# Glossary of business performance monitoring terms > Short definitions of the terms used across InnerMind Insights. Canonical page: https://inner-mind.ai/insights/glossary/ ## Performance Surveillance Intelligence (PSI) An always-on intelligence layer for business performance. It monitors continuously, learns what normal looks like for each part of the business, detects meaningful change, separates signal from noise, and helps leaders understand what changed, why it matters and what to do next. Unlike business intelligence, which waits to be asked, PSI does the noticing itself. Learn more: [Dashboards Wait for You to Look: Why Monitoring Beats Reporting](https://inner-mind.ai/insights/dashboards-wait-for-you-to-look/) ## Business performance monitoring The continuous comparison of each part of a business against its own normal, learned from its history, so that meaningful changes are surfaced and ranked by impact without anyone having to go looking for them. It is the opposite of reporting, which shows what you asked to see when you remember to look. Learn more: [Dashboards Wait for You to Look: Why Monitoring Beats Reporting](https://inner-mind.ai/insights/dashboards-wait-for-you-to-look/) ## Baseline A learned description of what normal looks like for one metric: its usual level, how much it normally varies, which way it is trending and its seasonal rhythm. A baseline is built from the metric's own history and is the yardstick a new value is judged against. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## Metric series One metric tracked consistently over time, under one name, in one unit and at one period, such as monthly revenue. Building a series means mapping variant labels like Revenue, Total Revenue and Net Sales to the same metric. Without a series there is no history, and without history there is no baseline. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## Normal range (expected range) The band of values a metric usually falls within, commonly set as its average plus or minus about 2.5 standard deviations over a trailing window. A value outside the normal range is unusual for this business, which makes it worth a look. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## Standard deviation (spread) A measure of how much a metric typically moves around its average. A metric with a small standard deviation is steady, so a modest change is significant; one with a large standard deviation is volatile, so the same change may be ordinary. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## Leave-one-out baseline A baseline calculated without the value being tested. Leaving that value out stops a genuinely unusual month from inflating the spread and widening the range it is being judged against. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## Trend A consistent movement of a metric in one direction over time. A trend is usually tested by fitting a straight line through recent values and requiring both a meaningful slope and a good fit, which separates a real drift from ordinary noise. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## R² (coefficient of determination) A number from 0 to 1 describing how closely a set of points follows a fitted line. In trend detection, an R² of about 0.5 or higher suggests the movement is consistent rather than random. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## Seasonality A pattern that repeats at the same point each year, such as a December peak or a summer slowdown. Ignoring seasonality makes normal swings look like problems and hides real ones; adjusting for it reliably needs at least two full years of monthly history. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## Seasonal index For each calendar month, its average value divided by the overall average. An index of 1.25 means that month typically runs 25% above an average month. Dividing a month's value by its index gives a deseasonalized value that can be compared fairly with the rest of the year. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## Cold start The period when a metric has too little history for a meaningful baseline, typically fewer than about four periods. A careful monitoring system holds off on flagging anything during a cold start rather than guessing. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## Signal A meaningful change in a business metric or document, such as a value outside its normal range or a sustained trend, as opposed to noise, which is ordinary variation. Separating signal from noise is the core job of business performance monitoring. Learn more: [Dashboards Wait for You to Look: Why Monitoring Beats Reporting](https://inner-mind.ai/insights/dashboards-wait-for-you-to-look/) ## Metric polarity Which direction is better for a given metric. Higher is better for revenue and margin; lower is better for costs, churn and days sales outstanding. Polarity is what turns "up" or "down" into "better" or "worse". Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/) ## Threshold alert An alert that fires when a metric crosses a fixed line set in advance. Threshold alerts only catch problems someone predicted, ignore seasonality, and cannot see a slow decline that stays inside the line. Learn more: [Dashboards Wait for You to Look: Why Monitoring Beats Reporting](https://inner-mind.ai/insights/dashboards-wait-for-you-to-look/) ## Alert fatigue The tendency to ignore alerts after too many of them turn out to be unimportant. It is usually caused by fixed thresholds and unadjusted seasonality, and it is reduced by comparing metrics to their own normal and ranking findings by impact. Learn more: [Dashboards Wait for You to Look: Why Monitoring Beats Reporting](https://inner-mind.ai/insights/dashboards-wait-for-you-to-look/) ## Deterministic analysis A calculation that always produces the same output from the same input. In business monitoring it means the same data always yields the same flagged changes in the same order, so every finding can be reproduced, audited and tested. Learn more: [Math Detects, AI Explains: How to Trust AI With Your Business Numbers](https://inner-mind.ai/insights/math-detects-ai-explains/) ## Days sales outstanding (DSO) The average number of days a business takes to collect payment after a sale. Rising DSO means more revenue is sitting in receivables instead of the bank; at $40,000 of daily sales, each extra day ties up about $40,000 of cash. Learn more: [Dashboards Wait for You to Look: Why Monitoring Beats Reporting](https://inner-mind.ai/insights/dashboards-wait-for-you-to-look/) ## Gross margin Revenue minus the direct cost of what was sold, as a percentage of revenue. Because it is a ratio, small moves in gross margin can represent large amounts of gross profit. Learn more: [What “Normal” Looks Like: How to Build a Baseline for Your Business Metrics](https://inner-mind.ai/insights/how-to-build-a-baseline-for-business-metrics/)