Klaviyo Product Certification cheat sheet

Klaviyo stores event data as metrics; a profile's timeline shows events, messages, and list/segment membership changes in reverse chronological order.

Profiles are keyed by email, phone, or a custom external_id; a single person can hold multiple identifiers, and merging combines duplicate profiles into one.

A metric is a named event (e.g., Placed Order) with a value and properties; flows and segments both reference metrics by name, so renaming a metric breaks references.

Catalogs store product or item data synced via feed or API; catalog items can be inserted into messages with dynamic blocks rather than static copy.

Klaviyo's built-in predictive analytics include CLV (customer lifetime value), churn risk, and next order date; each requires enough order history per profile to populate.

Sending infrastructure is shared across accounts but reputation is per sending domain; dedicated sending domains isolate deliverability from other senders.

A/B testing in Klaviyo splits recipients by a chosen variable (subject, content, send time) and picks a winner by open rate, click rate, or revenue.

Smart Sending suppresses a message to a profile that received another message within a set window (default 16 hours), preventing over-messaging across flows and campaigns.

A flow is a trigger plus an ordered series of blocks; triggers include list join, segment entry, metric (event) occurrence, price drop, and date property.

Flow filters apply to the whole flow and re-evaluate continuously; a profile that stops meeting a filter exits the flow at that point.

Conditional splits branch on profile properties, event history, or predicted values; each branch must be configured or profiles stall at the split.

Time delays can be duration-based or until a specific weekday/time; smart sending and quiet hours interact with delays to shift actual send times.

A profile can only be in a given flow once at a time by default; re-entry must be enabled explicitly and is capped by the re-entry setting.

Back-population is off by default; enabling it adds profiles who already meet the trigger condition when the flow goes live.

Flow messages inherit the flow's send channel (email, SMS, push); SMS flow messages require an SMS consent status of subscribed.

Cloning a flow copies blocks and settings but not analytics history; the clone starts with zero recipients and must be set live separately.

A list is a static, manually or form-driven collection; a segment is a dynamic, rule-based group recalculated continuously as profile data changes.

Segment definitions combine conditions with AND (all must match) or OR (any may match); nesting groups changes precedence, so read the logic tree top-down.

Segment conditions can reference profile properties, list membership, metric activity (e.g., Placed Order at least once in the last 30 days), and predictive properties.

'Has not placed order since starting checkout' style conditions use a defined time window; shortening the window shrinks the segment.

Suppression lists exclude profiles from sends; a profile on a suppression list still exists and can appear in segments.

Consent is tracked per channel: email subscription status (subscribed, never subscribed, unsubscribed) and SMS consent are independent fields.

Segment counts are approximate and refresh on a delay; a newly created segment may show zero until recalculation completes.

Sunset or engaged segments are typically built on 'Opened Email at least once in the last 60/90/180 days' to isolate active recipients for deliverability.

A campaign is a one-time send to a list or segment; it does not re-send automatically, unlike a flow triggered by an event.

Campaign send options include immediate send, scheduled send, and smart send time (optimizes per-recipient send hour based on prior open behavior).

Throttled sending spreads a campaign over a time window to protect deliverability and smooth traffic spikes.

A/B test campaigns can test up to a set number of variations and require a winning metric and a test percentage before the remainder sends.

Sign-up forms appear as popups, fly-ins, full-page, or embedded forms; each has display triggers (time delay, scroll, exit intent, URL) and targeting rules.

Form submission actions can add to a list, add to a segment via property, trigger a flow, and set consent; without a list action the profile may not be captured as subscribed.

Double opt-in requires the recipient to confirm via email before status becomes subscribed; single opt-in subscribes immediately on submit.

UTM parameters on campaign links feed attribution reports; missing UTMs cause revenue to be credited to direct traffic instead of the campaign.

Klaviyo stores event data as metrics; a profile's timeline shows events, messages, and list/segment membership changes in reverse chronological order.

Profiles are keyed by email, phone, or a custom external_id; a single person can hold multiple identifiers, and merging combines duplicate profiles into one.

A metric is a named event (e.g., Placed Order) with a value and properties; flows and segments both reference metrics by name, so renaming a metric breaks references.

Catalogs store product or item data synced via feed or API; catalog items can be inserted into messages with dynamic blocks rather than static copy.

Klaviyo's built-in predictive analytics include CLV (customer lifetime value), churn risk, and next order date; each requires enough order history per profile to populate.

Sending infrastructure is shared across accounts but reputation is per sending domain; dedicated sending domains isolate deliverability from other senders.

A/B testing in Klaviyo splits recipients by a chosen variable (subject, content, send time) and picks a winner by open rate, click rate, or revenue.

Smart Sending suppresses a message to a profile that received another message within a set window (default 16 hours), preventing over-messaging across flows and campaigns.

A flow is a trigger plus an ordered series of blocks; triggers include list join, segment entry, metric (event) occurrence, price drop, and date property.

Flow filters apply to the whole flow and re-evaluate continuously; a profile that stops meeting a filter exits the flow at that point.

Conditional splits branch on profile properties, event history, or predicted values; each branch must be configured or profiles stall at the split.

Time delays can be duration-based or until a specific weekday/time; smart sending and quiet hours interact with delays to shift actual send times.

A profile can only be in a given flow once at a time by default; re-entry must be enabled explicitly and is capped by the re-entry setting.

Back-population is off by default; enabling it adds profiles who already meet the trigger condition when the flow goes live.

Flow messages inherit the flow's send channel (email, SMS, push); SMS flow messages require an SMS consent status of subscribed.

