Metric definitions. How each number is computed.
The metric catalogue the panel, the reports, the alerts and the Query API use. Each metric is defined in one place and every screen reads the same definition: its name, definition, unit, which way is good, how it is computed and which date it counts by. The list is generated from the product itself.
This page is generated at build time from apps/api/locales/en/analytics.json, apps/api/src/features/analytics/domain/metric-catalogue.ts in the repository.
For every metric
- Rates are computed from the sums of numerator and denominator (
Σ numerator ÷ Σ denominator); daily or row rates are never averaged. - Distinct counts (visitors, converting people, converting accounts) are counted again for each period and row; daily values are not added up.
- A missing value is written
—;0really means zero. A metric of a source that is not connected is empty, not0. - Test, internal and bot traffic are left out by default and added from the filter bar. Traffic quality metrics always count bots.
- A metric that takes a role is narrowed by a business role:
conversions:signupcounts sign-ups only; without a role the app’s conversion roles count.
The names in the formula are the catalogue’s base measures.
Traffic
Visitors
Distinct people with a session (after identity merging). Daily values are not added up; they are counted again for every period and row.
- Name in a query
- visitors
- Unit
- Count
- Good direction
- Up is good
- Computation
- Distinct count: counted again for each period and row
- Formula
- Σ visitors
- Counting basis
- Event date
New visitors
Visitors seen for the first time: the person's first appearance falls within the period (within the day, week or month of a time breakdown).
- Name in a query
- new_visitors
- Unit
- Count
- Good direction
- Up is good
- Computation
- Distinct count: counted again for each period and row
- Formula
- Σ new_visitors
- Counting basis
- Event date
Sessions
Visits. 30 minutes of inactivity, midnight in the app's time zone or a new campaign start a new session. Counted on the day the session started.
- Name in a query
- sessions
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ sessions
- Counting basis
- Event date
Engaged sessions
Sessions with at least 10 s of active time, at least 2 page or screen views, or a conversion.
- Name in a query
- engaged_sessions
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ engaged_sessions
- Counting basis
- Event date
Engagement rate
Engaged sessions ÷ sessions.
- Name in a query
- engagement_rate
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ engaged_sessions ÷ Σ sessions
- Counting basis
- Event date
Bounce rate
Sessions without engagement ÷ sessions (1 − engagement rate). The single-page rate is a separate metric.
- Name in a query
- bounce_rate
- Unit
- Percent
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ bounces ÷ Σ sessions
- Counting basis
- Event date
Single-page rate
Sessions with only one page or screen view ÷ sessions.
- Name in a query
- single_page_rate
- Unit
- Percent
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ single_page_sessions ÷ Σ sessions
- Counting basis
- Event date
Page views
Page view calls (page). Counted on the day the view happened.
- Name in a query
- pageviews
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ pageviews
- Counting basis
- Event date
Screen views
Screen view calls of mobile apps (screen).
- Name in a query
- screen_views
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ screen_views
- Counting basis
- Event date
Views per session
Page and screen views of the sessions ÷ sessions.
- Name in a query
- views_per_session
- Unit
- Ratio
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ session_views ÷ Σ sessions
- Counting basis
- Event date
Average session duration
Active time of the sessions ÷ sessions. Active time is time in a visible tab; a hidden tab does not count.
- Name in a query
- avg_session_duration
- Unit
- Duration
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ active_ms ÷ Σ sessions
- Counting basis
- Event date
Events
Business and custom events (track); page, screen and identity calls excluded.
- Name in a query
- events
- Unit
- Count
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ events
- Counting basis
- Event date
People doing the event
Unique persons doing business events (track); merged identities count as one person. Next to the event count, it shows repeats.
- Name in a query
- event_persons
- Unit
- Count
- Good direction
- No direction
- Computation
- Distinct count: counted again for each period and row
- Formula
- Σ event_persons
- Counting basis
- Event date
Accounts doing the event
Unique accounts doing business events: the event’s account, else the person’s only account.
- Name in a query
- event_accounts
- Unit
- Count
- Good direction
- No direction
- Computation
- Distinct count: counted again for each period and row
- Formula
- Σ event_accounts
- Counting basis
- Event date
Events per person
Events ÷ people doing the event (from the totals).
- Name in a query
- events_per_person
- Unit
- Ratio
- Good direction
- No direction
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ events ÷ Σ event_persons
- Counting basis
- Event date
Entries
Sessions that started on this page.
