Intempt Docs
GuidesGetting Started

Attribute reference

Attributes are properties that help you define the specifics of an event, a user, or an account. Intempt has many attributes defined in its Data Model.

Attribute types

Attributes are properties that help you define the specifics of an event, a user, or an account. Intempt has many attributes defined in its Data Model.

Attribute typeDescription
Event attributesDescribes the events that are tracked within your website or application.
User attributesInformation associated with a user, such as their name, email, and other information.
Account attributesInformation associated with a group of users, such as companies. Examples of account attributes are company name, domain name, deal size, etc.
Calculated attributesA formula over one or more event-derived components (count, sum, average, min, max, or a value pulled from an event attribute).
RFM attributesScores a profile into one of 6 Recency/Frequency/Monetary segments based on purchase-like event data.
Qualify (Fit × Activity) attributesA weighted profile-fit filter crossed with a weighted, time-decaying activity filter, producing a 0-100 score. Labeled "Qualify" in the UI.
Likelihood attributesA trained model predicting the probability a profile performs a target event within a future window. Always displays as data type "Scored Tier."
Next Best Product attributesA trained model predicting which catalog product a profile is most likely to engage with next.
Default attributesDefault attributes (for events and user attributes) are automatically collected and populated when using Intempt's client-side SDKs.
Reserved attributesIntempt reserves certain attribute names (for events and user attributes) for special use cases, and these may or may not be automatically populated with values. The purpose of such Reserved attributes is for processing (e.g., event time stamping). Examples: timestamp, email, phone, name, and created.

AI-scored and calculated attribute types

The Attributes page's create menu groups five computed types under "AI Attribute Types" (each marked with a blu icon): Qualify, Calculate, Likelihood, Next Best Product, and RFM. These sit next to a separate "Manual Attribute Types" section (Text, Number, Date, Boolean) for fields with no computation, covered under Attributes.

RFM

RFM (Recency, Frequency, Monetary) scores a profile into one of 6 fixed behavioral segments based on purchase-like event data.

FieldDescription
ObjectUser or Account.
Score segmentLimits which profiles get scored. Defaults to "All users" / "All accounts."
Time windowThe lookback period for the recency/frequency/monetary calculation. Defaults to 30 days.
RecencyThe event that marks a profile's most recent activity.
FrequencyThe event counted to measure how often a profile is active.
MonetaryThe event and its numeric attribute (for example, order amount) summed to measure spend.

All three tabs (Recency, Frequency, Monetary) require an event before the attribute can be created.

SegmentDescription
ChampionsBought recently, buy often, and spend the most.
RegularsBuy on a regular basis with solid frequency.
PromisingRecent customers with average frequency.
New CustomersBought most recently but not often.
Needs AttentionAbove average recency, frequency, and monetary.
At RiskSpent big money and purchased often but long ago.

Use cases

  1. Identify Champions for a VIP loyalty offer.

  2. Re-engage At Risk high-value profiles with a win-back campaign before they churn.

  3. Route New Customers into an onboarding journey based on their RFM segment.

  4. Track how many profiles moved between segments after a campaign, using the attribute's Distribution tab.

  5. Score only a subset of your base, for example one product line's purchasers, using the Score segment field.

Qualify (Fit × Activity)

Qualify crosses a weighted profile-fit filter with a weighted, time-decaying activity filter to produce a 0-100 score. It appears as "Qualify" in the UI.

FieldDescription
ObjectUser or Account.
Score segmentLimits which profiles get scored. Defaults to "All users" / "All accounts."
Fit scoreProfile-attribute filters only (events, segments, and consents aren't available here). Weight from 1-100%.
Activity scoreEvent filters only (attributes and segments aren't available here). Weight from 1-100%, with a decay percentage per N days.
BucketFit + Activity level sumDescription
Ideal customers6 or higher (High + High)Great fit and highly engaged.
Strong5High on fit or activity.
Solid4Decent fit and engagement.
Developing3Weak on fit or activity.
Poor ICP fitBelow 3Unlikely to convert.

Use cases

  1. Score inbound leads by combining firmographic fit (industry, company size) with product usage activity.

  2. Route Ideal customers into a high-touch sales or onboarding journey.

  3. Exclude Poor ICP fit profiles from paid retargeting campaigns.

  4. Use the Activity score's decay setting so engagement reflects recent behavior rather than lifetime activity.

  5. Feed the Qualify bucket into a segment condition for sales routing rules.

Likelihood

Likelihood is a trained model that predicts the probability a profile performs a target event within a future window. It always displays as data type "Scored Tier."

