Segment attribute reference

This page lists every attribute you can use in the segment editor’s WHERE clause. For the grammar itself (groups, AND/OR, quoting, parentheses), see the segment query language reference. For a walkthrough of building and using segments, see Customers & segments.

How to read this page

Each attribute has a type, and the type decides which operators it accepts. Rather than repeating the same list on every row, the tables below name one of these operator sets:

Set	Operators
String	=, !=, CONTAINS, NOT CONTAINS, STARTS WITH, ENDS WITH, IN (...), NOT IN (...), IS NULL, IS NOT NULL
Enum	=, !=, IN (...), NOT IN (...), IS NULL, IS NOT NULL
Number	=, !=, >, >=, <, <=, BETWEEN ... AND ..., IS NULL, IS NOT NULL
Money	Same as Number. Values are major currency units: amount_spent > 500 means more than ₹500 (or $500, and so on)
Date	=, !=, <, >, <=, >=, BETWEEN ... AND ..., IN LAST n DAYS, NOT IN LAST n DAYS, IN NEXT n DAYS, IS NULL, IS NOT NULL (units: DAYS, WEEKS, MONTHS, YEARS)
Recurring date	=, BETWEEN, IN LAST, NOT IN LAST, IN NEXT, IN MONTH May, ON May 24, IS NULL, IS NOT NULL
Boolean	= true, = false, !=, IS NULL, IS NOT NULL
Tag	CONTAINS 'x', NOT CONTAINS 'x', CONTAINS ALL (...), CONTAINS ANY (...), IS NULL, IS NOT NULL
ID	=, !=, IN (...), NOT IN (...), IS NULL, IS NOT NULL. Values are IDs, written unquoted
Category	Same operator keys as ID, but subtree semantics (see taxonomy categories)
Complex	MATCHES (...), NOT_MATCHES (...), IS NULL, IS NOT NULL (see below)

List operators (IN, NOT IN, CONTAINS ALL, CONTAINS ANY) accept up to 100 values per condition. Enum values match case-insensitively; the tables list the canonical spellings.

Number conditions are capped at a magnitude of 1e12, and money values take at most three decimal places.

Not every attribute offers IS NULL

The operator sets above are the maximum an attribute of that type can accept. IS NULL / IS NOT NULL are only offered where a value can genuinely be absent, so these always-set attributes leave them out:

number_of_orders, amount_spent, customer_added_date, customer.account_anniversary, customer_account_status, email_subscription_status, customer.tax_exempt, customer.is_walkin, customer.is_guest, customer.has_verified_email, customer.has_verified_phone, customer.was_referred, customer.is_b2b_buyer, customer.has_purchased_gift_card, customer.email_can_receive_marketing, customer.email_open_tracking_allowed, and customer.email_click_tracking_allowed.

A customer who has never ordered has number_of_orders = 0, not an empty value, so write the comparison you mean. customer_language genuinely can be unset and keeps both operators.

Open and closed enums

Most enums are closed - only the spellings listed in these tables compile. Three are open, because their options are a convenience list rather than the whole set: company.id and company.name offer at most 250 companies, and loyalty.tier offers your store’s current tiers. A value outside the offered list still compiles and saves on those three; the editor warns that it doesn’t recognise the value, which keeps queries working on large stores and across tier renames.

Dates

Date values can be written as:

• an absolute date - 2026-05-25 (a datetime like 2026-05-25T10:30:00Z also works on datetime attributes)
• a named day - today, yesterday, tomorrow
• an offset from today - -30d, -4w, -6m, -1y, +7d (days, weeks, months, years; the unit letter may be uppercase, so +7D also works)
• a duration - IN LAST 30 DAYS, NOT IN LAST 6 MONTHS, IN NEXT 2 WEEKS

Dates are calendar-checked (2026-02-30 is rejected), timezone offsets are range-checked, and windows and offsets cap at roughly a century per unit - 36,600 days, 5,300 weeks, 1,200 months, or 100 years. A BETWEEN range whose bounds invert once they are resolved in the store timezone is reported as an empty range instead of silently matching nobody.

Recurring-date attributes (anniversaries, birthdays) additionally support IN MONTH May and ON May 24, which match every year. Those two operators are exclusive to recurring dates - customer.birthday, customer.account_anniversary, customer.first_order_anniversary, and customer date / date-time metafields. On a plain date attribute such as last_order_date they are rejected at validation; use a BETWEEN range or an IN LAST window instead.

