---

name: commerce-analytics description: Use when analyzing sales performance, customer metrics, inventory health, or generating forecasts.

Commerce Analytics Skill

Domain knowledge for business intelligence, sales analytics, customer insights, and demand forecasting.

Analytics Concepts

Time Periods

Period	Description	Use Case
today	Current calendar day	Real-time monitoring
last7days	Rolling 7 days	Weekly trends
last30days	Rolling 30 days	Monthly performance
this_month	Current calendar month	Month-to-date
last_month	Previous calendar month	Month comparison
this_year	Current calendar year	Year-to-date
all_time	All historical data	Lifetime value

Key Metrics

Metric	Formula	Meaning
AOV	Revenue / Orders	Average Order Value
LTV	Sum(Revenue per Customer)	Customer Lifetime Value
Return Rate	Returns / Orders * 100	Percentage returned
Conversion	Orders / Visitors	Purchase rate

Sales Summary Structure

{
  totalRevenue: 45230.00,      // Sum of all order totals
  orderCount: 156,             // Number of orders
  averageOrderValue: 290.00,   // Revenue / Orders
  itemsSold: 423,              // Sum of quantities
  uniqueCustomers: 89          // Distinct customers
}

Interpreting Sales Data

• Revenue up, Orders down = Higher AOV (upselling working)
• Revenue down, Orders up = Lower AOV (discounting impact)
• New customers high = Marketing working
• Returning customers high = Retention strong

Product Performance

Top Products Output

{
  sku: "WIDGET-001",
  name: "Premium Widget",
  unitsSold: 78,
  revenue: 15600.00,
  orderCount: 45
}

Product Metrics to Watch

Metric	High Value	Low Value
Units Sold	Popular item	Consider discontinuing
Revenue	High-value product	May need pricing review
Order Count	Frequently bought	Bundle opportunity

Product Performance Ratios

Revenue per Unit = Revenue / Units Sold
Units per Order = Units Sold / Order Count

Customer Analytics

Customer Metrics Output

{
  totalCustomers: 1250,           // All customers
  newCustomers: 45,               // New in period
  returningCustomers: 234,        // Repeat buyers
  averageLifetimeValue: 450.00,   // Average total spend
  averageOrdersPerCustomer: 2.3   // Order frequency
}

Customer Segmentation

Segment	Definition	Action
New	First purchase	Welcome, onboarding
Active	Purchased in 90 days	Maintain engagement
At Risk	No purchase 90-180 days	Re-engagement campaign
Lapsed	No purchase 180+ days	Win-back campaign
VIP	Top 10% by LTV	Premium treatment

Top Customers Output

{
  customerId: "uuid",
  name: "John Doe",
  email: "john@example.com",
  orderCount: 15,
  totalSpent: 2340.00,
  averageOrderValue: 156.00
}

Inventory Health

Health Overview Output

{
  totalSkus: 150,           // All products
  inStockSkus: 120,         // Available > 0
  lowStockSkus: 18,         // Below reorder point
  outOfStockSkus: 12,       // Available = 0
  totalValue: 125000.00     // Sum of inventory value
}

Stock Status Levels

Level	Condition	Priority
In Stock	Available > reorder point	Normal
Low Stock	Available <= reorder point	Monitor
Critical	Available <= reorder point/2	Urgent
Out of Stock	Available = 0	Emergency

Low Stock Item Output

{
  sku: "WIDGET-001",
  name: "Premium Widget",
  onHand: 15,
  allocated: 5,
  available: 10,
  reorderPoint: 20,
  averageDailySales: 3.5,
  daysOfStock: 2.8    // available / averageDailySales
}

Demand Forecasting

Forecast Output

{
  sku: "WIDGET-001",
  name: "Premium Widget",
  averageDailyDemand: 3.5,     // Historical average
  forecastedDemand: 105,       // Next 30 days
  confidence: 0.7,             // 70% confidence
  currentStock: 45,
  daysUntilStockout: 12,       // stock / daily demand
  recommendedReorderQty: 105,  // 30 days supply
  trend: "Rising"              // Rising, Falling, Stable
}

