afv-library/skills/tableau-next-semantic-model-generate/references/kpi-formulas.md

KPI Formulas & Aggregation Patterns

Business-logic reference for building calculated fields (_clc) and metrics (_mtc) on a Tableau Next semantic model. Provides lookup tables for custom aggregations, composite KPI formulas, and color semantics observed across finance, banking, sales, HR, retail, marketing, procurement, supply-chain, travel, and executive dashboards.

---

Table of Contents

• Time-Based Aggregation Patterns
• Composite KPI Patterns
• Color Coding Guide
• Palette Usage by Metric Type
• Metric Clarifications
• Industry-Specific KPI Patterns

---

Time-Based Aggregation Patterns

Native Tableau formulas for time-based business calculations. Replace [Date], [Measure], [Dimension] placeholders with actual field names from your semantic model.

Snapshot & Position Aggregations

Pattern	Tableau Formula	Business Use Case
Sum of (last period only)	SUM(IF [Date] = { INCLUDE : MAX([Date]) } THEN [Measure] ELSE NULL END)	Balance sheets, headcount snapshots, portfolio positions, inventory levels
Sum of (first period only)	SUM(IF [Date] = { INCLUDE : MIN([Date]) } THEN [Measure] ELSE NULL END)	Initial investment positions, opening balances
Average of (last period only)	AVG(IF [Date] = { INCLUDE : MAX([Date]) } THEN [Measure] ELSE NULL END)	Staff demographics at period end, current average values
Median of (last period only)	MEDIAN(IF [Date] = { INCLUDE : MAX([Date]) } THEN [Measure] ELSE NULL END)	Central tendency for period-end snapshots
Count Distinct of (last period only)	COUNTD(IF [Date] = { INCLUDE : MAX([Date]) } THEN [Dimension] ELSE NULL END)	Ending counts of unique items (Active Customers EOP, Open Positions)

When to use: For metrics that represent a state at a specific point in time rather than cumulative totals. Q1 value = last known value in Q1.

Time-Scaled Aggregations

Pattern	Tableau Formula	Business Use Case
Sum of (annualized)	SUM([Measure]) * (12 / COUNTD(DATEPART('month', [Date])))	Revenue projections; 1 month × 12, 2 months × 6, 3 months × 4
Sum of (monthly average)	SUM([Measure]) / COUNTD(DATEPART('year', [Date]) + DATEPART('month', [Date]))	Avg Monthly Inventory, Avg Monthly Assets

Annualization logic: Extrapolate partial-period values to full year.

Period Range Aggregations

Pattern	Tableau Formula	Business Use Case
Sum of (from to)	SUM(IF [Date] >= [FromDate] AND [Date] <= [ToDate] THEN [Measure] END)	Total Sales During Contract, Revenue in Date Range
Average of (from to)	AVG(IF [Date] >= [FromDate] AND [Date] <= [ToDate] THEN [Measure] END)	Avg Deal Size During Q1-Q3
Min/Max/Median (from to)	Same pattern with MIN, MAX, MEDIAN	Lowest inventory, peak revenue, median ticket
Count Distinct (from to)	COUNTD(IF [Date] >= [FromDate] AND [Date] <= [ToDate] THEN [Dimension] END)	Unique Customers Between Launch-EOY

When to use: For metrics that need to aggregate data only within a specific date range.

Lifecycle & Cohort Aggregations

Pattern	Tableau Formula	Business Use Case
Count Distinct of New	COUNTD(IF [Date] = { FIXED [Dimension] : MIN([Date]) } THEN [Dimension] END)	New Customers, New Products, New Hires
Count Distinct of Ending	COUNTD(IF [Date] = { FIXED [Dimension] : MAX([Date]) } THEN [Dimension] END)	Lost Customers, Discontinued Products
Sum of (for New)	SUM(IF [Date] = { FIXED [Dimension] : MIN([Date]) } THEN [Measure] END)	Revenue from New Customers only
Avg Lifetime of	DATEDIFF('day', { FIXED [Dimension] : MIN([Date]) }, { INCLUDE [Dimension] : MAX([Date]) }) / 365	Customer tenure (years), Product lifetime, Employee tenure

When to use: For cohort analysis, churn tracking, and new/lost item identification.

