Session
A group of interactions by one visitor that happen close together. A new session starts after a period of inactivity (30 minutes by default), at midnight, or when the visitor arrives again from a different source.
Glossary
Analytics terms, explained simply
What each term means, how it is usually calculated and where to find it in PathClue.
- 50 terms
- Plain language
- Formulas where useful
Traffic
Visits and where they come from
The basics of website and app traffic.
Pageview
One load of a page — or, in single-page apps, one route change. Apps record screen views instead.
Unique visitor
A distinct person or device in a period. Without cookies it is estimated within a day; with sign-in, people are counted across devices.
Bounce rate
The share of sessions with only one pageview and no meaningful interaction.
How it’s calculated: single-page sessions ÷ all sessions
Referrer
The page a visitor came from before landing on your site, such as a search engine or another website.
UTM parameters
Tags added to a link — utm_source, utm_medium, utm_campaign, utm_term, utm_content — that say which campaign a visit came from.
Click ID
A parameter ad platforms add to the landing URL, such as gclid, fbclid or ttclid, that identifies a paid click even without UTM tags.
Channel
A group of sources of the same kind — organic search, paid search, social, email, AI assistants, direct — defined by rules.
Landing page
The first page a visitor sees in a session; also called the entry page.
AI referral
A visit that comes from an AI assistant such as ChatGPT, Perplexity, Claude or Gemini, usually after it cited your page.
Behavior
What people do
Events and the analyses built on them.
Event
Something a user did, recorded with a name and a time — a click, a sign-up, a purchase.
Property
Extra detail attached to an event or a user, such as order value, plan or button name.
Autocapture
Letting the script record clicks, form submissions and page changes on its own, so you decide later which ones become named events.
Funnel
An ordered set of steps — view product, add to cart, checkout, pay — showing how many people reach each step and where they drop.
How it’s calculated: people at step n ÷ people at step 1
Conversion rate
The share of visitors or sessions that complete a goal.
How it’s calculated: conversions ÷ visitors × 100%
Retention
The share of users who come back and do something again after their first visit or action.
How it’s calculated: users active in period n ÷ users in the cohort
Path analysis
The sequences of pages or events people follow before or after a point you choose.
Heatmap
A picture of where people click or how far they scroll on a page, drawn as colour on top of it.
Session replay
A reconstruction of a visitor’s session — clicks, scrolls and page changes — that you can watch back. Inputs are usually masked.
Rage click
Several rapid clicks on the same spot — a sign that something didn’t respond as expected.
A/B test and significance
Showing two versions to similar visitors and checking whether the difference in results is larger than chance. A p-value below 0.05 is significant at 95% confidence.
Users
Who your users are
Profiles, identity and groups.
Identity resolution
Linking the devices and anonymous visits of one person into a single profile, usually when they sign in.
User profile
Everything known about one person — properties and a timeline of visits and events — on one page.
Segment
A group of users defined by rules on behaviour or properties, such as “visited pricing three times, never purchased”.
Cohort
A group of users who share a starting point — signed up in the same week, or were acquired through the same channel in the same month.
DAU / MAU
Daily and monthly active users — people who did something meaningful in a day or a month.
Stickiness
How often active users come back, usually daily active users divided by monthly active users.
How it’s calculated: DAU ÷ MAU
Account (group)
A company or workspace that users belong to — used to analyse B2B products by customer rather than by person.
Growth & revenue
Growth and money
The metrics on a growth board.
AARRR
Five stages for looking at growth: people arrive (acquisition), get value (activation), come back (retention), pay (revenue) and bring others (referral).
Goal
An event or page that counts as success, such as a sign-up or a purchase. Goals can be chained into milestones.
Average order value (AOV)
The average amount spent per order.
How it’s calculated: revenue ÷ orders
Repeat purchase rate
Of the customers who ordered in the chosen period, the share who ordered at least twice.
How it’s calculated: customers with 2+ orders ÷ all customers
Lifetime value (LTV)
The revenue a customer brings over time, often measured after 30, 60 or 90 days for a cohort.
How it’s calculated: cohort revenue to date ÷ cohort size
Payback period
How long it takes for the revenue from a cohort to cover what it cost to acquire it.
MRR
Monthly recurring revenue from subscriptions, split into new, expansion and churned MRR.
Churn
Customers or revenue lost in a period — cancelled subscriptions or users who stopped coming back.
Product health score
A single 0–100 score combining adoption, stickiness and growth (new and recovered users against those who dropped).
Attribution
Credit for conversions
Which marketing gets the credit.
Attribution model
A rule for sharing credit for a conversion among the touchpoints before it — last session, first touch, last non-direct or linear.
Last non-direct touch
Gives credit to the last source that was not a direct visit, so returning by bookmark doesn’t erase the campaign that brought someone.
ROI and ROAS
ROI weighs what you earned against what you spent overall; ROAS looks only at ad spend and the revenue it brought.
How it’s calculated: ROAS = revenue ÷ ad spend
Conversion API (postback)
Sending conversions from a server to an ad platform so its bidding learns from real results, without relying on browser pixels.
Offline conversion
A sale or deal that happens outside the website — a phone order, an in-store purchase — imported and linked back to the original click.
Privacy & data
Privacy, bots and data
How data is collected and kept clean.
Cookieless analytics
Counting visitors without storing cookies or other identifiers on the device — for example with a server-side ID that changes daily.
Bot traffic
Visits from software rather than people — crawlers, monitors, scrapers — which inflate traffic and lower conversion rates if mixed in.
AI crawler
A bot that fetches pages for AI training or AI answers, such as GPTBot, ClaudeBot or PerplexityBot.
robots.txt and llms.txt
robots.txt tells crawlers which paths they may fetch; llms.txt is a community proposal that guides AI models to a site’s important pages. Neither is enforceable.
Global Privacy Control (GPC)
A browser signal asking sites not to sell or share the visitor’s data; PathClue counts such visits anonymously.
Data retention
How long collected data is kept before it is deleted automatically.
Data processing agreement (DPA)
A contract that sets out how a processor handles personal data on a controller’s behalf, as required by GDPR Article 28.
MCP (Model Context Protocol)
An open protocol that lets AI assistants call tools and read data from other services — PathClue exposes read-only analytics tools through it.