Output tables

The module produces two intermediate touchpoint tables, two core tables (journeys and models), and six reports. Journeys, models and reports materialize in your outputs dataset (superform_outputs_*); the intermediate touchpoint tables materialize in the transformations dataset (superform_transformations_*).

Intermediate touchpoints

You normally consume the journeys table and the reports, not these. Only one is built, depending on touchpoint.mode.

Table	Granularity	Purpose
int_ga4_attribution_touchpoints_session	One row per traffic touch (one per ga4_sessions row) plus one row per matched conversion event	The flat touchpoint stream in session mode.
int_ga4_attribution_touchpoints_source_change	One row per intra-session source run plus one row per matched conversion event	The flat touchpoint stream in source_change mode.

ga4_attribution_journeys

Granularity: one row per user journey (journey_id), from first touch to its closing conversion or boundary. Traffic and conversion touches are stored as nested arrays.

Column	Description
journey_id	Unique ID - `FARM_FINGERPRINT(user_identifier
user_identifier	Identity key value (user_pseudo_id by default, else journey.user_identifier coalesced to user_pseudo_id).
user_identifier_type	Name of the identity-key column (e.g. user_pseudo_id).
journey_number	0-based sequence of this journey in the user's history.
journey_start_date	Date of the first touch (partition column).
journey_start_timestamp_utc	UTC timestamp of the first touch.
journey_end_date	Date of the last touch.
journey_length_days	Elapsed days, first to last touch.
days_to_conversion	Days from first touch to the closing conversion; NULL for ongoing/lapsed journeys.
user_pseudo_ids	Distinct user_pseudo_id values in the journey.
device_categories	Distinct device categories across traffic touches.
default_channel_groups	Distinct default channel groups across traffic touches.
ends_in_sale	TRUE if the journey has a closing conversion.
journey_length_steps	Number of traffic (non-conversion) touches.
revenue	Revenue across ALL conversion touches (closing + micro). Does not reconcile to models.attributed_revenue (closing only) - use closing_revenue for that.
closing_conversion_count	Closing conversion touches (0 or 1 by construction).
micro_conversion_count	Non-closing (micro) conversion touches.
closing_revenue	Revenue from closing conversions only (0 when none).
micro_revenue	Revenue from non-closing (micro) conversions (0 when none).
closing_conversion_date	Closing conversion date; NULL for ongoing/lapsed. Reports use this as conversion_date.
closing_conversion_name	Closing conversion name; NULL for ongoing/lapsed.
touches	Array of traffic touches, chronological (empty for a zero-touch conversion). Per-touch fields: touch_start_time, touch_id, session_id, user_pseudo_id, device_category, source, medium, campaign, default_channel_group.
conversions	Array of conversion touches (closing + non-closing), chronological. Per-conversion fields: touch_start_time, event_id, session_id, user_pseudo_id, conversion_name, revenue, is_closing.
is_ongoing	TRUE if no boundary closed the journey (still accumulating).
is_final	Data-freshness flag - TRUE once all touches are past the 72h late-data window.

ga4_attribution_models

Granularity: one row per credited traffic touch × model × closing conversion, aggregated to touch_date × conversion_date × closing_conversion_name × model_name × channel dimensions. Additive - see Attribution models.

Column	Description
touch_date	Date of the credited traffic touch (partition column).
conversion_date	Date of the closing conversion the credit is for.
closing_conversion_name	Closing conversion name - segment by conversion type ('(no touchpoint)' on the synthetic zero-touch row).
model_name	Attribution model. participation overcounts - compare fractional models only.
source / medium / campaign / default_channel_group / device_category	Channel dimensions of the touch ('(no touchpoint)' for a zero-touch journey).
attributed_revenue	Closing revenue apportioned to this touch (additive). SUM reconciles to closing revenue per fractional model; participation overcounts.
attributed_conversions	Share of the closing conversion on this touch (fractional for standard models, 1.0 each for participation; additive).
credited_touches	Credited touch-rows at this grain (pre-aggregation, not distinct touch_ids). position_based with 2 touches (no middle) renormalizes to 50/50.

Reports

All six reports are additive and partitioned on journey_start_date. They derive from the journeys table; store counts and SUM numerators, and compute rates/averages in BI (see Additivity).

ga4_attribution_channel_chain_report

Granularity: one row per journey_start_date × conversion_date × channel_chain × closing_conversion_name. The full ordered channel path and its outcomes.