FieldDescription
ObjectUser or Account.
Predict eventThe event the model is trained to predict. Required.
Prediction windowHow far into the future the model predicts, in days, weeks, or months. Defaults to 30 days.
Score segmentLimits which profiles get scored. Defaults to "All users" / "All accounts."

📘 Good to know

The Model Metrics tab shows AUC ROC, a confusion matrix, Precision/Recall/F1, and feature importance for each trained model version, with a delta chip against the previous version.

Use cases

  1. Predict which users are likely to convert in the next 30 days and target them with a conversion campaign.

  2. Predict churn risk by scoring the likelihood of a cancellation event.

  3. Compare model versions in the Model Metrics tab to confirm a retrain improved AUC ROC.

  4. Score only trial users, using the Score segment field, to focus prediction on one lifecycle stage.

  5. Use the feature importance list to see which behaviors drive the prediction.

Next Best Product

Next Best Product is a trained model that predicts which catalog product a profile is most likely to engage with next.

FieldDescription
ObjectUser or Account.
Product filtersInclude, Exclude, and Pin groups scoping which catalog products the model can recommend.
Score segmentLimits which profiles get scored. Defaults to "All users" / "All accounts."
Refresh cadenceOnce, Daily, or Weekly.

Use cases

  1. Personalize a product recommendation block with each profile's Next Best Product.

  2. Exclude out-of-stock or discontinued items with the Exclude filter.

  3. Pin a specific product, such as a new launch, so it always surfaces as a candidate.

  4. Set a Daily refresh cadence so recommendations stay current as catalog and behavior change.

  5. Compare Top-K Accuracy across model versions to confirm a retrain improved recommendation quality.

Calculate

Calculate builds an attribute from a formula over one or more event-derived components.

AggregationNeeds an attribute?
CountNo.
Count unique ofYes, string attributes only.
Sum ofYes, numeric.
Average ofYes, numeric.
Min ofYes, numeric.
Max ofYes, numeric.
Value from event attributeYes, numeric. Also shows an "Event matching order" picker (First event / Last event).

The formula defaults to A + B + ... for however many components you add, or you can write a custom expression using component letters, numbers, and + - * / ( ).

Use cases

  1. Build a lifetime order value attribute by summing a purchase event's amount attribute.

  2. Count how many times a profile triggered a specific event, using the Count aggregation.

  3. Combine two components with a custom formula, for example average order value as (total spend) / (order count).

  4. Use Count unique of to measure how many distinct products a profile has viewed.

  5. Use Value from event attribute with First event to capture a profile's very first order amount.

Default attributes

Out of the box, Intempt provides a wealth of automatically captured events and attributes to help you quickly get started with analysis and targeting. Reference the platform-specific sections below for the full list of all of the autocaptured data available when you install Intempt.

📘 How does Intempt define a session?

A session is a period of activity from a single user in your app or website. It can include many pageviews or events. On web and Android, a session ends after 30 minutes of inactivity. On iOS, a session ends after 5 minutes of inactivity, regardless of whether the app's in the background or foreground.

Note that the default events and attributes aren’t modifiable.

JavaScript SDK

Events

EventDefinition
View pageThis event is triggered every time a user opens a new page on your website.
Page leaveThis event is recorded when a user leaves a page.
Click onCapture this event every time a user clicks on a specific element, like buttons or links. It's crucial for understanding user behavior around key interactions on your site.
Change onThis event occurs whenever a user makes a change to a field in a form, such as selecting an option from a dropdown or toggling a switch. It helps track user preferences and input patterns.
Submit onTriggered when a user submits a form, this event is vital for tracking conversions and understanding how effectively your forms lead to user actions.
Session startThis event marks the beginning of a user's session on your site.
Session endRecorded when a user's session ends, either by closing the browser, expiring the session timeout, or manually logging out.