Complex attributes and MATCHES

Complex attributes (orders_placed, products_purchased, storefront.product_viewed, and so on) represent a set of records per customer - orders, order lines, events. You query them with MATCHES:

orders_placed MATCHES (count >= 3, sum_amount > 10000)

• Sub-conditions are separated by commas, which mean AND.
• Record-level subfields (date, channel, sku, …) filter which records count. Aggregate subfields (count, sum_amount, avg_rating, …) are computed over the records that pass those filters.
• MATCHES () with no sub-conditions means “has at least one record”.
• NOT_MATCHES (...) means “has no record matching these sub-conditions”.
• IS NULL on the parent means the customer has never had such a record (never placed an order, never abandoned a checkout); IS NOT NULL means they have at least one, ever.

Subfields are only valid inside their parent’s MATCHES (...) - you can’t write orders_placed.count at the top level.

Taxonomy categories

The category attributes (products_purchased.category, storefront.product_viewed.category) take a product taxonomy category key such as aa-1-13. Matching is hierarchical: a condition on a category also matches every descendant category, so category = 'aa-1' (Clothing) matches products in any clothing subcategory. IS NULL matches uncategorized products.

Note

Attributes marked Sellerlane-only in the tables below have no Shopify equivalent - keep that in mind if you’re porting segment definitions from a Shopify store.

Customer profile

Basic identity, account, and lifecycle facts about the customer record.

Attribute	Type	Operators	Notes
customer.email	text	String	Customer email address.
customer.phone	text	String	Customer phone number.
customer.first_name	text	String
customer.last_name	text	String
customer_account_status	enum	Enum	ENABLED, DISABLED, INVITED, DECLINED.
customer_added_date	date	Date	Date the customer record was created.
customer.email_verified_at	date	Date	Empty until the email is verified.
customer.phone_verified_at	date	Date	Empty until the phone is verified.
customer_email_domain	text	String	Domain part of the email, e.g. gmail.com. = and IN match case-insensitively; the editor suggests the domains your customers actually use.
customer_tags	tag list	Tag	Tags on the customer profile.
customer.tax_exempt	boolean	Boolean
customer_language	text	String	Preferred language/locale.
customer.login_channel	enum	Enum	email, sms.
customer.last_login_at	datetime	Date
customer.is_guest	boolean	Boolean	True when the customer has no verified email or phone.
customer.has_verified_email	boolean	Boolean
customer.has_verified_phone	boolean	Boolean
customer.account_anniversary	recurring date	Recurring date	Yearly anniversary of the signup date.
customer.first_order_anniversary	recurring date	Recurring date	Yearly anniversary of the first order.
customer.birthday	recurring date	Recurring date	Recurring month/day, evaluated in the store timezone.
customer.is_walkin	boolean	Boolean	POS walk-in record with no contact details. Sellerlane-only.
customer.is_b2b_buyer	boolean	Boolean	The customer holds an active company (wholesale) membership.
company.id	enum	Enum	The company the customer is an active member of. Open enum - the picker lists at most 250 companies.
company.name	enum	Enum	Name spelling of the same membership. Open enum.

customer_tags CONTAINS 'vip' AND customer.has_verified_email = true
AND customer.birthday IN NEXT 7 DAYS

Location

Location attributes match against any of the customer’s saved addresses. For radius targeting, see Geo distance.

Attribute	Type	Operators	Notes
customer_countries	text	String	ISO country code across the customer’s addresses, e.g. IN.
customer_regions	text	String	States/provinces, stored as a country-region composite, e.g. in-tn.
customer_cities	text	String	Cities, stored as a country-region-city composite with punctuation stripped, e.g. in-tn-chennai.
customer.postal_code	text	String	Postal codes across the customer’s addresses.

Region and city values are composites, not the plain names - a city is matched as in-tn-chennai, never as Chennai. You don’t have to construct them: all four attributes suggest values from your store’s own addresses, and the dropdown shows the readable label (Chennai, TN, IN) while inserting the composite the query needs. See value suggestions.

customer_countries = 'IN' AND customer_cities IN ('in-mh-mumbai', 'in-mh-pune')

Spending & order stats

Lifetime aggregates over the customer’s order history. Money values are major currency units.

