rate.
Click "Take Action"
Record what you did
Confirm
Estimated Revenue Impact
Each Insight Card includes an estimated monthly revenue impact that helps you understand the financial significance of a pricing gap. Instead of just showing you the price difference, Nesika AI calculates what that gap could mean for your monthly revenue.
Why Revenue Impact?
📊 How It's Calculated
The estimated monthly revenue impact is calculated using this formula:
Monthly Revenue Impact = Price Gap per Unit × Estimated Monthly Unit SalesWhere:
- Price Gap per Unit — the absolute dollar difference between your price and the primary competitor's price
- Estimated Monthly Unit Sales — a volume estimate based on the product's price tier (see table below)
📦 Estimated Monthly Unit Sales by Price Tier
Since actual unit sales data is not available, Nesika AI uses price as a proxy for sales volume. Lower-priced products typically sell in higher volumes.
| Product Price | Est. Monthly Units (Low) | Est. Monthly Units (High) | Typical Products |
|---|---|---|---|
| ≤ $10 | 1,000 | 5,000 | Consumables, small accessories |
| $10.01 – $25 | 500 | 2,000 | Everyday essentials, mid-range items |
| $25.01 – $50 | 200 | 1,000 | Mid-price consumer goods |
| $50.01 – $100 | 100 | 500 | Higher-value household items |
| $100.01 – $300 | 50 | 200 | Premium goods, electronics |
| > $300 | 10 | 50 | High-ticket items, specialty products |
💡 Example
Suppose you sell a product at $9.99 and a competitor prices it at $8.99:
- Price gap: $1.00 per unit
- Price tier: ≤ $10 → estimated 1,000–5,000 units/month
- Estimated impact: $1,000–$5,000 per month in revenue at risk
The Insight Card would display: "Estimated monthly impact: $1,000.00–$5,000.00 (revenue risk)"
Based on est. 1,000–5,000 units/month at AUD 1.00 price gap per unit
🏷️ Impact Types
Shown for Competitor Undercut insights — the competitor's lower price means you risk losing this amount of monthly revenue to them.
Shown for Price Increase Opportunity insights — your price is significantly below competitors, so you could potentially capture this additional monthly revenue by raising your price.
Estimates, Not Exact Numbers
Insight Lifecycle
Each insight follows a lifecycle from creation to resolution. Understanding the lifecycle helps you manage your insights effectively.
Status Flow
Automatic Invalidation
Filtering & Searching
When you have many insights, use the built-in filtering and search tools to find exactly what you need:
🔎 Search
Use the search bar to find insights by product name or SKU. Results update instantly as you type.
📋 Filter by Type
Show only specific insight types — Competitor Undercut or Price Increase Opportunity. Additional types like Promo Detected will be available as they are released.
⚡ Filter by Severity
Focus on what matters most by filtering to Critical, High, Medium, or Low severity insights.
📌 Filter by Status
View insights by their current status — New, Viewed, Dismissed, Snoozed, or Actioned.
📅 Date Range
Filter insights by the date they were generated to focus on recent changes or review historical data.
📥 Export to CSV
Download your filtered insights as a CSV file for offline analysis, reporting, or sharing with your team.
Combine Filters
Configuration & Settings
Insight Cards are generated automatically after each scan completes. You can customize how insights are generated to match your business needs through the Settings page.
All settings have sensible defaults that work well for most businesses. You only need to change them if you want to fine-tune the insight generation to better match your specific pricing strategy and product catalog.
Settings categories
- Consensus & Confidence — How much agreement and certainty is required before generating an insight
- Pricing Thresholds (Weber's Law) — What size price gap is considered significant
- Defense & Offense Gaps — Fine-tuned thresholds for undercut vs. opportunity detection
- Lifecycle — How long before duplicate insights can be re-generated
Where to Configure
Changes Apply to Future Scans
Settings Reference (All Defaults)
Below is a complete reference of all configurable insight generation settings, their default values, and what they control. These can be overridden per workspace in Settings → Insights Configuration.
🎯 Consensus & Confidence
Controls how much agreement and certainty is required before generating insights.
| Setting | Default | Description |
|---|---|---|
| Consensus Threshold | 1.0 (100%) | All competitors must agree on the pricing direction. Set to 0.8 to allow insights when 80% of competitors agree. |
| Min Confidence Threshold | 0.9 (90%) | Minimum confidence score required for an insight to be generated. Lower this to see more insights with less certainty. |
| Min Match Confidence | 0.7 (70%) | Minimum product matching confidence required. Products with lower match quality are excluded from insight generation. |
📐 Pricing Thresholds (Weber's Law)
Defines what constitutes a "significant" price gap using the two-tier Weber's Law approach.
| Setting | Default | Description |
|---|---|---|
| Price Sensitivity Threshold | $20.00 | The price point that separates "low-ticket" from "high-ticket" items. Items below this use dollar thresholds; items at or above use percentage thresholds. |
| Low-Ticket Dollar Threshold | $1.00 | Minimum absolute dollar difference required for low-ticket items (priced below the sensitivity threshold). |
| High-Ticket Percent Threshold | 0.10 (10%) | Minimum percentage difference required for high-ticket items (priced at or above the sensitivity threshold). |
| Undercut Threshold | 10% | Percentage threshold specifically for detecting when a competitor has undercut your price. |
🛡️ Defense & Offense Gaps
Fine-grained control over the minimum price gap required for defense (competitors undercutting you) and offense (price increase opportunities) insights.
| Setting | Default | Description |
|---|---|---|
| Defense Gap — Low-Ticket | 0.05 (5%) | Minimum gap to trigger a "competitor undercut" insight for low-ticket items. |
| Defense Gap — High-Ticket | 0.005 (0.5%) | Minimum gap to trigger a "competitor undercut" insight for high-ticket items. |
| Offense Gap — Low-Ticket | 0.10 (10%) | Minimum gap to trigger a "price increase opportunity" insight for low-ticket items. |
| Offense Gap — High-Ticket | 0.02 (2%) | Minimum gap to trigger a "price increase opportunity" insight for high-ticket items. |
📊 Confidence Scoring
Controls how confidence scores are calculated and grow over time.
| Setting | Default | Description |
|---|---|---|
| Initial Confidence Score | 0.5 (50%) | Starting confidence when a pricing pattern is first observed. Each subsequent confirming scan increases this value. |
| Confidence Increment | 0.15 (15%) | Percentage of current score added per confirming scan. The actual increase is the smaller of this percentage or the max increment. |
| Max Confidence Increment | 0.1 (+10%) | Maximum absolute confidence increase per step. Caps the growth rate to ensure gradual, stable progression. |
| High Match Quality Threshold | 0.8 (80%) | Confidence above this value earns the "high quality match" flag on the insight card. |
🔄 Lifecycle & Deduplication
Controls insight validity periods and deduplication behavior.
| Setting | Default | Description |
|---|---|---|
| Deduplication TTL | 72 hours | How long before the same insight can be generated again. Prevents duplicate insights from back-to-back scans. |
| Valid Until | 30 days | Default validity period for an insight. After this, the insight automatically expires if not acted upon. |
| Recommendation Expiry | 3 days | How long a specific recommendation stays valid before it should be re-evaluated with fresh data. |