For the complete documentation index, see llms.txt. This page is also available as Markdown.

Schema

This documentation describes the standard tables in your Bonsai data warehouse, enabling reliable access to core marketing and business data with consistent naming, data types, and performance.

Overview

These tables power Bonsai’s incrementality modeling framework by structuring media inputs and business outcomes into statistically modeled results. They serve as the source of truth for measuring true incremental lift, separating baseline performance from marketing-driven impact across channels and time.


MMM Feature Data

raw_feature_data_validation_table schema Description: The data validation table for input data versus Bonsai Incrementality feature model data.
field_name
is_nullable
data_type
definition

client_number

YES

INT64

An internal client identifier.

date

YES

DATE

The calendar date of the event.

channel

YES

STRING

The marketing channel associated with the feature.

platform

YES

STRING

The ad platform.

account_id

YES

STRING

The ad platform account ID from which a campaign derived.

campaign

YES

STRING

An organized effort to promote a product or service.

campaign_id

YES

STRING

The ad platform ID of the campaign.

feature

YES

STRING

The strategic segment of marketing activity measured in the Bonsai incrementality model.

impressions

YES

FLOAT64

Ad impressions.

clicks

YES

INT64

Ad clicks.

reach

YES

INT64

Media reach by users.

conversions

YES

FLOAT64

Media conversions, as measured by ad platform conversion data.

cost

YES

FLOAT64

Media ad cost.

MMM Results

incrementality_daily_results_table schema Description: The table of results from all live Bonsai incrementality models.
field_name
is_nullable
data_type
definition

client_number

YES

INT64

An internal client identifier.

date

YES

DATE

The calendar date of the event.

channel

YES

STRING

The marketing channel associated with the feature.

feature

YES

STRING

The strategic segment of marketing activity measured in the Bonsai incrementality model.

model_number

YES

INT64

The model number for the client in Bonsai incrementality products.

training_number

YES

INT64

The training number for the Bonsai incrementality model.

type

YES

STRING

The type of result: Base, Base+, or Incremental.

live_flag

YES

INT64

Denotes whether an Incrementality model is live in the Bonsai platform.

spend

YES

FLOAT64

Media ad cost.

incremental_dv

YES

FLOAT64

Incremental outcomes.

incremental_dv_value

YES

FLOAT64

Incremental outcome value.

incrementality_daily_predicted_actual_table schema Description: Bonsai Incrementality model results compared to actual business outcomes.
field_name
is_nullable
data_type
definition

client_number

YES

INT64

An internal client identifier.

date

YES

DATE

The calendar date of the event.

model_number

YES

INT64

The model number for the client.

training_number

YES

INT64

The training number for the model.

actual_value

YES

FLOAT64

The actual measured business outcome.

predicted_value

YES

FLOAT64

The predicted outcome from the Bonsai incrementality model.

Forecaster

forecaster_validator_table schema Description: The Budget Planner forecast validation data, comparing forecasts to actual outcomes.
field_name
is_nullable
data_type
definition

client_number

YES

INT64

An internal client identifier.

date

YES

DATE

The calendar date of the event.

year

YES

INT64

The calendar year (YYYY).

month

YES

INT64

The month of the year.

channel

YES

STRING

The marketing channel associated with the feature.

feature

YES

STRING

The strategic segment of marketing activity measured in the model.

model_number

YES

INT64

The model number for the client.

type

YES

STRING

The type of result: Base, Base+, or Incremental.

bid_ratio

YES

FLOAT64

Estimated auction position relative to historical position.

spend_forecast

YES

FLOAT64

The spend forecasted.

incremental_dv_actual

YES

FLOAT64

Incremental outcomes, actual.

incremental_dv_forecast

YES

FLOAT64

Incremental outcomes forecasted.

incremental_return_actual

YES

FLOAT64

Incremental return, actual.

incremental_return_forecast

YES

FLOAT64

Incremental return forecasted.

incremental_return_rank

YES

INT64

Incremental return volume rank (descending).

max_inc_return_forecast

YES

FLOAT64

Maximum incremental return forecasted.

min_inc_return_forecast

YES

FLOAT64

Minimum incremental return forecasted.

forecaster_table schema Description: The table of all Budget Planner forecasts, for all scenarios.
field_name
is_nullable
data_type
definition

client_number

YES

INT64

An internal client identifier.

date

YES

DATE

The calendar date of the forecast.

year

YES

INT64

The calendar year (YYYY).

month

YES

INT64

The month of the year.

weekday

YES

INT64

The calendar weekday (1 = Sunday).

channel

YES

STRING

The marketing channel associated with the feature.

feature

YES

STRING

The strategic segment of marketing activity measured in the model.

model_number

YES

INT64

The model number for the client.

forecast_number

YES

INT64

Bonsai internal ID for the Budget Planner forecast result.

type

YES

STRING

The type of result: Base, Base+, or Incremental.

Bid_Ratio

YES

FLOAT64

Estimated auction position relative to historical position.

spend_base

YES

FLOAT64

Base spend.

spend_forecast

YES

FLOAT64

Forecasted spend.

click_base

YES

FLOAT64

Base click estimate.

click_forecast

YES

FLOAT64

Forecasted click totals.

impression_base

YES

FLOAT64

Base expected impressions.

impression_forecast

YES

FLOAT64

Forecasted impressions.

reach_base

YES

FLOAT64

Base media reach.

reach_forecast

YES

FLOAT64

Forecasted media reach.

incremental_dv_base

YES

FLOAT64

Base business outcomes.

incremental_dv_forecast

YES

FLOAT64

Forecasted incremental outcomes.

incremental_dv_value_base

YES

FLOAT64

Base business outcome value.

incremental_dv_value_forecast

YES

FLOAT64

Forecasted incremental outcome value.

incremental_return_forecast

YES

FLOAT64

Forecasted incremental return.

max_return_forecast

YES

FLOAT64

Maximum incremental return forecasted across all features.

next_yr_cpc_index

YES

FLOAT64

Forecasted CPC index.

next_yr_demand_index

YES

FLOAT64

Forecasted demand index.

next_yr_value_index

YES

FLOAT64

Forecasted customer response index.

Last updated