Attribute	Type	Operators	Notes
amount_spent	money	Money	Customer-profile spend across paid-ish, uncanceled orders after raw payment refunds. Includes tax, shipping, and money paid to purchase a gift card; it is not analytics Total sales.
number_of_orders	number	Number	Count of the same paid-ish, uncanceled, non-test customer-profile orders.
average_order_value	money	Money	Average customer-profile spend per paid order (amount_spent ÷ number_of_orders). Despite the stable query key, this is not the analytics Average order value metric.
first_order_date	date	Date	Empty for customers who have never ordered.
last_order_date	date	Date	Empty for customers who have never ordered.

amount_spent > 500 AND number_of_orders >= 3
AND last_order_date NOT IN LAST 90 DAYS

The editor labels average_order_value as Average spend per paid order to avoid confusing it with analytics AOV. Analytics AOV is order-time merchandise after discounts divided by eligible non-gift-card orders and deliberately excludes tax, shipping, gift-card issuance, and later refunds. See Financial semantics.

Predictions & RFM

Recency/frequency/monetary scoring from the latest analytics snapshot, plus spend predictions. The three prediction attributes only appear in the editor once your store has more than 100 qualifying orders and a generated prediction snapshot. RFM is lifetime-scoped: chronology uses eligible order placement, while monetary value uses event-net total sales through the snapshot. See RFM and actual lifetime value.

Attribute	Type	Operators	Notes
rfm_group	enum	Enum	DORMANT, AT_RISK, PREVIOUSLY_LOYAL, NEEDS_ATTENTION, ALMOST_LOST, LOYAL, PROMISING, ACTIVE, NEW, CHAMPIONS, PROSPECTS.
rfm.recency_days	number	Number	Store-local days since the latest eligible order was placed.
rfm.frequency	number	Number	Lifetime eligible orders placed through the snapshot.
rfm.monetary	money	Money	Lifetime event-net total sales through the snapshot, floored at zero per customer.
rfm.r_score	number	Number	Recency score.
rfm.f_score	number	Number	Frequency score.
rfm.m_score	number	Number	Monetary score.
rfm.fm_score	number	Number	Combined frequency-monetary score.
rfm.rfm_score	text	String	Combined score string, e.g. 543.
rfm.first_order_date	date	Date	First order date in the snapshot.
rfm.last_order_date	date	Date	Last order date in the snapshot.
predicted_spend_tier	enum	Enum	HIGH, MEDIUM, LOW.
predicted_ltv	money	Money	Predicted lifetime value. Sellerlane-only.
prediction_confidence	number	Number	Model confidence, 0 to 1. Sellerlane-only.

rfm_group IN ('CHAMPIONS', 'LOYAL') OR predicted_spend_tier = 'HIGH'

Marketing consent

Email marketing consent and tracking preferences. See the consent guide for what each state means.

Attribute	Type	Operators	Notes
email_subscription_status	enum	Enum	SUBSCRIBED, UNSUBSCRIBED, NOT_SUBSCRIBED, PENDING, INVALID, REDACTED.
customer.email_can_receive_marketing	boolean	Boolean	The effective “may I email this person” flag.
customer.email_open_tracking_allowed	boolean	Boolean
customer.email_click_tracking_allowed	boolean	Boolean
customer.marketing_consent_updated_at	datetime	Date	When consent last changed.

email_subscription_status = 'SUBSCRIBED'
AND customer.marketing_consent_updated_at > -30d

Orders placed

orders_placed is a complex attribute over the customer’s orders. Use MATCHES with the subfields below; orders_placed IS NULL means the customer has never placed an order.

