Skip to main content

Snowflake Data Transfer Overview

Set up data transfers via Snowflake Share

Updated today

Overview

JamLoop’s Platform gives customers flexible options to access and analyze campaign data in the way that best fits their reporting, attribution and BI workflows. Customers can either configure custom aggregate reports directly in the platform, or set-up dynamic data transfers via Snowflake Share.

Summary of Export Types

There are three main types of data transfer:

  1. Aggregate

    1. Grouped and summarized views of campaign performance across chosen dimensions

    2. May or may not include attribution

  1. Log-Level Delivery (*Snowflake Share only)

    1. Granular, impression-level reporting showing every ad delivered.

    2. Can be used to calculate metrics such as household reach and frequency

  1. Log-Level Attribution (*Snowflake Share only)

    1. Granular conversion-focused reporting that ties conversions back to attributed impressions

Each export type is designed for different levels of transparency and analytical depth. By aligning the transfer type with the customer use cases, JamLoop ensures brands can validate performance, power internal modeling, and integrate seamlessly into BI environments.

Delivery Methods

  • Direct Download (*aggregate reports only)

    • Accessible directly from JamLoop’s reporting interface.

    • Best for ad hoc or smaller-scale exports

    • Custom Reports are emailed (aggregates only) and include a download link

    • Reports can be sent directly to customer inboxes

    • Useful for lightweight, human-readable exports and when multiple team members need regular access.

    • Exports with more than 100k rows are not supported by an email export

  • Snowflake Share

    • Aggregate & Log-Level options

    • Data delivered directly into the customer’s Snowflake environment

    • Eliminates file transfers and enables near-real-time access for advanced analytics teams.

    • Requires customer to setup access ahead of time

    • In some cases, JamLoop may support direct integrations with a BI tool so long as the BI system has an integration and/or can receive data from Snowflake’s cloud-based warehouse

Scheduling & Frequency

  • Frequency Options

    • Custom Reports

      • Daily

      • Weekly

      • Monthly

    • Snowflake Share

      • Dynamic / Real-Time

  • Timing Cadence

    • Exports are recommended to be delivered during off-peak hours (e.g. early morning UTC/local time) to capture completed activity from the prior period, and/or account for attribution which processes daily between 5:00 - 7:00AM Eastern Time.

  • Report Date-Range Rules

    • Aggregates: based on impression timescale

    • Log-Level Delivery: all impressions that occurred within the report range.

    • Log-Level Conversion: all conversions in the conversion time-scale, but impressions may precede the range depending on attribution window. (e.g. a conversion on 1/25/2026 may be attributed to an impression occurring 1/5/2026)

  • Date Timezone Consideration

    • Users can choose the timezone to be used for all date fields included in the reports

    • Defaults to UTC

    • Timezones need to be provided with regional information to account for DST

    • When a timezone is chosen that has DST, the report that spans the change to or from DST will skip an hour or contain multiple rows for the same hour respectively

  • Delayed Data Consideration

    • In certain cases, impression data from Display or specialty publishers may be reported with a delay. If a campaign includes these publishers, exports may initially reflect incomplete data until reconciliation is complete. To account for this, customers can either:

      • Use rolling report ranges (e.g., a rolling 7-day window) to allow late impressions to be captured.

      • Schedule exports on a lagged cadence (e.g., reporting on the prior week after a short delay) to ensure reconciled impression counts.

  • File Format

    • CSV

    • dynamic table via Snowflake

  • Locale handling

    • All numbers will be formatted following US-EN locale formatting rules

    • Numbers will use a period character as the decimal separator, with no thousands separator

    • Dates will be delivered using the ISO 8601 standard format YYYY-MM-DD (year, month, day) or YYYY-MM-DDTHH:mm:ss (with time)

Export Types Detail

  1. Aggregate Export (with or without attribution)

Use Cases

  • High-level campaign performance reporting

  • Executive-friendly views of delivery and outcomes without handling large data files

  • Flexible grouping by dimensions such as date, campaign, line, publisher, product, etc.

