Two dashboards can disagree while both are doing exactly what they were configured to do. One counts events; another counts orders. One groups by a reporting day that ends at midnight in a different time zone. Before explaining a spike, check the unit in each row, the definition behind the metric and the delay between activity and reporting. This guide moves from users and events to dimensions, cohorts and data quality. It helps teams ask what a number represents before turning it into a decision.
🧭 Jump to a term
People and activity: User · Session · Pageview · Analytics event · Event property
Describe and measure: Dimension · Metric · Data grain · Data freshness · Reporting time zone
Explore journeys and trust the data: Cohort analysis · Funnel analysis · Path analysis · Data quality · Source of truth
⭐ Know these first
Start with User, Session, Analytics event, Cohort analysis, Data grain. Then follow the grouped learning order below.
📎 How to read this page
What it means gives the precise meaning. Operator translation gives the version you might hear in a real ecommerce meeting. In real life shows an illustrative example. Watch out and the confusion boxes show where a familiar term can mislead.
📈 Read the relationships first
These combinations are diagnostic hypotheses, not proof of causality. Compare the same period and scope, then investigate the mechanism.
👤 Users ↑ + sessions flat
Usually means: More recognized identities may not mean more visits. Check next: identity rules, device changes and consent.
📈 Event count ↑ + order count flat
Usually means: A change in instrumentation or event duplication may explain the gap. Check next: Compare event definitions with the order ledger.
🕒 Report refresh ↑ + current-day revenue ↑
Usually means: Late arriving data may backfill today’s report. Check next: data freshness before calling a trend.
🧮 Metric by dimension + totals differ
Usually means: A join or grain mismatch may duplicate records. Check next: Confirm what each row represents before aggregation.
People and activity
01 · 🟢 Core
User = A person or recognized identity interacting with a digital property
🧠 What it means
A person or recognized identity interacting with a digital property
💬 OPERATOR TRANSLATION
“Analytics may identify a person across visits only when its identity rules and consent permit; user counts are tool-defined, not a headcount.”
🛍️ In real life
One shopper browses on a phone and later signs in on a laptop; the analytics identity rules determine whether those actions connect.
🔗 Related: Session · Pageview · Analytics event · ↑ all terms
02 · 🟢 Core
Session = A grouped period of interaction under a tool’s session rules
🧠 What it means
A grouped period of interaction under a tool’s session rules
💬 OPERATOR TRANSLATION
“A session groups activity according to an analytics product’s start, inactivity and campaign rules. A session is not necessarily one visit or one person.”
🛍️ In real life
A shopper browses, leaves, then returns after the configured timeout; the tool may record a new session.
🔗 Related: User · Pageview · Analytics event · ↑ all terms
03 · 🔵 Operations
Pageview = A recorded view or load of a page
🧠 What it means
A recorded view or load of a page
💬 OPERATOR TRANSLATION
“A pageview is a tracking event that records a page being viewed under the implementation’s rules. Reloads, virtual page changes and blocked tracking affect counts.”
🛍️ In real life
A shopper opens a product page, reloads it and opens another page; the analytics setup may record three pageviews.
🔗 Related: User · Session · Analytics event · ↑ all terms
04 · 🟢 Core
Analytics event = A recorded interaction or occurrence
🧠 What it means
A recorded interaction or occurrence
💬 OPERATOR TRANSLATION
“An analytics event is a named occurrence, such as a purchase, search or add-to-cart action. Collection rules, consent and deduplication determine what reaches reports.”
🛍️ In real life
The storefront records an add_to_cart event when a shopper adds a product to the basket.
🔗 Related: User · Session · Pageview · ↑ all terms
05 · 🔵 Operations
Event property = A detail attached to an event
🧠 What it means
A detail attached to an event
💬 OPERATOR TRANSLATION
“An event property adds context to a recorded event, such as product ID, value or page location. Names, types and privacy treatment should be governed in the tracking plan.”
🛍️ In real life
A purchase event carries currency, order value and item identifiers as properties.
🔗 Related: User · Session · Pageview · ↑ all terms
Describe and measure
06 · 🔵 Operations
Dimension = A descriptive attribute used to group or filter data
🧠 What it means
A descriptive attribute used to group or filter data
💬 OPERATOR TRANSLATION
“A dimension describes a data point, such as channel, device or country. Definitions and available values depend on the source and reporting model.”
🛍️ In real life
An analyst groups orders by acquisition channel to compare paid search with email.
🔗 Related: Metric · Data grain · Data freshness · ↑ all terms
07 · 🔵 Operations
Metric = A quantitative measure calculated from data
🧠 What it means
A quantitative measure calculated from data
💬 OPERATOR TRANSLATION
“A metric is a numeric measure, such as orders, revenue or conversion rate. Confirm its formula, scope, currency, attribution and time basis before comparing tools.”
🛍️ In real life
A team compares completed orders week over week using the same order and date rules.