Subfield	Type	Operators	Notes
date	date	Date	Order creation date.
paid_date	date	Date
canceled_date	date	Date
closed_date	date	Date
count	number	Number	Number of matched orders.
amount	money	Money	A single matched order’s total.
sum_amount	money	Money	Sum of matched order totals.
sum_net_sales	money	Money
sum_gross_sales	money	Money
sum_discount	money	Money
sum_refund	money	Money
average_order_amount	money	Money	Average across matched orders.
min_order_amount	money	Money
max_order_amount	money	Money
location_id	ID	ID	POS/store location of the order.
channel	enum	Enum	online, pos.
financial_status	enum	Enum	pending, authorized, partially_paid, paid, partially_refunded, refunded, voided.
fulfillment_status	enum	Enum	unfulfilled, partially_fulfilled, fulfilled.
lifecycle_status	enum	Enum	open, closed, canceled.
shipping_country	enum	Enum	ISO country code, free input.
shipping_region	text	String
shipping_city	text	String
billing_country	enum	Enum	ISO country code, free input.
billing_region	text	String
billing_city	text	String
used_store_credit	boolean	Boolean
store_credit_amount	money	Money	Store credit applied to the order.
used_gift_card	boolean	Boolean
gift_card_amount	money	Money	Gift card amount applied to the order.
is_first_order	boolean	Boolean	True on the customer’s first order.
customer_order_index	number	Number	Order sequence number for the customer (1 = first).
payment_method_kind	enum	Enum	gateway, manual. Sellerlane-only.
gateway_provider	enum	Enum	stripe, razorpay, payu, adyen, xendit, paystack, mercadopago, tap, twoctwop, flutterwave, airwallex, cashfree. Sellerlane-only.
payment_amount_type	enum	Enum	Payment rule applied at checkout: full, percentage, fixed (deposits), cod. Sellerlane-only.
is_cod	boolean	Boolean	Cash on delivery. Sellerlane-only.
has_discount	boolean	Boolean	Any discount applied.
discount_code	text	String	Discount code used on the order.
has_refund	boolean	Boolean
has_return	boolean	Boolean	Returned units or return value exist.
units_returned	number	Number
metafield.<namespace>.<key>	metafield	By value type	Order metafield condition - see Custom fields.

orders_placed MATCHES (count >= 3, sum_amount > 10000, date IN LAST 6 MONTHS)
AND orders_placed NOT_MATCHES (is_cod = true, has_return = true)

Products purchased

Three complex attributes cover the line-item view of order history: products_purchased (by product), ordered_variant (by variant), and ordered_collection (by collection membership of purchased products). products_purchased IS NULL means the customer has never bought anything.

products_purchased

Subfield	Type	Operators	Notes
date	date	Date	Order date of the matched line.
count	number	Number	Number of matched order lines.
quantity	number	Number	Quantity on a single matched line.
sum_quantity	number	Number	Total quantity across matched lines.
sum_line_amount	money	Money	Total net sales from matched lines.
id	ID	ID	The purchased product.
variant_id	ID	ID	The purchased variant.
sku	text	String
vendor	text	String
product_type	text	String	The merchant-defined product type label.
tag	tag list	Tag	Product tags at order time.
category	category	Category	Taxonomy category, includes all subcategories.
product_kind	enum	Enum	Platform kind: physical, digital, gift_card, event, bundle. Sellerlane-only.
brand_id	ID	ID
metafield.<namespace>.<key>	metafield	By value type	Product metafield condition.

ordered_variant

Subfield	Type	Operators	Notes
date	date	Date
count	number	Number	Number of matched lines.
quantity	number	Number	Total purchased quantity.
sum_line_amount	money	Money
variant_id	ID	ID
product_id	ID	ID	Parent product.
sku	text	String
metafield.<namespace>.<key>	metafield	By value type	Variant metafield condition.

ordered_collection

Subfield	Type	Operators	Notes
date	date	Date
count	number	Number	Matched lines whose product is in the collection.
collection_id	ID	ID
metafield.<namespace>.<key>	metafield	By value type	Collection metafield condition.

products_purchased MATCHES (category = 'aa-1-13', date IN LAST 90 DAYS)
OR ordered_variant MATCHES (sku STARTS WITH 'TSHIRT-')

Abandoned checkout

One scalar plus the abandoned_checkout complex attribute over the customer’s abandoned carts/checkouts.

Attribute	Type	Operators	Notes
abandoned_checkout_date	date	Date	Date of the customer’s abandoned checkout.
abandoned_checkout	complex	Complex	Subfields below.