Demand Trends

Trend	Pattern	Action
Rising	Demand increasing	Order more
Stable	Consistent demand	Maintain levels
Falling	Demand decreasing	Reduce orders

Reorder Recommendations

When daysUntilStockout < leadTime:

• Order immediately
• Quantity = forecastedDemand + safetyStock

Revenue Forecasting

Revenue Forecast Output

{
  period: "Period +1",          // Future period
  forecastedRevenue: 48000.00,  // Point estimate
  lowerBound: 40800.00,         // 80% confidence lower
  upperBound: 55200.00,         // 80% confidence upper
  confidenceLevel: 0.8,         // 80%
  basedOnPeriods: 12            // Historical periods used
}

Forecast Granularity

Granularity	Best For	Accuracy
Day	Short-term planning	Higher variance
Week	Operational planning	Moderate
Month	Strategic planning	More stable

Interpreting Confidence Intervals

Forecasted: $48,000
Lower (80%): $40,800
Upper (80%): $55,200

There's an 80% chance revenue will fall between
$40,800 and $55,200

Order Status Breakdown

Status Breakdown Output

{
  pending: 12,      // Awaiting confirmation
  confirmed: 8,     // Confirmed, not processing
  processing: 15,   // Being prepared
  shipped: 45,      // In transit
  delivered: 120,   // Completed
  cancelled: 5,     // Cancelled by customer/merchant
  refunded: 3       // Refunded
}

Operational Health Indicators

Ratio	Formula	Good Target
Completion Rate	Delivered / Total	> 95%
Cancellation Rate	Cancelled / Total	< 3%
Fulfillment Rate	Shipped / (Pending+Confirmed+Processing)	Improving

Return Metrics

Return Metrics Output

{
  totalReturns: 23,
  returnRatePercent: 4.5,
  totalRefunded: 3450.00
}

Return Rate Benchmarks

Rate	Assessment	Action
< 3%	Excellent	Maintain quality
3-5%	Average	Monitor reasons
5-10%	High	Investigate causes
> 10%	Critical	Quality review

Common Return Reasons

1. Defective - Quality control issue
2. Wrong item - Fulfillment error
3. Not as described - Listing accuracy
4. Changed mind - Normal behavior
5. Better price found - Competitive pricing
6. Damaged - Shipping issue

Analytics Workflows

Weekly Business Review

1. get_sales_summary(period: "last7days")
2. get_top_products(period: "last7days", limit: 5)
3. get_order_status_breakdown(period: "last7days")
4. get_return_metrics(period: "last7days")

Monthly Planning

1. get_sales_summary(period: "last_month")
2. get_customer_metrics(period: "last_month")
3. get_revenue_forecast(periodsAhead: 3, granularity: "month")
4. get_demand_forecast(daysAhead: 30)

Inventory Review

1. get_inventory_health()
2. get_low_stock_items(threshold: 20)
3. get_demand_forecast(skus: [critical_skus], daysAhead: 14)

Best Practices

1. Compare periods - Always show context vs prior period
2. Focus on trends - Single data points are less meaningful
3. Segment analysis - Break down by product, customer, region
4. Action-oriented - Every insight should suggest an action
5. Set benchmarks - Define what "good" looks like for your business
6. Regular cadence - Schedule analytics reviews weekly/monthly

Common Questions and Answers

"Is this a good month?"

Compare to:

• Same month last year (seasonality)
• Previous month (trend)
• Monthly average (benchmark)

"Which products should I reorder?"

Look for:

• daysUntilStockout < 14
• trend = "Rising"
• High revenue contribution

"Are customers coming back?"

Check:

• returningCustomers / (totalCustomers - newCustomers)
• Average orders per customer
• Customer lifetime value trend

"Is my pricing right?"

Analyze:

• AOV trends
• Return rate
• Revenue per unit sold