- Name in a query
- entries
- Unit
- Count
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ entries
- Counting basis
- Event date
Exits
Sessions that ended on this page.
- Name in a query
- exits
- Unit
- Count
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ exits
- Counting basis
- Event date
Exit rate
Exits ÷ page views.
- Name in a query
- exit_rate
- Unit
- Percent
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ exits ÷ Σ pageviews
- Counting basis
- Event date
Consented sessions
Sessions whose events carried analytics consent (or whose source does not ask for it).
- Name in a query
- consented_sessions
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ consented_sessions
- Counting basis
- Event date
Consent rate
Consented sessions ÷ all sessions. Sessions without a consent signal (server) are in the denominator.
- Name in a query
- consent_rate
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ consented_sessions ÷ Σ sessions
- Counting basis
- Event date
Mobile sessions
Sessions whose device type is mobile.
- Name in a query
- mobile_sessions
- Unit
- Count
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ mobile_sessions
- Counting basis
- Event date
Behaviour
Engagement time per view
Active time of the page views ÷ page views that reported leaving.
- Name in a query
- avg_engagement_time
- Unit
- Duration
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ active_ms ÷ Σ measured_views
- Counting basis
- Event date
Average scroll depth
Average deepest scroll of the page views that reported leaving.
- Name in a query
- avg_scroll_depth
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ scroll_sum ÷ Σ scroll_count × 0.01
- Counting basis
- Event date
Clicks
Recorded clicks.
- Name in a query
- clicks
- Unit
- Count
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ clicks
- Counting basis
- Event date
Rage clicks
At least 3 clicks on the same spot within 1 s: something does not work as expected.
- Name in a query
- rage_clicks
- Unit
- Count
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ rage_clicks
- Counting basis
- Event date
Dead clicks
Clicks followed by no change on the page within 1 s: an element taken for a button or link.
- Name in a query
- dead_clicks
- Unit
- Count
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ dead_clicks
- Counting basis
- Event date
JS errors
JavaScript errors caught on the page.
- Name in a query
- js_errors
- Unit
- Count
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ js_errors
- Counting basis
- Event date
Frustrated sessions
Sessions with a rage click, a dead click or a JS error.
- Name in a query
- frustrated_sessions
- Unit
- Count
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ frustrated_sessions
- Counting basis
- Event date
Frustration rate
Frustrated sessions ÷ all sessions.
- Name in a query
- frustration_rate
- Unit
- Percent
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ frustrated_sessions ÷ Σ sessions
- Counting basis
- Event date
Traffic quality
Bot sessions
Sessions flagged as bots; counted whatever the screen's bot switch (so the bot share can be watched).
- Name in a query
- bot_sessions
- Unit
- Count
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ bot_sessions
- Counting basis
- Event date
Sessions from an ad click
Sessions whose landing carried an ad platform's click id (gclid, fbclid, li_fat_id…).
- Name in a query
- ad_click_sessions
- Unit
- Count
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ ad_click_sessions
- Counting basis
- Event date
Repeated ad clicks
Ad-click sessions of a visitor who also arrived from an ad click within the 30 minutes before: a sign of click fraud or a broken landing page.
- Name in a query
- repeat_ad_clicks
- Unit
- Count
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ repeat_ad_clicks
- Counting basis
- Event date
Repeated ad click rate
Repeated ad clicks ÷ sessions from an ad click.
- Name in a query
- repeat_ad_click_rate
- Unit
- Percent
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ repeat_ad_clicks ÷ Σ ad_click_sessions
- Counting basis
- Event date
Conversions and revenue
Conversions
Events whose definition has a conversion role; with a role chosen, only that role. Counted on the day the event happened.
Takes a business role.
- Name in a query
- conversions:<role>
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ conversions
- Counting basis
- Event date
Conversion rate
Sessions that reached a conversion role (the chosen role) ÷ sessions.
Takes a business role.
- Name in a query
- conversion_rate:<role>
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ converted_sessions ÷ Σ sessions
- Counting basis
- Event date
Converting people
Unique persons doing an event of a conversion role (with a role chosen, that role); merged identities count once. Not added over days: counted again for each period. With an attributed dimension, the persons whose conversions are credited to the row.
Takes a business role.
- Name in a query
- converters:<role>
- Unit
- Count
- Good direction
- Up is good
- Computation
- Distinct count: counted again for each period and row
- Formula
- Σ conversion_persons
- Counting basis
- Event date
Converting accounts
Unique accounts doing an event of a conversion role: the event's account, else the person's only account (with a role chosen, that role; for purchase the paying accounts). Not added over days.