Column	Description
channel_chain	The path - ordered >-delimited channels; part of the grain.
path_length	Channels in the path (constant per channel_chain).
path_length_bucket	path_length bucketed: 1, 2, 3, 4-5, 6+.
first_channel / last_channel	First / last channel in the path.
num_distinct_channels	Distinct channels in the path (path_length counts repeats; this does not).
journeys	Journeys on this path (additive).
closed_journeys / ongoing_journeys / lapsed_journeys	Journeys that closed / still open / closed without a sale.
no_touch_journeys	Journeys with zero traffic touches (chain = '(no touchpoint)').
sum_traffic_touches	SUM of traffic touches (additive).
total_conversion_count	SUM of all conversion touches, closing + micro (additive).
closing_conversion_count / micro_conversion_count	SUM of closing / non-closing conversions.
total_revenue / micro_revenue	SUM of closing-conversion / micro-conversion revenue.
sum_journey_length_days	SUM of journey length in days.
sum_days_to_conversion	SUM of first-touch-to-conversion days, closed journeys.

ga4_attribution_journey_summary_report

Granularity: one row per journey_start_date × conversion_date × path_length × closing_conversion_name. Journey metrics by path length.

Column	Description
path_length	Channels in the path (the grain; raw integer - bucket in BI).
path_length_bucket	Convenience bucket over path_length (1, 2, 3, 4-5, 6+); descriptor, not grain.
journeys	Journeys at this path length (additive).
closed_journeys / ongoing_journeys / lapsed_journeys	conversion_rate = closed_journeys / journeys.
no_touch_journeys	Zero-traffic-touch journeys (fall in path_length 1).
total_conversion_count / closing_conversion_count / micro_conversion_count	Additive conversion counts.
total_revenue / micro_revenue	SUM of closing / micro revenue.
sum_journey_length_days / sum_days_to_conversion	Additive day sums for derived averages.

ga4_attribution_channel_assist_report

Granularity: one row per journey_start_date × conversion_date × channel × closing_conversion_name. Per-channel participation and assist - an influence view (values double-count across the channels of a journey; for credit use the models table).

Column	Description
channel	The participating channel (a value of the configured journey.channel_grouping).
journeys_participated	Journeys whose path includes this channel (any position; once per journey, overlaps across channels).
journeys_as_first / journeys_as_last	Journeys where this channel is first / last (the closing channel).
journeys_assisted	Journeys where this channel participated but was not the closing channel.
closed_journeys_participated / _as_first / _as_last / _assisted	The same four counts restricted to closing journeys.
participated_revenue	Closing revenue of journeys this channel participated in (influence, not credit).
assisted_revenue	Closing revenue of journeys this channel assisted (not the closing channel).
micro_conversion_count / micro_revenue / closing_conversion_count	Conversions/revenue in journeys this channel participated in (influence; double-counts).

ga4_attribution_channel_transition_report

Granularity: one row per journey_start_date × conversion_date × from_channel × to_channel. Sankey/flow edges between consecutive channels.

Column	Description
from_channel	Source node - a channel, or '(start)'.
to_channel	Destination node - a channel, or a terminal node: '(conversion: <name>)' if closed, else '(no conversion)'.
transitions	Times this consecutive pair occurs (edge weight; repeats within a journey counted).
closed_transitions	Transitions within journeys that ended in a closing conversion.
journeys_with_transition	Distinct journeys containing this edge (vs transitions, which counts repeats).
closed_journeys_with_transition	Distinct closing journeys containing this edge.

ga4_attribution_conversion_lag_report

Granularity: one row per journey_start_date × conversion_date × closing_conversion_name × lag_bucket (closed journeys only). Time-to-conversion distribution.

Column	Description
lag_bucket	days_to_conversion bucketed: 0 (same day), 1 day, 2-7 days, 8-30 days, 31+ days (the grain).
lag_bucket_order	Integer sort key (1-5) for chronological chart ordering.
closed_journeys	Closing journeys in this bucket.
closing_conversion_count	SUM of closing conversions in this bucket (= closed_journeys when one closing per journey).
revenue	SUM of closing-conversion revenue in this bucket.
sum_days_to_conversion	SUM of first-touch-to-closing days (additive).

ga4_attribution_micro_conversion_report

Granularity: one row per journey_start_date × conversion_name, where conversion_name is a micro-conversion (a conversions[] entry not in journey.closing_conversions). Micro-conversion volume and its lift on final outcomes.

Column	Description
conversion_name	The micro-conversion (the grain).
event_count	Occurrences of this micro-conversion (additive).
event_revenue	Revenue on this micro-conversion's own touches (additive).
journeys_with	Journeys containing this micro-conversion. conv_rate_with = closed_journeys_with / journeys_with.
closed_journeys_with	Journeys with this micro-conversion that reached a closing conversion - compare conv_rate_with to the overall rate for lift.
closing_conversion_count_with	Closing conversions in journeys containing this micro-conversion (additive).
closing_revenue_with	Closing revenue of journeys containing this micro-conversion (final outcome, not the micro's own revenue). Summing across conversion_name double-counts a journey.