ZMS Campaign SKU Daily

KPI Aggregation & Granularity: The calculated KPIs in this dataset—specifically ropi, roas, cvr, ctr, cpc, roas_direct, clicks_conversion_rate_direct, and cpsi_direct—are provided at the baseline granularity of campaign_id, dt, config_sku, and country. If you alter or roll up the granularity of your analysis, do not directly sum (SUM) or average (AVG) these KPI columns. Instead, recalculate the metrics using the explicit formulas mentioned in the schema description below (i.e., aggregate the base metrics first, then apply the formula). Example usage can be seen in the Example Measures section.

Overview

This dataset delivers daily insights for Zalando Partner Marketing Services (ZMS) campaigns, with a specific focus on Performance Objective Campaigns structured at the granular config_sku level.

By tracking key metrics—such as viewable_impressions, ad_clicks, budget_spent, and items_sold, partners can monitor exactly how distinct SKUs perform across various campaigns and target countries. These granular insights enable partners to evaluate marketing effectiveness, optimise campaign budgets, and seamlessly align inventory management with active marketing initiatives.

Property	Description
Data granularity	`campaign_id`, `dt`, `config_sku`, `country`
History available	Data available from `2024-01-01`
Update frequency	Daily
Data retention	Past 2 years + current year
Primary keys	`campaign_id`, `dt`, `config_sku`, `country`
Partition columns	`dt`
SLOs	Daily around 1:00 PM UTC (for the previous day's data)
Notice period for changes	See Versioning and Deprecation Policy
Support	Support available via partner-care@zalando.de. No 24/7 support available.

Data Refresh Strategy

Partners can choose either a full load or an incremental load approach when consuming this dataset, depending on what best fits their pipeline.

Option A: Incremental Load (recommended)

How it works: Merge new and updated records into your local table using account_id, campaign_id, dt, country, and config_sku as the merge keys. Rows matching these keys should be updated/replaced with the latest version; new key combinations should be inserted.
Identifying changed records: Use the updated_at timestamp to request only rows that have changed since your last successful load (e.g. WHERE updated_at > <last_max_updated_at_you_processed>).
Why choose this? Lower processing cost and faster refreshes, since only new/changed rows since your last run need to be transferred and merged.
Soft delete handling: The dataset includes an is_deleted column to indicate records that have been logically deleted. Partners should exclude records where is_deleted = true when performing attribution KPI calculations to ensure accurate reporting.

Option B: Full Load

How it works: Overwrite your entire local table with the full dataset on each retrieval.
Why choose this? Simplest to implement and operate — no merge logic required, and guarantees your table always reflects the latest state, including retroactive corrections to historical data (up to 2 years).
Soft delete handling: The dataset includes an is_deleted column to indicate records that have been logically deleted. Partners should exclude records where is_deleted = true when performing attribution KPI calculations to ensure accurate reporting.

Tracking refreshes

Regardless of approach, partners can reference the updated_at timestamp column to verify exactly when a given record was last refreshed.