Session attributes

AttributeDescriptionExampleUser attribute
Session IDRandomly generated ID used to identify Intempt user session5801636456131197975false
ReferrerWebpage that linked to your site and initiated the session. If the user navigated directly...<https://google.com/>true
Full referrerLike "Referrer", but also contains the protocol, port, and path if they exist<https://www.google.com/search?q=handmade+leather+wallets>true
UTM sourceUTM parameter attributing the source for the user's sessionnewsletter, email, Googletrue
UTM mediumUTM parameter attributing the medium for the user's sessionsocial, display, cpm, ppc, cpc, organic, affiliate, email...true
UTM termUTM parameter attributing the keyword for the user's sessionrunning+shoes, luxury+car+rentals, organic+dog+foodtrue
UTM contentUTM parameter attributing the campaign content for the user's sessionsidebar_banner, footer_link, header_button, textlink_articletrue
UTM campaignUTM parameter attributing the campaign for the user's sessionSpring_Sale_2024, Newsletter_March_2024, Product_Launch_Q2...true
SourceWhere this event came from - which Intempt snippet or SDK the data was captured bywebfalse
Landing page queryQuery parameters of the first page of the user's session?utm_source=Google&utm_medium=CPC&utm_campaign=sell&utm_t...false
Landing page hashHash route of the first page of the user's session# install for help.intempt.com/docs#installfalse
Landing pageFirst page of the user's sessionhttps://www.google.com/url?q=https://eventsnearyou.lovable.app/&sa=D&source=docs&ust=1772717752833020&usg=AOvVaw3uZQkU2dNOlMal3ze9kagEtrue
BrowserUser's browser nameChrome 122.0.6261true
Device typeUser's device type - "Mobile", "Tablet", or "Desktop"Desktoptrue
PlatformUser's operating system nameMac OS X 12.5.1true
IP addressUser's IP address78.56.199.152true
CountryUser's country location as determined by IP addressLithuaniatrue
RegionUser's region as determined by IP addressVilniustrue
CityUser's city as determined by IP addressVilniustrue
Session duration (seconds)The duration between the "Session Start" and the "Session End" events in seconds1542false
Session Event CountThe number of events during a session15false
Session End Event NameThe original event name that triggered the Session End eventpurchasefalse

Pageview attributes

AttributeDescriptionExampleUser attribute
URLFull page URL<https://www.intempt.com/>false
DomainWebsite domain, including subdomainapp.intempt.comfalse
QueryQuery parameters of the URL, e.g., ?utm_id=1234 for intempt.com?utm_id=1234?utm_source=Google&utm_medium=CPC&utm_campaign=sell=false
PathPath of the URL following the domain pointing to a specific page, e.g., /company/companyfalse
Window widthWidth of the user's screen in pixels1920false
HashHash route of the first page of the user's session# install for help.intempt.com/docs#installfalse
TitleTitle of the pageIntemptfalse
Previous pageThe previous page visited in this session/false

Event attributes

AttributeDescriptionExampleUser attribute
Hrefhref property of the link - primarily used for clicks and anchor tags/companyfalse
Target tagTag name of the event target's DOM elementINPUT, DIV, or Afalse
Target IDID of the event target's DOM element, e.g.# loginfalse
Target classClass name of the event target's DOM elementprimary-btnfalse
Target textText of the event targetSign upfalse
HierarchyFull hierarchy of the event targetdiv#Navbar.navbar.w-nav[data-collapse=medium][data-easing2=ease][data-...false

iOS SDK

Events

EventDefinition
TouchRecorded when a user taps or touches a UI control, such as buttons, sliders, or any interactive elements.
View ScreenCaptured every time a user views a screen within the app. Useful for tracking navigation and screen engagement.
Leave ScreenLogged when a user exits a screen, useful for understanding user flows and potential interface issues.
ActionTriggered by an interaction like submitting a form or clicking a link, helps identify effective calls-to-action.
Edit FieldCaptures changes made to a form field (change recorded, not the text value), useful for form interaction data.
App Install/UpgradeTracks app installations or upgrades, capturing version and build details, key for understanding adoption rates.
Session startLogs the start of a new user session, essential for analyzing session frequency and engagement.
Session endRecorded after 5 minutes of inactivity, whether in background or foreground, for session length analysis.