Time Interval Calculations

Pattern	Tableau Formula	Business Use Case
Avg Years between	DATEDIFF('day', [StartDate], [EndDate]) / 365	Avg Years to Maturity, Avg Investment Horizon
Avg Months between	DATEDIFF('day', [StartDate], [EndDate]) / 365 * 12	Avg Months to Close, Loan Duration
Avg Days between	DATEDIFF('minute', [StartDate], [EndDate]) / (24 * 60)	Days to Close, Days Between Booking-Travel
Avg Hours between	DATEDIFF('minute', [StartDate], [EndDate]) / 60	Avg Response Time
Avg Minutes between	DATEDIFF('minute', [StartDate], [EndDate])	Avg Call Duration
Avg Seconds between	DATEDIFF('second', [StartDate], [EndDate])	Avg Page Load Time

---

Composite KPI Patterns

Real-world KPI formulas from production dashboards (finance, banking, sales).

Financial Statement Patterns

Income Statement (Profit & Loss)

Pattern	Formula	Example
Gross Profit	Revenues - Cost of Revenues	KPI1 - KPI2 where KPI1=Total Revenues, KPI2=COGS
Gross Margin %	Gross Profit / Revenues	(Revenues - COGS) / Revenues
Operating Income	Gross Profit - (OpEx + Depreciation)	KPI2 - (KPI4 + KPI12)
Operating Income %	Operating Income / Revenues	Operating Income / Total Revenues
Net Income	Operating Income - (Interests + Taxes)	KPI13 - KPI15

Tableau Implementation:

<kpi-simple name='KPI1' aggregation='SUM' canonical-attribute='IF STARTSWITH(UPPER(Dim2),"REVENUES") THEN Value END' />
<kpi-simple name='KPI11' aggregation='SUM' canonical-attribute='IF STARTSWITH(UPPER(Dim2),"COST OF REVENUES") THEN Value END' />
<kpi-composite name='KPI2' caption='Gross Profit' kpi1-name='KPI1' calculation='-' kpi2-name='KPI11' />

Balance Sheet

Pattern	Formula	Example
Working Capital	Current Assets - Current Liabilities	KPI17 - KPI18 using snapshot aggregation
Total Assets	Current Assets + Non-Current Assets	Sum all asset categories with snapshot aggregation
Total Equity	Total Assets - Total Liabilities	KPI21 - KPI22
Current Ratio	Current Assets / Current Liabilities	KPI17 / KPI18
Quick Ratio	(Cash + AR) / Current Liabilities	KPI24 / KPI18
Debt to Equity	Total Liabilities / Total Equity	KPI22 / KPI7

Critical: Use snapshot aggregation (SUM(IF [Date] = MAX([Date]) THEN ... END)) for balance sheet accounts.

Banking & Wealth Management Patterns

Assets under Management (AuM)

Pattern	Formula	Business Meaning
Net New Money (NNM)	Inflows - Outflows	zn(Inflows) + zn(Outflows) (Outflows are negative)
AuM Balance	Last period AuM (snapshot)	Use snapshot aggregation: SUM(IF [Date] = { INCLUDE : MAX([Date]) } THEN [AuM] END)
AuM per Client	Total AuM / Nb of Clients	KPI3 / KPI6
Return on Assets %	Annualized Income / Avg Monthly AuM	SUM(Income)*12/COUNTD(Month) / (SUM(AuM)/COUNTD(Month))
New Clients	Count new occurrences	COUNTD(IF [Date] = { FIXED [Client] : MIN([Date]) } THEN [Client] END)
Lost Clients	Count ending occurrences	COUNTD(IF [Date] = { FIXED [Client] : MAX([Date]) } THEN [Client] END)
Client Asset & Liabilities (CAL)	AuM + Lending Stock	KPI3 + KPI17
Net New Lending (NNL)	Lending Inflows + Lending Outflows	KPI21 + KPI23