Subfield	Type	Operators	Notes
date	date	Date	When the cart was abandoned.
count	number	Number	Number of abandoned checkouts.
amount	money	Money	Cart total.
line_count	number	Number	Lines in the cart.
item_quantity	number	Number	Items in the cart.
channel	enum	Enum	online, pos.
recovered	boolean	Boolean	Whether the checkout later converted to an order.
recovered_date	date	Date
used_store_credit	boolean	Boolean
store_credit_amount	money	Money
used_gift_card	boolean	Boolean
gift_card_amount	money	Money

abandoned_checkout MATCHES (date IN LAST 7 DAYS, recovered = false, amount > 1000)

Store credit

store_credit_accounts matches on the customer’s store credit account state. store_credit_accounts IS NULL means the customer has never had a store credit account.

Subfield	Type	Operators	Notes
balance	money	Money	Current balance.
currency	enum	Enum	ISO currency code, free input.
expiring_amount	money	Money	Portion of the balance that has an expiry date.
next_expiry_date	date	Date
last_credit_date	date	Date	Last time credit was issued.
last_debit_date	date	Date	Last time credit was spent.
issued_amount	money	Money	Lifetime credit issued.
redeemed_amount	money	Money	Lifetime credit redeemed.
expired_amount	money	Money	Lifetime credit expired.
ledger_count	number	Number	Ledger entries on the account.

store_credit_accounts MATCHES (balance > 0, next_expiry_date IN NEXT 30 DAYS)

Gift cards

Sellerlane-only. gift_cards matches on gift cards assigned to the customer (cards they hold and can spend), plus one purchase-side boolean.

Attribute	Type	Operators	Notes
gift_cards	complex	Complex	Subfields below.
customer.has_purchased_gift_card	boolean	Boolean	The customer has bought a gift card, for themselves or as a gift.

Subfield	Type	Operators	Notes
balance	money	Money	Remaining spendable balance across active, unexpired cards (net of checkout reservations).
count	number	Number	Number of matched cards.
state	enum	Enum	active, disabled, expired.
expiry_date	date	Date	IN NEXT 30 DAYS makes an expiring-soon winback.
last_used_date	date	Date	Last redemption against an order.

gift_cards MATCHES (state = 'active', expiry_date IN NEXT 30 DAYS)

Loyalty

Sellerlane-only. loyalty matches on the customer’s loyalty account. loyalty IS NULL means the customer has never had a loyalty account.

Subfield	Type	Operators	Notes
points_balance	number	Number	Current redeemable points.
tier	enum	Enum	Your store’s tier names. An open enum: the editor offers your current tiers, and a value outside that list still compiles (with a warning), so a segment survives a tier rename.
status	enum	Enum	enrolled, paused, excluded, merged.
enrolled_at	date	Date
last_earned_at	date	Date
last_redeemed_at	date	Date
lifetime_earned_points	number	Number
lifetime_redeemed_points	number	Number
next_expiration_date	date	Date	Next points expiry - good for expiring-points winbacks.

loyalty MATCHES (status = 'enrolled', points_balance >= 500,
  next_expiration_date IN NEXT 14 DAYS)

Reviews

Sellerlane-only. product_review matches on reviews the customer has submitted.

Subfield	Type	Operators	Notes
date	date	Date	When the review was submitted.
count	number	Number	Number of matched reviews.
rating	number	Number	Star rating on a matched review (1-5).
avg_rating	number	Number	Average rating across matched reviews.
verified	boolean	Boolean	Verified purchase.
has_photo	boolean	Boolean	Review includes a visible photo.
has_video	boolean	Boolean	Review includes a visible video.
product_id	ID	ID	The reviewed product.

product_review MATCHES (rating >= 4, has_photo = true, date IN LAST 12 MONTHS)

Referrals

Sellerlane-only. referrals_made matches referrals where the customer is the referrer; customer.was_referred covers the other direction.

Attribute	Type	Operators	Notes
referrals_made	complex	Complex	Subfields below.
customer.was_referred	boolean	Boolean	The customer joined through someone else’s referral.

Subfield	Type	Operators	Notes
date	date	Date	When the referral was attributed.
count	number	Number	Number of matched referrals.
status	enum	Enum	pending, qualified, rewarded, disqualified.

referrals_made MATCHES (status = 'rewarded') AND customer.was_referred = false

Event tickets

Sellerlane-only. event_tickets matches ticket reservations from the customer’s orders for event products.