Session attributes

AttributeDescriptionExampleUser attribute
Session IDRandomly generated ID used to identify Intempt user session5801636456131197975false
SourceWhere this event came from - which Intempt snippet or SDK the data was captured by: either Web, iOS or NodeJS.web, iOSfalse
DeviceUser’s device nameiPhonefalse
Device typeUser's device type - "Mobile", "Tablet", or "Desktop"Desktoptrue
PlatformUser's operating system nameMac OS X 12.5.1true
CountryUser's country location as determined by IP addressLithuaniatrue
RegionUser's region as determined by IP addressVilniustrue
CityUser's city as determined by IP addressVilniustrue
iOS vendor IDA unique identifier provided by Apple used to identify the device's vendor or maker.6D92078A-8246-4BA4-AE5B-76104861E7DCfalse
iOS advertiser IDAdvertising identifier for an iOS device. Set to all zeroes if the user has “Limit Ad Tracking” enabledAEBE52E7-03EE-455A-B3C4-E57283966239false
App nameThe name of the application as it appears on the App Store and on the home screen of the deviceMyFitnessAppfalse
App identifierA unique identifier for the app, usually in reverse DNS format, used by the system and services like notificationscom.mycompany.myfitnessappfalse
App versionVersion of the app1.4.2false
Session Start Event NameThe original event name that triggered the Session End event.purchasefalse

Pageview (screen) attributes

AttributeDescriptionExampleAttribute
ViewControllerThe name of the view controller class within the app's code that manages a portion of the user interfaceUserProfileViewControllerfalse
ViewController accessibilityIdentifierA developer-assigned identifier used in UI tests to find and interact with the UI elementsuserProfileScreenfalse
ViewController accessibilityLabelUsed by accessibility features (like VoiceOver) to describe the purpose or content of the screen or UI element to users with visual impairmentsUser Profile Screenfalse

Event attributes

AttributeDescriptionExampleAttribute
Action methodName of the action method triggered by the eventloginButtonWasPressedfalse
Target textThe text of the event targetAdd to cartfalse
Target view classClass name of an iOS action’s targetUITableCellViewfalse
Target view nameThe instance variable name of an iOS action’s targetloginButtonViewfalse
Target accessibility identifierA developer-assigned identifier used to find and interact with UI elements during UI testing. It's unique within...loginButtonfalse
Target accessibilityLabelA label used by accessibility features (like VoiceOver) to describe a UI element to users with visual impairmentsLogin Buttonfalse
HierarchyRepresents the path through the app's view hierarchy to reach a specific elementMainView > LoginForm > loginButtonfalse
Previous version codeA numerical value representing the version of the app that was previously installed on the device103false
Previous build typeDescribes the type of build (e.g., Alpha, Beta, Release) for the previous versionBetafalse
Current version codeA numerical value representing the current version of the app. This increments with each new update or release104false
Current build typeBuild type of newly upgraded application packageReleasefalse
Is upgradeTrue if this was an upgrade, false if this was an install or downgradetruefalse
App visibility stateDescribes whether the app is currently in the foreground or background. Absence of a value indicates an unknown...Foregroundfalse

Reserved attributes

Users

User attributeDescription
NameWho matches the name value you specify. First name and last name are treated as a single value.
First nameWho match any part of a first name value only.
Last nameWho matches any part of a last name value only.
EmailWho matches the email address value you specify.
Phone numberWho match the phone number value you specify.
CityWho match the city value you specify.
RegionWho matches the region value you specify.
Postal codeWho matches the postal code value you specify.
CountryWho match the country value you specify.
CreatedDate when the user was created
First seenWhose first interaction was on the relative or absolute date you specify.
Last seenWhose last interaction was on the relative or absolute date you specify.
Initial UTM campaign sourceWho matches the initial UTM campaign source value you specify.
Initial UTM campaign nameWho matches the initial UTM campaign name value you specify.
Initial UTM campaign mediumWho matches the initial UTM campaign medium value you specify.
Initial UTM campaign contentWho matches the initial UTM campaign content value you specify.
Initial UTM campaign termWho matches the initial UTM campaign term value you specify.
Initial ReferrerReferring URL when the user first arrived on your site. Defaults to "$direct" if the user is not referred.
Initial Referring DomainReferring domain at first arrival.

Accounts

Account attributeDescription
Account nameWho are linked to the account name you specify.
IndustryWho are linked to the account that has the industry value you specify.
EmployeesWho are linked to the account with the number of employees you specify.
WebsiteWho is linked to the account that has the website value you specify.
CityWho are linked to the account that has the city value you specify.
RegionWho are linked to the account that has the region value you specify.
Postal codeWho are linked to the account that has the postal code value you specify.
CountryWho are linked to the account that has the country value you specify.
CreatedWho is linked to the account record that was created.
First seenWho have the first seen value you specify and are linked to an account.
Last seenWho has the last seen value you specify and is linked to an account.

On this page