Table Reference

zms_campaign_sku_daily_share.direct_data_sharing.zms_campaign_sku_daily

Schema

Column name	Format	Description
`account_id`	`string`	Zalando Partner Account ID (included as partition key for downstream processing).
`merchant_id`	`string`	Zalando Partner ID (BPID).
`campaign_id`	`string`	Unique ID of the campaign (n_code).
`campaign_name`	`string`	Name of the campaign.
`campaign_objective`	`string`	Objective of the campaign. One of: Conversion, Consideration, Awareness.
`start_date`	`date`	Campaign start date.
`end_date`	`date`	Campaign end date.
`dt`	`date`	Event record date.
`country`	`string`	ISO country code of campaign (e.g. DE, AT).
`config_sku`	`string`	Zalando article variant (e.g. ON321N1QS-K11).
`merchant_sku`	`string`	Partners reference code for a SKU (Stock Keeping Unit) at the article and colour level.
`article_name`	`string`	Name of the article (e.g. PL BB Cap).
`brand_code`	`string`	Brand code of the article.
`brand_name`	`string`	The name of the brand.
`gender`	`string`	The gender the product is intended for (e.g. female, male, unisex).
`viewable_impressions`	`bigint`	The number of times a user has been exposed to at least 50% of your ad content (25% for big ad format).
`ad_clicks`	`bigint`	The number of times a user has clicked on your ads.
`budget_spent`	`double`	Your total campaign budget, including discounts, vouchers and free media.
`partner_spent`	`double`	The amount you have invested in your campaign, excluding discounts, vouchers and free media.
`items_sold`	`bigint`	The number of items sold after users clicked on your ads (before cancellations and returns).
`attributed_gmv`	`double`	The value of items sold after users clicked on your ads (before cancellations and returns).
`ropi`	`double`	Return On Partner Invest: Calculated as Attributed GMV divided by Partner Invest and excluding any budget spent on TikTok, Pinterest and/or Snapchat. Formula:`SUM(attributed_gmv) / SUM(partner_spent)`
`roas`	`double`	Return On Ad Spend: Calculated as Attributed GMV divided by Budget Spent and excluding any budget spent on TikTok, Pinterest and/or Snapchat. Formula:`SUM(attributed_gmv) / SUM(budget_spent)`
`cvr`	`double`	Conversion Rate: The percentage of clicks on your ads that have led to sales (excluding clicks from TikTok, Pinterest and/or Snapchat). Formula:`SUM(items_sold) / SUM(ad_clicks)`
`ctr`	`double`	Click-through Rate: Calculated as Clicks divided by Viewable Impressions. Formula:`SUM(ad_clicks) / SUM(viewable_impressions)`
`cpc`	`double`	Cost Per Click: The amount you pay for each click on your ads. Formula:`SUM(budget_spent) / SUM(ad_clicks)`
`created_at`	`timestamp`	Timestamp in UTC when the data was created.
`updated_at`	`timestamp`	Timestamp in UTC when the data was updated.
`sales_direct`	`bigint`	The number of items sold - before cancellations and returns - after a user clicked on an ad of the exact same SKU.
`gmv_direct`	`double`	The value of items sold - before cancellations and returns - after users clicked on an ad of the exact same SKU.
`roas_direct`	`double`	Return on Ad Spend: The GMV your business earns for each euro it spends on advertising. Calculated as Direct GMV divided by Budget Spent. Formula:`SUM(gmv_direct) / SUM(budget_spent)`
`clicks_conversion_rate_direct`	`double`	The percentage of clicks on your ads that have led to a sale of the exact same SKU. Formula: `SUM(sales_direct) / (SUM(ad_clicks)`
`cpsi_direct`	`double`	The advertising cost per individual item sold of the exact same SKU. Formula:`SUM(budget_spent) / SUM(sales_direct)`
`is_deleted`	`boolean`	Soft delete flag: true when the row was deleted due to attribution correction.

Example Measures

Campaign Overview at Article level (ROAS, ROPI)

This query evaluates campaign returns at the config_sku level . It calculates the Return on Ad Spend (ROAS) and Return on Partner Invest (ROPI). Optionally report can be expanded to different granularities like country, dt.

SELECT
    campaign_id,
    campaign_name,
    config_sku,
    SUM(viewable_impressions) AS viewable_impressions,
    SUM(ad_clicks) AS ad_clicks,
    SUM(budget_spent) AS budget_spent,
    SUM(partner_spent) AS partner_spent,
    SUM(attributed_gmv) AS attributed_gmv,
    SUM(attributed_gmv) / NULLIF(SUM(budget_spent), 0) AS roas,
    SUM(attributed_gmv) / NULLIF(SUM(partner_spent), 0) AS ropi
FROM zms_campaign_sku_daily_share.direct_data_sharing.zms_campaign_sku_daily
WHERE is_deleted IS FALSE
GROUP BY 1, 2, 3

Top SKUs by Direct Sales and Efficiency

Ranks advertised SKUs by sales_direct and roas_direct to identify which specific SKUs drive the most direct conversions. Use this to prioritize high-performing SKUs for future campaigns and budget allocation.