Takes a business role.
- Name in a query
- converting_accounts:<role>
- Unit
- Count
- Good direction
- Up is good
- Computation
- Distinct count: counted again for each period and row
- Formula
- Σ conversion_accounts
- Counting basis
- Event date
Conversions with a known source
Conversions credited, under the chosen attribution model, to a touch whose channel is neither direct nor unassigned ÷ all conversions (sums of the weights).
- Name in a query
- known_source_rate
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ known_source ÷ Σ credited_conversions
- Counting basis
- Event date
Revenue
Gross amount of sales (events of the refund, checkout start and failed payment roles excluded) in the app's currency. Amounts in other currencies are converted at the TCMB rate of the event's day.
- Name in a query
- revenue
- Unit
- Amount (the app’s currency)
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ revenue
- Counting basis
- Event date
Refunds
Amount of the refund role's events (positive whatever their sign), in the app's currency.
- Name in a query
- refunds
- Unit
- Amount (the app’s currency)
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ refunds
- Counting basis
- Event date
Net revenue
Net amount of sales (the amount before tax when the event gave it, else the gross amount) − the refunds' net amount.
- Name in a query
- net_revenue
- Unit
- Amount (the app’s currency)
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ net_revenue
- Counting basis
- Event date
Revenue per visitor
Revenue ÷ visitors.
- Name in a query
- revenue_per_visitor
- Unit
- Amount (the app’s currency)
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ revenue ÷ Σ visitors
- Counting basis
- Event date
Conversions not from ads
The share of conversions the attribution service credits to touches not resolved to an ad, or to no touch (under the chosen model; organic, direct, referral…). With conversions from ads it adds up to all conversions.
Takes a business role.
- Name in a query
- non_ad_conversions:<role>
- Unit
- Count
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ non_ad_conversions
- Counting basis
- Event date
Advertising
Ad spend
Spend the ad platforms report (excluding VAT), converted to the app's currency with the day's rate. Counted on the day it was spent (the ad account's calendar).
- Name in a query
- spend
- Unit
- Amount (the app’s currency)
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ spend
- Counting basis
- Event date
Impressions
How often the ads were shown on the platform.
- Name in a query
- impressions
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ impressions
- Counting basis
- Event date
Ad clicks
Ad clicks the platform counts (not the clicks on your pages).
- Name in a query
- ad_clicks
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ ad_clicks
- Counting basis
- Event date
Click-through rate
Ad clicks ÷ impressions (from the totals).
- Name in a query
- ctr
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ ad_clicks ÷ Σ impressions
- Counting basis
- Event date
Cost per click
Spend ÷ ad clicks (from the totals).
- Name in a query
- cpc
- Unit
- Amount (the app’s currency)
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ spend ÷ Σ ad_clicks
- Counting basis
- Event date
Cost per mille
Spend ÷ impressions × 1000.
- Name in a query
- cpm
- Unit
- Amount (the app’s currency)
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ spend ÷ Σ impressions × 1,000
- Counting basis
- Event date
Video views
Views of video ads as the platform counts them.
- Name in a query
- video_views
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ video_views
- Counting basis
- Event date
View rate
Video views ÷ impressions of the video ads.
- Name in a query
- video_view_rate
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ video_views ÷ Σ video_impressions
- Counting basis
- Event date
Cost per view
Spend of the video ads ÷ video views.
- Name in a query
- cost_per_view
- Unit
- Amount (the app’s currency)
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ video_spend ÷ Σ video_views
- Counting basis
- Event date
Watched 25 % of the video
Views reaching at least a quarter of the video, as the platform reports them; empty where the platform does not report it.
- Name in a query
- video_p25
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ video_p25
- Counting basis
- Event date
Watched 50 % of the video
Views reaching at least half of the video, as the platform reports them.
- Name in a query
- video_p50
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ video_p50
- Counting basis
- Event date
Watched 75 % of the video
Views reaching at least three quarters of the video, as the platform reports them.
- Name in a query
- video_p75
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ video_p75
- Counting basis
- Event date
Watched the whole video
Views reaching the end of the video, as the platform reports them.
- Name in a query
- video_p100
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ video_p100
- Counting basis
- Event date
Platform conversions
Conversions the ad platform counts itself (its own pixel and attribution window); never mixed with Postlip Metrics' conversions.