Subfield	Type	Operators	Notes
start_date	date	Date	Event start. IN NEXT finds upcoming attendees; IN LAST is a past-attendance proxy.
product_id	ID	ID	The event product.
quantity	number	Number	Tickets on a matched reservation.
canceled	boolean	Boolean	Reservation was canceled.
count	number	Number	Number of matched reservations.

event_tickets MATCHES (start_date IN NEXT 30 DAYS, canceled = false)

Back-in-stock requests

Sellerlane-only. stock_notifications matches the customer’s back-in-stock requests. Only signed-in requests can match - guest email-only requests have no customer record.

Subfield	Type	Operators	Notes
product_id	ID	ID	Product the customer is waiting on.
variant_id	ID	ID	Variant the customer is waiting on.
status	enum	Enum	active, notified, canceled, expired.
date	date	Date	When the request was made.
count	number	Number	Number of matched requests.

stock_notifications MATCHES (status = 'active', date IN LAST 30 DAYS)

Self-service requests

Sellerlane-only. self_service_requests matches the customer’s buyer self-service requests (cancellations, returns, redeliveries, and so on).

Subfield	Type	Operators	Notes
type	enum	Enum	order_cancel, return, event_reschedule, event_cancel, digital_redeliver, digital_redownload, gift_card_resend.
status	enum	Enum	requested, approved, in_transit, received, completed, declined, withdrawn, expired.
date	date	Date	When the request was submitted.
count	number	Number	Number of matched requests.

self_service_requests MATCHES (type = 'return', status = 'completed',
  date IN LAST 6 MONTHS)

Email engagement

Eight complex attributes, one per email event. All share the same subfields; email.clicked adds a click-specific one. IS NULL on any of them means the event has never happened for the customer. The older shopify_email.* spelling is a legacy alias — saved segments using it keep validating, but new queries should use email.*.

Attribute	Meaning
email.sent	A marketing email was sent.
email.delivered	An email was delivered.
email.opened	An email was opened.
email.clicked	A link in an email was clicked.
email.bounced	An email bounced.
email.failed	An email failed to send.
email.unsubscribed	The customer unsubscribed from an email.
email.marked_as_spam	The customer reported an email as spam.

Subfield	Type	Operators	Notes
date	date	Date	When the event happened.
count	number	Number	Number of matched events.
campaign_id	ID	ID	The campaign.
activity_id	ID	ID	The activity: a campaign or one specific standalone/automation email — matching either id. The editor suggests both, emails shown by subject.
link_url	text	String	Clicked link URL. email.clicked only; the editor suggests URLs that have actually been clicked.

email.delivered MATCHES (date IN LAST 90 DAYS)
AND email.opened IS NULL

Note

message_id and tracked_link_id were removed from this catalog. Both were internal per-recipient identifiers with no picker and no way for you to know a value, so they could never produce a useful condition - use campaign_id or activity_id to scope by send, and link_url to scope by link. Segments saved earlier that still reference the old keys keep validating, so nothing you built breaks.

Storefront behavior

Thirteen complex attributes over the customer’s storefront analytics events. Events only attribute to a customer when the visit is identified (logged in or recognized), and consent settings apply.

Attribute	Event
storefront.product_viewed	Viewed a product page.
storefront.collection_viewed	Viewed a collection page.
page_viewed	Viewed any storefront page.
product_added_to_cart	Added a product to the cart.
product_removed_from_cart	Removed a product from the cart.
cart_viewed	Viewed the cart.
checkout_started	Started checkout.
checkout_completed	Completed checkout.
search_submitted	Submitted a storefront search.
discount_code_entered	Entered a discount code.
discount_code_applied	A discount code was applied.
discount_code_rejected	A discount code was rejected.
payment_failed	A payment attempt failed.

All thirteen share these subfields - and the nine generic events have only these:

Subfield	Type	Operators	Notes
date	date	Date	Event date.
count	number	Number	Number of matched events.
path	text	String	Storefront path, e.g. /collections/sale. The editor suggests paths from recent traffic.
campaign_id	text	String	Free-text client tracking value from the visit’s utm_campaign-style parameters. Not a marketing campaign ID - for that, use the email engagement attributes.
message_id	text	String	Free-text client tracking value from the visit’s tracking parameters, on the same footing as campaign_id.
value_amount	money	Money	Event value, where applicable.