Cloning a flow copies blocks and settings but not analytics history; the clone starts with zero recipients and must be set live separately.

A list is a static, manually or form-driven collection; a segment is a dynamic, rule-based group recalculated continuously as profile data changes.

Segment definitions combine conditions with AND (all must match) or OR (any may match); nesting groups changes precedence, so read the logic tree top-down.

Segment conditions can reference profile properties, list membership, metric activity (e.g., Placed Order at least once in the last 30 days), and predictive properties.

'Has not placed order since starting checkout' style conditions use a defined time window; shortening the window shrinks the segment.

Suppression lists exclude profiles from sends; a profile on a suppression list still exists and can appear in segments.

Consent is tracked per channel: email subscription status (subscribed, never subscribed, unsubscribed) and SMS consent are independent fields.

Segment counts are approximate and refresh on a delay; a newly created segment may show zero until recalculation completes.

Sunset or engaged segments are typically built on 'Opened Email at least once in the last 60/90/180 days' to isolate active recipients for deliverability.

A campaign is a one-time send to a list or segment; it does not re-send automatically, unlike a flow triggered by an event.

Campaign send options include immediate send, scheduled send, and smart send time (optimizes per-recipient send hour based on prior open behavior).

Throttled sending spreads a campaign over a time window to protect deliverability and smooth traffic spikes.

A/B test campaigns can test up to a set number of variations and require a winning metric and a test percentage before the remainder sends.

Sign-up forms appear as popups, fly-ins, full-page, or embedded forms; each has display triggers (time delay, scroll, exit intent, URL) and targeting rules.

Form submission actions can add to a list, add to a segment via property, trigger a flow, and set consent; without a list action the profile may not be captured as subscribed.

Double opt-in requires the recipient to confirm via email before status becomes subscribed; single opt-in subscribes immediately on submit.

UTM parameters on campaign links feed attribution reports; missing UTMs cause revenue to be credited to direct traffic instead of the campaign.

Klaviyo stores event data as metrics; a profile's timeline shows events, messages, and list/segment membership changes in reverse chronological order.

Profiles are keyed by email, phone, or a custom external_id; a single person can hold multiple identifiers, and merging combines duplicate profiles into one.

A metric is a named event (e.g., Placed Order) with a value and properties; flows and segments both reference metrics by name, so renaming a metric breaks references.

Catalogs store product or item data synced via feed or API; catalog items can be inserted into messages with dynamic blocks rather than static copy.

Klaviyo's built-in predictive analytics include CLV (customer lifetime value), churn risk, and next order date; each requires enough order history per profile to populate.

Sending infrastructure is shared across accounts but reputation is per sending domain; dedicated sending domains isolate deliverability from other senders.

A/B testing in Klaviyo splits recipients by a chosen variable (subject, content, send time) and picks a winner by open rate, click rate, or revenue.

Smart Sending suppresses a message to a profile that received another message within a set window (default 16 hours), preventing over-messaging across flows and campaigns.

A flow is a trigger plus an ordered series of blocks; triggers include list join, segment entry, metric (event) occurrence, price drop, and date property.

Flow filters apply to the whole flow and re-evaluate continuously; a profile that stops meeting a filter exits the flow at that point.

Conditional splits branch on profile properties, event history, or predicted values; each branch must be configured or profiles stall at the split.

Time delays can be duration-based or until a specific weekday/time; smart sending and quiet hours interact with delays to shift actual send times.

A profile can only be in a given flow once at a time by default; re-entry must be enabled explicitly and is capped by the re-entry setting.

Back-population is off by default; enabling it adds profiles who already meet the trigger condition when the flow goes live.

Flow messages inherit the flow's send channel (email, SMS, push); SMS flow messages require an SMS consent status of subscribed.

Cloning a flow copies blocks and settings but not analytics history; the clone starts with zero recipients and must be set live separately.

A list is a static, manually or form-driven collection; a segment is a dynamic, rule-based group recalculated continuously as profile data changes.

Segment definitions combine conditions with AND (all must match) or OR (any may match); nesting groups changes precedence, so read the logic tree top-down.

Segment conditions can reference profile properties, list membership, metric activity (e.g., Placed Order at least once in the last 30 days), and predictive properties.

'Has not placed order since starting checkout' style conditions use a defined time window; shortening the window shrinks the segment.

Suppression lists exclude profiles from sends; a profile on a suppression list still exists and can appear in segments.

Consent is tracked per channel: email subscription status (subscribed, never subscribed, unsubscribed) and SMS consent are independent fields.

Segment counts are approximate and refresh on a delay; a newly created segment may show zero until recalculation completes.

Sunset or engaged segments are typically built on 'Opened Email at least once in the last 60/90/180 days' to isolate active recipients for deliverability.

A campaign is a one-time send to a list or segment; it does not re-send automatically, unlike a flow triggered by an event.

Campaign send options include immediate send, scheduled send, and smart send time (optimizes per-recipient send hour based on prior open behavior).

Throttled sending spreads a campaign over a time window to protect deliverability and smooth traffic spikes.

A/B test campaigns can test up to a set number of variations and require a winning metric and a test percentage before the remainder sends.

Sign-up forms appear as popups, fly-ins, full-page, or embedded forms; each has display triggers (time delay, scroll, exit intent, URL) and targeting rules.

Form submission actions can add to a list, add to a segment via property, trigger a flow, and set consent; without a list action the profile may not be captured as subscribed.

Double opt-in requires the recipient to confirm via email before status becomes subscribed; single opt-in subscribes immediately on submit.

UTM parameters on campaign links feed attribution reports; missing UTMs cause revenue to be credited to direct traffic instead of the campaign.