- Name in a query
- platform_conversions
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ platform_conversions
- Counting basis
- Event date
Platform conversion value
The value the platform gives its own conversions, in the app's currency.
- Name in a query
- platform_conversion_value
- Unit
- Amount (the app’s currency)
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ platform_conversion_value
- Counting basis
- Event date
Cost per platform conversion
Spend ÷ platform conversions.
- Name in a query
- platform_cpa
- Unit
- Amount (the app’s currency)
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ spend ÷ Σ platform_conversions
- Counting basis
- Event date
Platform ROAS
Platform conversion value ÷ spend.
- Name in a query
- platform_roas
- Unit
- Ratio
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ platform_conversion_value ÷ Σ spend
- Counting basis
- Event date
Conversions from ads
The app's conversions credited by the attribution service to touches whose campaign resolved to an ad (under the chosen model); with a role chosen, only that role. Counted on the day the conversion happened.
Takes a business role.
- Name in a query
- ad_conversions:<role>
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ ad_conversions
- Counting basis
- Event date
Revenue from ads
Net revenue (excluding tax, refunds deducted) of the conversions from ads, by their credit.
- Name in a query
- ad_revenue
- Unit
- Amount (the app’s currency)
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ ad_revenue
- Counting basis
- Event date
Cost per lead
Spend ÷ leads from ads (role: lead).
- Name in a query
- cpl
- Unit
- Amount (the app’s currency)
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ spend ÷ Σ ad_conversions
- Counting basis
- Event date
Cost per conversion
Spend ÷ conversions from ads; with a role chosen, the cost per that role (e.g. cost per sign-up).
Takes a business role.
- Name in a query
- cpa:<role>
- Unit
- Amount (the app’s currency)
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ spend ÷ Σ ad_conversions
- Counting basis
- Event date
ROAS
Net revenue from ads ÷ spend (both excluding tax).
- Name in a query
- roas
- Unit
- Ratio
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ ad_revenue ÷ Σ spend
- Counting basis
- Event date
Click → conversion rate
Conversions from ads ÷ ad clicks (from the sums; with a role chosen, that role). Noisy for campaigns with few clicks: lists rank with a minimum-clicks threshold.
Takes a business role.
- Name in a query
- ad_conversion_rate:<role>
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ ad_conversions ÷ Σ ad_clicks
- Counting basis
- Event date
Planned budget
The budget plan spread evenly over the days (monthly budget ÷ days of the month).
- Name in a query
- budget
- Unit
- Amount (the app’s currency)
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ budget
- Counting basis
- Event date
Budget pacing
Spend ÷ the planned budget of the same days: 100 % is on plan.
- Name in a query
- budget_pacing
- Unit
- Percent
- Good direction
- No direction
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ spend ÷ Σ budget
- Counting basis
- Event date
Web analytics connections
GA4 sessions
Sessions of the connected Google Analytics 4 property, by the property's day. The definition differs from the pixel's (GA4 counts per browser, the pixel per tab and with a new campaign).
- Name in a query
- ga4_sessions
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ ga4_sessions
- Counting basis
- Event date
GA4 engaged sessions
Sessions GA4 counts as engaged (10 s+, 2+ pages or a key event).
- Name in a query
- ga4_engaged_sessions
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ ga4_engaged_sessions
- Counting basis
- Event date
GA4 engagement rate
GA4 engaged sessions ÷ GA4 sessions (from the sums).
- Name in a query
- ga4_engagement_rate
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ ga4_engaged_sessions ÷ Σ ga4_sessions
- Counting basis
- Event date
GA4 daily active users (average)
The average of GA4's daily active users over the days with data. A count of the day: days are never added up and shown as unique users.
- Name in a query
- ga4_active_users
- Unit
- Decimal number
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ ga4_active_users ÷ Σ ga4_days
- Counting basis
- Event date
GA4 key events
Events marked as key events in the GA4 property.
- Name in a query
- ga4_key_events
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ ga4_key_events
- Counting basis
- Event date
GA4 revenue
GA4's purchase revenue, converted from the property's currency to the app's with the day's TCMB rate; a day without a rate is not counted.
- Name in a query
- ga4_revenue
- Unit
- Amount (the app’s currency)
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ ga4_revenue
- Counting basis
- Event date
Clicks from Google
Search result clicks as Search Console reports them, by Google's day. The last days arrive a few days late.