The three product-context events (storefront.product_viewed, product_added_to_cart, product_removed_from_cart) additionally support:

Subfield	Type	Operators	Notes
id	ID	ID	The product.
variant_id	ID	ID	The variant.
tag	tag list	Tag	The product’s current tags.
category	category	Category	The product’s current taxonomy category, includes all subcategories.
metafield.<namespace>.<key>	metafield	By value type	Product metafield condition.

storefront.collection_viewed additionally supports id (the collection) and metafield.<namespace>.<key> (collection metafield).

Those extras are scoped to the events that have them. The product parameters exist only on the three product events, the collection parameters only on storefront.collection_viewed, and the remaining nine events (page_viewed, cart_viewed, checkout_started, checkout_completed, search_submitted, the three discount_code_* events, and payment_failed) accept nothing beyond the six shared subfields. Writing checkout_completed MATCHES (id = '…') is an error you see while typing, not a condition that quietly matches nobody.

There are no search_query or discount_code subfields: the analytics pipeline doesn’t store the raw query or code, so those can’t be filtered on.

storefront.product_viewed MATCHES (category = 'aa-1-13', date IN LAST 14 DAYS)
AND products_purchased IS NULL

Geo distance

customer_within_distance matches customers whose identified storefront visit places them within a radius of a coordinate. It uses a MATCHES form with fixed parameters rather than subfields:

customer_within_distance MATCHES (coordinates = (13.08, 80.27), distance_km = 25)

Use distance_km or distance_mi. Latitude runs -90 to 90, longitude -180 to 180, and the distance must be positive. IS NULL / IS NOT NULL match customers with no known / any known visit location.

Custom fields (metafields)

Any metafield definition with Use as filter enabled becomes a segment attribute, as long as its value type is filterable. Where it appears depends on the owner type:

Owner type	Where you use it	Written as
Customer	Top level	metafields.<namespace>.<key>
Order	Inside orders_placed MATCHES (...)	metafield.<namespace>.<key>
Product	Inside products_purchased, storefront.product_viewed, product_added_to_cart, product_removed_from_cart	metafield.<namespace>.<key>
Variant	Inside ordered_variant MATCHES (...)	metafield.<namespace>.<key>
Collection	Inside ordered_collection, storefront.collection_viewed	metafield.<namespace>.<key>

Note the spelling difference: plural metafields. at the top level, singular metafield. inside a MATCHES.

Value types and operators

The metafield’s value type decides the operator set:

Value type	Operators	Notes
Single-line text, multi-line text, URL	String	= and IN are case-sensitive for text metafields.
Color	String	Compared case-insensitively.
Integer, decimal, rating	Number	Rating comparisons are checked against the definition’s min/max scale.
Money	Money	Major currency units.
Date, date-time	Date	Customer date metafields also support the recurring operators IN MONTH and ON <month day> - ideal for birthdays and renewal dates.
Boolean	Boolean
Dimension, volume, weight	Number + unit	Write the unit after the value: metafield.specs.weight > 2 kg. Units - dimension: mm, cm, m, in, ft, yd; volume: ml, cl, l, m3, us_fl_oz, us_pt, us_qt, us_gal, imp_fl_oz, imp_pt, imp_qt, imp_gal; weight: g, kg, lb, oz. Comparison is unit-normalized, so 2 kg matches a value stored as 2000 g.
References (product, variant, collection, customer, page, blog post, location, file, metaobject)	ID	Compare by referenced ID.
JSON, rich text, link	Not filterable	These types never appear in the segment editor.

List metafields

List metafields compare at the owner level - the condition looks at the whole list:

• =, CONTAINS, and other positive operators match when any element matches.
• Negative operators (!=, NOT CONTAINS, NOT IN) match only when no element matches.
• Text and number lists additionally offer CONTAINS ANY (...) (at least one of the given values is an element) and CONTAINS ALL (...) (every given value is an element), like tags.

When a definition is deleted

If you delete a metafield definition that segments reference, Sellerlane removes those conditions from the affected segments automatically, refreshes their membership, and records a note on each segment naming the deleted field. A segment whose rules end up empty matches no one until you edit it. The note clears the next time you save the segment’s rules.

metafields.custom.fit_preference = 'slim'
AND orders_placed MATCHES (metafield.custom.gift_wrap = true)

See also

• Segment query language - grammar, grouping, and editor behavior.
• Customers & segments - building, previewing, exporting, and using segments.
• Metafields - defining the custom fields you segment on.