Lending & Loans

Pattern	Formula	Business Meaning
Total Revenue	Interests + Fees	KPI2 + KPI4 (monthly totals)
Revenue Rate %	(Interests + Fees) / Loan Amortization	KPI11 / KPI3
Delinquent Loans Rate %	Delinquent Count / Total Active Loans	KPI14 / KPI5
Default Rate %	Defaulted Count / Total Active Loans	KPI15 / KPI5
Client Liabilities (EOP)	Outstanding balance at period end	SUM(IF [Date] = { INCLUDE : MAX([Date]) } THEN [ClientLiabilities] END)

Sales & Travel Patterns

Pattern	Formula	Example
Average Spend per Customer	Total Amount / Count Distinct Customers	KPI1 / KPI7
Share of Metric	AVG(IF condition THEN 1 ELSE 0 END)	AVG(if upper(Dim11)="Y" then 1 else 0 end) for "% Online"
Over-pricing %	(Actual - Lowest) / Lowest	(TicketAmount - LowestFare) / LowestFare
Days Between Events	DATEDIFF('minute', [Date1], [Date2]) / (24 * 60)	Days between booking and travel
Tickets per Traveler	Total Tickets / Distinct Travelers	KPI5 / KPI7

Supply Chain Patterns

Pattern	Formula	Business Meaning
Product Demand Index	Integer scale 1-5 (Low to High)	Direct measure, no calculation
Logistics Reliability	Integer scale 1-5	Direct measure, use MEDIAN or AVG
Inflation Impact	Integer scale 1-5	Direct measure for risk scoring

---

Usage Notes

Aggregation Selection Criteria

1. Snapshot metrics (balance sheet, headcount, inventory): Use SUM(IF [Date] = MAX([Date]) THEN ... END)
2. Flow metrics (revenue, expenses, transactions): Use standard SUM, AVG, COUNT
3. New/lost tracking: Use COUNTD(IF [Date] = { FIXED [Dim] : MIN/MAX([Date]) } THEN [Dim] END)
4. Time-between calculations: Use DATEDIFF('day', [Start], [End]) with appropriate unit
5. Annualized projections: Use SUM([Measure]) * (12 / COUNTD(Month))
6. Period-constrained aggregations: Use SUM(IF [Date] >= [From] AND [Date] <= [To] THEN ... END)

---

Cross-References

• See metric-design.md for Tableau Next semantic metric patterns
• See field-types.md for calculated measurement vs dimension guidance
• See tableau-functions.md for Tableau expression functions

---

Metric Clarifications

Before implementing a KPI, determine which variant matches the business question. Many pattern names suggest lifetime or strategic metrics, but the formula may be period-specific.

Period vs Lifetime Metrics

Type	When to Use	Formula Pattern	Example
Period	"Revenue per customer this month"	SUM(Measure) / COUNTD(Dimension)	Total Sales / Active Customers
Lifetime	"Total value of a customer over their entire relationship"	{ FIXED [Dimension]: SUM(Measure) }	{ FIXED [Customer]: SUM([Revenue]) }

Rule: If the metric name includes "Lifetime," "LTV," "CLV," or "Total Value," use an LOD expression to aggregate at the entity level first, then average or sum as needed.

Simple vs Fully-Loaded Costs

Type	When to Use	Example
Simple	Quick operational view	CAC = Marketing Spend / New Customers
Fully-Loaded	Strategic investment decisions	CAC = (Marketing + Sales salaries + Overhead + Tools) / New Customers

Rule: Document which cost components are included. "CAC" without qualification typically means marketing-only; "fully-loaded CAC" includes sales and overhead.