- Name in a query
- gsc_clicks
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ gsc_clicks
- Counting basis
- Event date
Impressions on Google
Times a page of the site appeared in Google's search results.
- Name in a query
- gsc_impressions
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ gsc_impressions
- Counting basis
- Event date
Search click-through rate
Clicks from Google ÷ impressions on Google (from the sums; never an average of daily rates).
- Name in a query
- gsc_ctr
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ gsc_clicks ÷ Σ gsc_impressions
- Counting basis
- Event date
Average position
Average position weighted by impressions: Σ (position × impressions) ÷ Σ impressions. Lower is better (1 = the first result).
- Name in a query
- gsc_position
- Unit
- Decimal number
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ gsc_position_sum ÷ Σ gsc_impressions
- Counting basis
- Event date
Mailing and SMS
Sent
Messages the provider accepted (a cohort by send time).
- Name in a query
- email.sent, sms.sent
- Unit
- Count
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ sent
- Counting basis
- Cohort (day sent)
Delivered
Messages the provider reported delivered.
- Name in a query
- email.delivered, sms.delivered
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ delivered
- Counting basis
- Cohort (day sent)
Opened
Messages a person opened at least once (automatic prefetches left out).
- Name in a query
- email.opened
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ opened
- Counting basis
- Cohort (day sent)
Automatic opens
Messages whose opens all look automatic (seconds after the send, security scanners).
- Name in a query
- email.machine_opened
- Unit
- Count
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ machine_opened
- Counting basis
- Cohort (day sent)
Clicked
Messages with at least one tracked click.
- Name in a query
- email.clicked
- Unit
- Count
- Good direction
- Up is good
- Computation
- Sum: days and rows add up
- Formula
- Σ clicked
- Counting basis
- Cohort (day sent)
Bounced
Messages that bounced permanently (the address is suppressed).
- Name in a query
- email.bounced
- Unit
- Count
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ bounced
- Counting basis
- Cohort (day sent)
Complaints
Messages the recipient marked as spam.
- Name in a query
- email.complained
- Unit
- Count
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ complained
- Counting basis
- Cohort (day sent)
Unsubscribed
Messages whose recipient unsubscribed.
- Name in a query
- email.unsubscribed
- Unit
- Count
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ unsubscribed
- Counting basis
- Cohort (day sent)
Failed
Messages the provider refused for good.
- Name in a query
- email.failed, sms.failed
- Unit
- Count
- Good direction
- Down is good
- Computation
- Sum: days and rows add up
- Formula
- Σ failed
- Counting basis
- Cohort (day sent)
Delivery rate
Σ delivered ÷ Σ sent.
- Name in a query
- email.delivery_rate, sms.delivery_rate
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ delivered ÷ Σ sent
- Counting basis
- Cohort (day sent)
Open rate
Σ opened ÷ Σ delivered.
- Name in a query
- email.open_rate
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ opened ÷ Σ delivered
- Counting basis
- Cohort (day sent)
Click rate
Σ clicked ÷ Σ delivered.
- Name in a query
- email.click_rate
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ clicked ÷ Σ delivered
- Counting basis
- Cohort (day sent)
Click-to-open rate
Σ clicked ÷ Σ opened.
- Name in a query
- email.click_to_open_rate
- Unit
- Percent
- Good direction
- Up is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ clicked ÷ Σ opened
- Counting basis
- Cohort (day sent)
Bounce rate
Σ bounced ÷ Σ sent.
- Name in a query
- email.bounce_rate
- Unit
- Percent
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ bounced ÷ Σ sent
- Counting basis
- Cohort (day sent)
Complaint rate
Σ complaints ÷ Σ delivered.
- Name in a query
- email.complaint_rate
- Unit
- Percent
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ complained ÷ Σ delivered
- Counting basis
- Cohort (day sent)
Unsubscribe rate
Σ unsubscribed ÷ Σ delivered.
- Name in a query
- email.unsubscribe_rate
- Unit
- Percent
- Good direction
- Down is good
- Computation
- Rate: sum of numerators ÷ sum of denominators
- Formula
- Σ unsubscribed ÷ Σ delivered
- Counting basis
- Cohort (day sent)
Credits
The provider's credits of the SMS parts sent.
- Name in a query
- sms.credits
- Unit
- Count
- Good direction
- No direction
- Computation
- Sum: days and rows add up
- Formula
- Σ credits
- Counting basis
- Cohort (day sent)
Followers
Followers at the end of the period (the last known value; days are not added up). History starts with the first sync.