SELECT
    campaign_id,
    campaign_name,
    config_sku,
    article_name,
    brand_name,
    SUM(budget_spent)                                           AS budget_spent,
    SUM(sales_direct)                                           AS sales_direct,
    SUM(gmv_direct)                                             AS gmv_direct,
    SUM(gmv_direct)   / NULLIF(SUM(budget_spent), 0)            AS roas_direct,
    SUM(budget_spent) / NULLIF(SUM(sales_direct), 0)            AS cpsi_direct,
    SUM(sales_direct) / NULLIF(SUM(ad_clicks), 0)               AS clicks_conversion_rate_direct
FROM zms_campaign_sku_daily_share.direct_data_sharing.zms_campaign_sku_daily
WHERE is_deleted IS FALSE
GROUP BY 1, 2, 3, 4, 5
ORDER BY sales_direct DESC

Direct vs Attributed Performance by SKU

Compares direct attribution (gmv_direct, roas_direct) with broader click attribution (attributed_gmv, roas) at the SKU level. A large gap between the two signals a strong halo effect — customers purchasing other products after clicking your ad — which can inform how you evaluate the full value of a campaign.

SELECT
    config_sku,
    article_name,
    brand_name,
    SUM(budget_spent)                                           AS budget_spent,
    SUM(sales_direct)                                           AS sales_direct,
    SUM(items_sold)                                             AS items_sold_attributed,
    SUM(gmv_direct)                                             AS gmv_direct,
    SUM(attributed_gmv)                                         AS attributed_gmv,
    SUM(gmv_direct)     / NULLIF(SUM(budget_spent), 0)          AS roas_direct,
    SUM(attributed_gmv) / NULLIF(SUM(budget_spent), 0)          AS roas_attributed
FROM zms_campaign_sku_daily_share.direct_data_sharing.zms_campaign_sku_daily
WHERE is_deleted IS FALSE
GROUP BY 1, 2, 3
ORDER BY gmv_direct DESC

Changelog (non-breaking changes)

2026-09-22 - v1.4.0 (General Availability)

Summary: Dataset moved from pilot mode to General Availability.
Details: The backend serving this data has moved to a more reliable source maintaining the same existing schema of the dataset.
Availability: partners that were using the dataset during the pilot phase will continue their operations normally. New partners can request access to the data from the contact support button at the bottom of this page.
Impact: values for created_at, updated_at columns are all changed due to an initial overwrite job to switch the backend data source. Moving forward, created_at & updated_at columns will resume their normally expected behaviour.
Action required: None

2026-08-25 — v1.3.0 (Data load change)

Summary: Dataset now supports both full load and incremental load consumption patterns.
Details: Partners may either (a) fully overwrite their local table on each refresh, or (b) incrementally merge new/changed records using account_id, campaign_id, dt, country, and config_sku as merge keys, filtering on updated_at to identify changed rows since their last load. Both approaches must exclude/apply is_deleted = true rows appropriately (see Data Refresh Strategy section).
Availability: Available immediately; no change to underlying data or schema.
Impact: None expected for existing consumers. Partners currently doing full loads may continue unchanged; incremental load is an optional alternative approach.
Action required: None. Partners wanting lower processing costs may optionally switch to the incremental merge pattern described above.

2026-07-28 — v1.2.0 (Additive schema change)

Summary: Added 1 new column for soft deletion
Details: New column allows partner to understand when an row was deleted from an upstream source. Added columns: is_deleted. Column is non-nullable.
Availability: Available from the release date 2026-07-28.
Impact: None expected for existing consumers. This is purely additive and does not modify existing columns or their behavior.
Action required: None. Existing queries will continue to work.

2026-07-08 — v1.1.0 (Additive schema change)

Summary: Added 5 new columns for direct attribution metrics at the SKU level
Details: New columns enable granular analysis of direct (same-SKU) attribution vs. broader click-through attribution at the article variant level. Added columns: sales_direct, gmv_direct, roas_direct, clicks_conversion_rate_direct, cpsi_direct. All columns are non-nullable and represent metrics specific to clicks on the exact same SKU.
Availability: Available from the release date 2026-07-08.
Impact: None expected for existing consumers. This is purely additive and does not modify existing columns or their behavior.
Action required: None. Existing queries will continue to work. Consumers can optionally add the new direct attribution columns to their pipelines to measure SKU-specific campaign effectiveness and identify halo effects.

Contact Support