Snapshot vs Flow Metrics

Type	When to Use	Aggregation	Example
Snapshot	State at a point in time	SUM(IF [Date] = MAX([Date]) THEN ... END)	MRR, Headcount (EOP), Inventory
Flow	Activity over a period	SUM, COUNT	Revenue, Orders, New Customers

Rule: Use snapshot aggregation for balance-sheet style metrics; use SUM/COUNT for income-statement style metrics.

Attribution Models

For marketing metrics (ROAS, CPA, Conversion Rate), clarify:

• Total: All revenue in period ÷ ad spend (simple, may over-attribute)
• Attributed: Revenue/conversions filtered to ad-sourced customers only (requires attribution dimension)

---

Industry-Specific KPI Patterns

Business formulas extracted from 61 production dashboard templates across multiple industries.

Sales & E-Commerce

Customer Lifetime Value (CLV) Patterns

Real CLV answers: "How much total revenue will this customer generate over their entire relationship?" Use these patterns when the user asks for CLV, LTV, or Customer Lifetime Value.

Pattern	Formula	Business Use Case
CLV per Customer	{ FIXED [Customer]: SUM([Revenue]) }	Total revenue per customer over entire history
Average CLV	AVG([CLV per Customer])	Portfolio average lifetime value
Predictive CLV	AOV × Purchase Frequency × Customer Lifespan	Forward-looking estimate; use DATEDIFF('day', { FIXED [Customer] : MIN([Date]) }, { INCLUDE [Customer] : MAX([Date]) }) / 365 for lifespan
Cohort CLV	SUM([CLV per Customer]) where Customer in cohort	Total value of a cohort

When to use each: CLV per Customer is the base calc field; Average CLV for KPI cards; Predictive CLV when you need forward-looking estimates without full history.

Period Metrics (not lifetime):

Pattern	Formula	Business Use Case
Sales per Customer	Total Sales / Active Customers	KPI1 / KPI2 - Average customer spend in period; for lifetime value, see CLV patterns above
New Customers	COUNTD(IF [Date] = { FIXED [Customer] : MIN([Date]) } THEN [Customer] END)	Count first-time customers in period
Sales Margin %	Total Margin / Total Sales	KPI4 / KPI1 - Profitability percentage
Sales Costs %	(Sales - Margin) / Sales	KPI9 / KPI1 - Cost ratio
Average Selling Price	Total Sales / Total Quantity	KPI1 / KPI5 - Price per unit
Gross Discount %	(Volume at List Price - Sales Incl VAT) / Volume at List Price	Discount effectiveness
Net Discount %	(Volume at List Price - Sales Excl VAT) / Volume at List Price	After-tax discount
VAT Amount	Sales Incl VAT - Sales Excl VAT	Tax calculation

E-Commerce Specific:

Pattern	Formula	Business Use Case
Customer Acquisition Cost (CAC)	Marketing Costs / New Customers	SUM(IF [Date] = { FIXED [Customer] : MIN([Date]) } THEN [MarketingCosts] END) / COUNTD(IF [Date] = { FIXED [Customer] : MIN([Date]) } THEN [Customer] END)
CLV/CAC Ratio	Average CLV / CAC	Profitability of acquisition; CLV = { FIXED [Customer]: SUM([Revenue]) } averaged (see CLV patterns above)
Conversion Rate	Orders / Visits	Visitor-to-customer conversion
Average Order Value (AOV)	Total Revenue / Number of Orders	Revenue per transaction
Cart Abandonment %	(Carts Created - Orders) / Carts Created	Lost sales opportunity

Recurring Revenue (SaaS/Subscription)

Pattern	Formula	Business Use Case
MRR (Monthly Recurring Revenue)	SUM(IF [Date] = { INCLUDE : MAX([Date]) } THEN [MonthlyRecurringRevenue] END)	Snapshot at period end
ARR (Annual Recurring Revenue)	SUM(IF [Date] = { INCLUDE : MAX([Date]) } THEN 12 * [MonthlyRecurringRevenue] END)	Annualized recurring revenue
Total Revenue	SUM(ZN(OneTimeFee) + ZN(MonthlyRecurringRevenue))	Recurring + one-time fees
Recurring Revenue %	Period Recurring Revenue / Total Revenue	KPI5 / KPI1 - Subscription mix
Contract Duration (months)	DATEDIFF('day', [StartDate], [EndDate]) / 365 * 12	Average contract length
Revenue per Account	Total Revenue / Active Accounts	KPI1 / KPI2 - Period metric; for account lifetime value, use { FIXED [Account]: SUM([Revenue]) }
New Active Accounts	COUNTD(IF [Date] = { FIXED [Account] : MIN([Date]) } THEN [Account] END)	First-time contracts

