KPI catalog

Generated from kpis/catalog.yaml - the single source of truth for the Python engine, the SQL reference queries, the story site and the dashboard. 28 KPIs in 5 domains. Every KPI page shows its definition, how to read it, the common misreadings, its SQL, the Python function and a worked example taken from the test suite.

Service operations & SLA

KPI	Business question	Direction	Target	Owner	Current
First response time (median)	How long does a customer wait before a human acknowledges the ticket?	Lower is better ↓	≤ 2.0 h	Service Desk Lead	✓ 0.9 h
Resolution time (median)	How long does it take to solve a customer's problem end to end?	Lower is better ↓	≤ 24.0 h	Service Desk Manager	✓ 11.1 h
SLA compliance	What share of tickets did we resolve within the time we promised?	Higher is better ↑	≥ 90.0%	Service Desk Manager	✓ 94.4%
Aged backlog (> 7 days)	How much of the open work is already old enough to hurt the SLA?	Lower is better ↓	≤ 10.0%	Queue Team Lead	✕ 76.7%
First contact resolution (FCR)	How often do we solve the problem in a single interaction?	Higher is better ↑	≥ 70.0%	Service Desk Manager	! 62.5%
Reopen rate	How often is a "resolved" ticket not really resolved?	Lower is better ↓	≤ 5.0%	Quality Lead (Service)	✓ 4.1%

Maintenance & asset reliability

KPI	Business question	Direction	Target	Owner	Current
Mean time between failures (MTBF)	How long does an asset run, on average, before it fails?	Higher is better ↑	≥ 150.0 h	Reliability Engineer	! 118.0 h
Mean time to repair (MTTR)	Once an asset fails, how long until it runs again?	Lower is better ↓	≤ 4.0 h	Maintenance Supervisor	! 5.3 h
Availability	Of the time the asset was supposed to run, how much could it actually run?	Higher is better ↑	≥ 97.0%	Maintenance Manager	! 95.7%
Preventive maintenance compliance	Are we doing the preventive work we planned, when we planned it?	Higher is better ↑	≥ 90.0%	Maintenance Planner	! 86.6%
Planned maintenance share	How much of our maintenance effort is planned rather than reactive?	Higher is better ↑	≥ 80.0%	Maintenance Manager	! 65.5%

Inventory & demand

KPI	Business question	Direction	Target	Owner	Current
Fill rate (units)	Of the units customers asked for, how many did we ship from stock?	Higher is better ↑	≥ 95.0%	Supply Planner	✕ 79.1%
On time in full (OTIF)	How many orders arrived complete and on the promised date?	Higher is better ↑	≥ 90.0%	Logistics Manager	✕ 74.5%
Inventory turnover	How many times a year does the stock we hold get sold?	Higher is better ↑	≥ 8.0x / year	Inventory Controller	! 5.6x / year
Days of cover	If demand continues as usual, how many days will current stock last?	Inside a band ↔	14.0 days to 35.0 days	Supply Planner	✕ 63.1 days
Stockout rate (SKU-days)	How often is an item simply not available?	Lower is better ↓	≤ 3.0%	Supply Planner	! 5.9%
Obsolete stock share	How much of our inventory value has stopped moving?	Lower is better ↓	≤ 5.0%	Inventory Controller	✕ 27.5%
Forecast error (MAPE)	On average, by what percentage does each item's forecast miss?	Lower is better ↓	≤ 40.0%	Demand Planner	✕ 389.2%
Forecast error (WAPE)	Of all the units we sold, how many did the forecast get wrong?	Lower is better ↓	≤ 30.0%	Demand Planner	✕ 77.0%
Forecast bias	Do we systematically forecast too much or too little?	Inside a band ↔	-5.0% to 5.0%	Demand Planner	✓ -2.1%

Customers

KPI	Business question	Direction	Target	Owner	Current
Customer retention (quarter over quarter)	Of the customers who bought last quarter, how many bought again this quarter?	Higher is better ↑	≥ 50.0%	Customer Success / CRM Lead	✓ 62.8%
Repeat purchase rate	What share of customers come back for a second order?	Higher is better ↑	≥ 60.0%	CRM Lead	✓ 72.4%
Revenue at risk (RFM "At risk")	How much of our revenue comes from good customers who have stopped buying?	Lower is better ↓	≤ 10.0%	CRM Lead	✓ 9.4%
Net Promoter Score (simulated survey)	Would our customers recommend us?	Higher is better ↑	≥ +30	Customer Experience Lead	! +12

Data quality as a KPI

KPI	Business question	Direction	Target	Owner	Current
Completeness	Are the fields we need actually filled in?	Higher is better ↑	≥ 99.0%	Data Steward	! 96.3%
Uniqueness	Is each real-world record stored exactly once?	Higher is better ↑	≥ 99.9%	Data Steward	✕ 96.8%
Validity	Do the values respect the business rules?	Higher is better ↑	≥ 98.0%	Data Steward	✓ 98.9%
Timeliness	Is the data available soon enough after the event to act on it?	Higher is better ↑	≥ 95.0%	Data Engineer	✓ 97.4%

Current = value on the full dataset of the case (illustrative). Status icons: ✓ on target, ! watch, ✕ act.