  • Ideal for teams that want topline metrics for dashboards, reconciliation or benchmarking

Considerations

  • Attribution-enabled aggregates may appear incomplete if the window exceeds the report range. For example, a weekly file may exclude conversions that occur outside that 7-day window.

  • To mitigate this, rolling or lagged reports are recommended:

    • Rolling: Sending a 30 day report each send, with the most recent day (previous day) added, and the earliest date (31 days prior) dropped

    • Lagged: 30-day lag to align with a 30-day lookback window so that today’s report shows data from 31 days ago.

  • Reach is not dynamically calculable across multiple dimensions in aggregated files

Schema / Example Fields

  • Dimensions

    • Date

    • Campaign Name

    • Line Item Name

    • Media Name (Publisher)

    • Product Type

    • Creative Name

    • Zip Code

    • DMA Name

    • DMA Code

    • Screen Type

  • Metrics

    • Spend

    • Impressions

    • IP Reach 1 day

    • IP Reach 7 day

    • Frequency

    • Quartile 1

    • Quartile 2

    • Quartile 3

    • Quartile 4

    • Completed Views

    • VCR (video completion rate)

    • Clicks

    • Conversion(s)

      • Responses

      • + any pixel events

  1. Log-Level Delivery Export

Use Cases

  • Granular impression-level data for customers with internal BI or data science resources.

  • Enables full flexibility in slicing, filtering, and modeling performance data.

  • Required for brands that want to dynamically calculate reach across any permutation of dimensions

Considerations

  • File sizes can be very large, requiring storage and processing infrastructure.

  • Exports include every impression, even those with no conversions

  • More technical expertise is needed to integrate, analyze, and interpret the data compared to aggregate files

Schema / Example Fields

  • Impression Timestamp

  • Impression Date

  • Impression Time

  • IP Address (SHA256 Salted Hash)

  • Impression ID

  • Campaign Name / ID

  • Line Item Name / ID

  • Media Name (Publisher)

  • Product Type

  • Creative Name

  • Zip Code

  • DMA Name

  • DMA Code

  • Screen Type

  • Quartile 1

  • Quartile 2

  • Quartile 3

  • Quartile 4

  • Completes

  • Clicks

  1. Log-Level Attribution Export

Use Cases

  • Detailed conversion-level reporting for validation and attribution analysis

  • Analyze attributed events’ custom value-parameters passed through even pixel (e.g. order_id)

  • Enables advertisers to understand exactly which impressions contributed to each conversion

  • Supports both Last-Touch Attribution (LTA) and Multi-Touch Attribution (MTA) methodologies

    • LTA each conversion is tied to a single impression

    • MTA each conversion may appear multiple times, with all contributing impressions listed

Considerations

  • File sizes are smaller than delivery logs but still require technical resources to manage.

  • Depending on attribution methodology, conversion rows may be duplicated (MTA) or limited to one per conversion.

  • Reports are based on the conversion timescale: all conversions that occur within the report start and end dates are included. However, attributed impressions for those conversions may fall outside the report period (earlier than the report start date), depending on the lookback window.

  • Reach cannot be directly calculated unless paired with delivery logs

Schema / Example Fields

  • Conversion Timestamp

  • Conversion Date

  • Conversion Time

  • Conversion Event Name

  • Conversion IP Address (hashed SHA 256)

  • Pixel ID

  • Event ID

  • Page URL

  • Referrer URL

  • Revenue (macro)

  • Order ID (macro)

  • User ID (macro)

  • Product (macro)

  • Custom (macro)

  • Impression Timestamp

  • Impression IP Address (hashed SHA256)

  • Impression ID

  • Campaign Name / ID

  • Line Item Name / ID

  • Media Name (Publisher)

  • Product Type

  • Creative Name

  • Zip Code

  • DMA Name

  • DMA Code

  • Screen Type

Did this answer your question?