Key Insight: Use snapshot aggregation for MRR/ARR (last period value), not SUM.

Retail & Inventory

Pattern	Formula	Business Use Case
Sales per Store	Total Sales / Active Stores	KPI1 / KPI7 - Store productivity
Distinct Products Sold	COUNTD(Product)	Catalog depth
Average Inventory Value	SUM(On-Hand Amount) / COUNTD(Date)	Average inventory holding
Total On-Hand Inventory	SUM(IF [Date] = { INCLUDE : MAX([Date]) } THEN [OnHandAmount] END)	Snapshot inventory value
Inventory Turnover	(Days in Year × Total Sales) / Total On-Hand Inventory	(DAYSinYEAR * KPI1) / KPI101 - Turns per year; higher = faster sell-through
Days Sales of Inventory (DSI)	Total On-Hand Inventory / (Total Sales / 365)	Days to sell current inventory; use Inventory / Daily Sales for correct time dimension
Average Selling Price	Total Sales / Total Sales Units	KPI1 / KPI50 - Price per unit
Out-of-Stock Positions	SUM(IF On-Hand Units = 0 THEN 1 END)	Count of stockouts
Out-of-Stocks %	Out-of-Stock Positions / Inventory Positions	KPI13 / KPI14 - Stockout rate

Procurement & Supply Chain

Pattern	Formula	Business Use Case
Spend per Vendor	Total Spend / Active Vendors	KPI1 / KPI2 - Vendor concentration
New Vendor #	COUNTD(IF [Date] = { FIXED [Vendor] : MIN([Date]) } THEN [Vendor] END)	First-time suppliers
Average Buying Price	Purchasing Volume / Total Purchase Quantity	KPI1 / KPI5 - Unit cost
Spend Under Management (SuM)	SUM(IF [Date] >= [StartDate] AND [Date] <= [EndDate] THEN [Amount] END)	Contracted spend in period
Ongoing Contracts	COUNTD(IF [Date] >= [StartDate] AND [Date] <= [EndDate] THEN [Contract] END)	Active contracts during period
Active Suppliers	COUNTD(IF [Date] >= [StartDate] AND [Date] <= [EndDate] THEN [Supplier] END)	Suppliers with active contracts
Avg Contract Size	Spend Under Management / Ongoing Contracts	KPI1 / KPI2 - Contract value

Digital Marketing & Advertising

Google Analytics:

Pattern	Formula	Business Use Case
Bounce Rate %	Bounced Sessions / Total Sessions	Single-page visits
Pages per Session	Total Pageviews / Total Sessions	Engagement depth
Avg Session Duration	DATEDIFF('second', [SessionStart], [SessionEnd])	Time on site
Conversion Rate %	Goal Completions / Total Sessions	Success rate
Cost per Click (CPC)	Total Ad Spend / Total Clicks	Click efficiency
Click-Through Rate (CTR) %	Clicks / Impressions	Ad engagement

Google/Facebook/Twitter Ads:

Pattern	Formula	Business Use Case
Cost per Acquisition (CPA)	Total Spend / Conversions	KPI1 / KPI3 - Conversion cost
Return on Ad Spend (ROAS)	Revenue / Ad Spend	KPI4 / KPI1 - Total period revenue ÷ ad spend; for attributed ROAS, filter revenue to ad-sourced conversions
CPM (Cost per 1000 Impressions)	(Total Spend / Impressions) × 1000	Reach cost
Conversion Rate %	Conversions / Clicks	Click-to-conversion rate
Quality Score	(CTR × Relevance × Landing Page Experience)	Ad platform ranking
Cost per Lead (CPL)	Total Spend / Leads Generated	Lead acquisition cost