🔗 Related: Dimension · Data grain · Data freshness · ↑ all terms
08 · 🟢 Core
Data grain = The unit represented by one row or record
🧠 What it means
The unit represented by one row or record
💬 OPERATOR TRANSLATION
“Data grain states what one row represents, such as an order, line item, event or customer-day. Joining tables at mismatched grains can multiply values.”
🛍️ In real life
A sales table has one row per order line, so summing order value directly repeats orders with multiple lines.
🔗 Related: Dimension · Metric · Data freshness · ↑ all terms
09 · 🔵 Operations
Data freshness = How recently data has been updated
🧠 What it means
How recently data has been updated
💬 OPERATOR TRANSLATION
“Data freshness is the delay between an event occurring and its availability in a dataset or report. Processing schedules and late-arriving records create lags.”
🛍️ In real life
A dashboard refreshes hourly, so this morning’s orders are incomplete until the next load.
🔗 Related: Dimension · Metric · Data grain · ↑ all terms
10 · 🔵 Operations
Reporting time zone = Time zone used to assign timestamps to reporting dates
🧠 What it means
Time zone used to assign timestamps to reporting dates
💬 OPERATOR TRANSLATION
“A reporting time zone determines which calendar day receives an event near midnight. Compare reports only after confirming the property and export time-zone rules.”
🛍️ In real life
An order at 00:30 UTC may belong to the prior local reporting day for a West Coast store.
🔗 Related: Dimension · Metric · Data grain · ↑ all terms
Explore journeys and trust the data
11 · 🟢 Core
Cohort analysis = Comparison of groups sharing a starting characteristic or time
🧠 What it means
Comparison of groups sharing a starting characteristic or time
💬 OPERATOR TRANSLATION
“Cohort analysis tracks groups defined by a common starting event or period over time. Retention and repeat purchase results depend on cohort entry, observation window and exclusions.”
🛍️ In real life
A retailer compares repeat purchase in January and February first-order cohorts after 30 days.
🔗 Related: Funnel analysis · Path analysis · Data quality · ↑ all terms
12 · 🔵 Operations
Funnel analysis = Analysis of progression through defined steps
🧠 What it means
Analysis of progression through defined steps
💬 OPERATOR TRANSLATION
“Funnel analysis measures how users or events progress through an ordered set of steps. Entry, order, conversion window and repeat-entry rules change the result.”
🛍️ In real life
A team measures product view, add to cart, checkout and purchase completion for eligible sessions.
🔗 Related: Cohort analysis · Path analysis · Data quality · ↑ all terms
13 · 🔵 Operations
Path analysis = Analysis of sequences taken through pages or events
🧠 What it means
Analysis of sequences taken through pages or events
💬 OPERATOR TRANSLATION
“Path analysis explores sequences of pages or events before or after a selected step. It is descriptive and can expose common routes without proving why users took them.”
🛍️ In real life
An analyst examines the pages shoppers view between landing and purchase.
🔗 Related: Cohort analysis · Funnel analysis · Data quality · ↑ all terms
14 · 🔵 Operations
Data quality = Fitness of data for its intended use
🧠 What it means
Fitness of data for its intended use
💬 OPERATOR TRANSLATION
“Data quality covers properties such as accuracy, completeness, consistency and validity for a defined purpose. A clean chart cannot repair missing or duplicated inputs.”
🛍️ In real life
A sudden duplicate purchase event inflates reported orders until the tagging issue is fixed.
🔗 Related: Cohort analysis · Funnel analysis · Path analysis · ↑ all terms
15 · 🔵 Operations
Source of truth = The agreed authoritative source for a specified measure
🧠 What it means
The agreed authoritative source for a specified measure
💬 OPERATOR TRANSLATION
“A source of truth is the agreed system and definition used for a particular business question. Different sources can each be authoritative for different workflows.”
🛍️ In real life
Finance uses the order ledger for recognized sales, while web analytics explains observed browsing behavior.
🔗 Related: Cohort analysis · Funnel analysis · Path analysis · ↑ all terms
🔀 Users vs sessions vs pageviews
Users describe recognized identities; sessions group activity under tool rules; pageviews count page-view events. These are different units.
🔀 Dimension vs metric
A dimension describes or groups records; a metric quantifies them. Channel is a dimension; order count is a metric.
🔀 Cohort vs funnel analysis
Cohorts compare groups that share an entry characteristic over time. Funnels measure progression through ordered steps.
🔀 Source of truth vs one universal dashboard
The authoritative source depends on the question. Finance, order operations and behavior analytics may each own different measures.
🤔 Still confused?
Follow this thread: User → Dimension → Cohort analysis. That sequence moves from the basic object or relationship to the decisions and checks it supports.
Sources and scope
Primary documentation checked on 27 September 2026. Platform features and eligibility can change; examples and cartoon situations are illustrative.