Email Marketing:

Pattern	Formula	Business Use Case
Open Rate %	Emails Opened / Emails Delivered	Email engagement
Click Rate %	Clicks / Emails Delivered	Link engagement
Click-to-Open Rate (CTOR) %	Clicks / Opens	Engaged reader action
Unsubscribe Rate %	Unsubscribes / Emails Delivered	List health
Bounce Rate %	Bounced Emails / Total Emails Sent	Deliverability
Conversion Rate %	Conversions / Emails Delivered	Campaign effectiveness

Sales Performance & Quotas

Pattern	Formula	Business Use Case
Sales vs Quota	Total Sales - Total Quota	KPI1 - KPI2 - Variance amount
Quota Attainment %	Total Sales / Total Quota	KPI1 / KPI2 - Achievement rate
Market Share %	Our Sales / Total Market Size	KPI2 / KPI1 - Competitive position

RFM Analysis (Customer Segmentation)

Recency, Frequency, Monetary analysis - Template focuses on segmentation visualizations rather than formulas. Primary KPIs:

• Total Sales: SUM(SalesAmount)
• Active Customers #: COUNTD(Customer)

Segmentation happens in visualization layer using FIXED LOD calculations on last purchase date, purchase count, and average spend.

---

Advanced Aggregation Use Cases by Industry

Finance: Balance Sheet Accounts

• Use: Snapshot aggregation for all asset/liability accounts
• Why: Balance sheets are snapshots at period end, not cumulative totals
• Formula: SUM(IF [Date] = { INCLUDE : MAX([Date]) } THEN [Measure] END)

Banking: Portfolio Positions

• Use: Snapshot aggregation for AuM, Lending Stock
• Why: Account balances are point-in-time values
• Pattern: COUNTD(IF [Date] = { FIXED [Client] : MIN/MAX([Date]) } THEN [Client] END) for client churn

Sales: New/Lost Customer Tracking

• Use: COUNTD(IF [Date] = { FIXED [Customer] : MIN([Date]) } THEN [Customer] END) for new customers
• Why: First transaction ever per customer
• Pattern: New Customers = first transaction in period

Retail: Inventory Management

• Use: Snapshot aggregation for on-hand inventory, SUM for sales
• Why: Inventory is snapshot; sales are cumulative
• Pattern: Inventory Turnover = (Annual Sales) / (Average Inventory)

Recurring Revenue: MRR/ARR

• Use: Snapshot aggregation for MRR (not SUM!)
• Why: MRR is the recurring revenue at period end, not summed across days
• Formula: SUM(IF [Date] = { INCLUDE : MAX([Date]) } THEN [MonthlyRecurringRevenue] END)

Procurement: Contract Duration

• Use: SUM(IF [Date] >= [From] AND [Date] <= [To] THEN [Amount] END) for active contracts
• Why: Only count/sum contracts active between start/end dates
• Pattern: Spend Under Management = contracts overlapping the analysis period

Digital Marketing: Time-Based Metrics

• Use: DATEDIFF('second', [Start], [End]) for session duration, response time
• Why: Native time difference calculations
• Pattern: Avg Session Duration = DATEDIFF('second', [SessionStart], [SessionEnd])

---

Coverage: 61 production dashboard templates across Finance (P&L, Balance Sheet, Cash Flow), Banking (AuM, Lending, Loans), Sales (Basic, Expert, Margin, Discount/VAT, RFM, Market Share), Recurring Revenue (SaaS/Subscription), Retail (Sales, Inventory), E-Commerce, Digital Marketing (Google Analytics, Google Ads, Facebook Ads, Twitter Ads, Email Marketing), Procurement (Spend Analytics, Purchasing, Contracts), Supply Chain, Travel, and CEO Cockpits.