Basics
What's the difference between Data Blending and Data Conversion?
What's the difference between Data Blending and Data Conversion?
Both live on the Data integration tab of every data source’s settings page, gated by the same on/off toggle.Data Blending layers complementary data with advertising data. Use it when a non-ad source — for example, GA4, a CRM, or an MMP — should sit alongside ad data in the same widget. The ad rows stay primary; the non-ad source adds columns.Data Conversion turns a non-ad source into a standalone ad channel. Once converted, the source shows up in ad data filters and can have other data sources blended onto it the same way they would blend onto Meta Ads or Google Ads. Use it when a channel Adriel doesn’t natively support is tracked in a spreadsheet, CSV, or database.
Why are blend settings required at all?
Why are blend settings required at all?
Ad-platform data sources and non-ad data sources have different shapes. The dashboard cannot guess how a custom field aligns with the ad-data schema, so the configuration tells it how to do the join. Without it, the non-ad source has no anchor to the ad rows.
How is the Data integration tab accessed?
How is the Data integration tab accessed?
Open the Connections menu, pick the data source, open Data source settings, and switch to the Data integration tab. From there, turn the toggle on and pick Data Blending or Data Conversion.Earlier versions also allowed editing blend settings from the pencil icon on a widget’s settings panel. That pencil icon now redirects to the data source settings page — blend and conversion configuration live on the data source, not on individual widgets.
Data Blending: join types
What is the difference between Left Join and Full Join?
What is the difference between Left Join and Full Join?
Left Join keeps the ad rows and adds matching columns from the non-ad source. Rows from the non-ad source that have no ad match are dropped.Full Join does everything Left Join does, plus it keeps non-ad source rows that have no ad match — adding them as new rows.Worked example. Two datasets:
- Google Ads with Campaign ID and Link Click.
- A Google Sheet with Sheet: Campaign ID and Sheet: Link Click.
In short: Left Join adds columns; Full Join adds columns and rows.
Left Join was selected but rows behave as Full Join
Left Join was selected but rows behave as Full Join
Left Join is only honored for these 10 standard marketing breakdowns:
- Channel
- Ad account
- Campaign
- Campaign name
- Ad set
- Ad set name
- Ad
- Ad name
- Keyword
- Keyword text
Full Join was selected but rows behave as Left Join
Full Join was selected but rows behave as Left Join
A filter on a marketing dimension that is not present in any breakdown mapping or metric mapping is partially or totally ignored. The fix is either to add the filtered dimension to the mapping, or to refine the filter to use a dimension that is included.
Why does a blended widget show fewer rows or zero values than the raw source?
Why does a blended widget show fewer rows or zero values than the raw source?
Blending uses Left Join by default, which scopes the blended source to values that also appear in a marketing source. When no marketing source is added, or the join keys don’t overlap, the blended metrics collapse to nothing. Attaching a Google Sheet budget onto existing campaigns is the intended Left Join case.Switch to Full Join only when the blended source should contribute rows of its own. Note that Full Join can slow dashboards significantly and enforces every mapped filter on both sides simultaneously.
Data Blending: mapping strategies
Case 1 — ID mapping (recommended when possible)
Case 1 — ID mapping (recommended when possible)
Use IDs (campaign ID, ad ID) as the join key. IDs are unique across ad accounts and channels, so this is the most accurate option. Default to this whenever the non-ad source includes IDs.
Case 2 — Name mapping
Case 2 — Name mapping
Use names (campaign name, ad set name) as the join key when IDs are not available. To avoid duplication when the same name exists across different ad accounts or channels, include parent-level names (for example, map both Ad Account and Campaign together).
Case 3 — Date-only / total matching
Case 3 — Date-only / total matching
For pure visual side-by-side comparison without a dimension join:
- Step 1: Select a date dimension.
- Step 2: Select Full Join.
- Step 3: Leave the widget dimension mappings empty.
- Step 4: Set Total Mapping to No grouping.
The data source has no ad-account dimension. How can ad accounts still be mapped?
The data source has no ad-account dimension. How can ad accounts still be mapped?
Tracking sources often omit ad account names. Use unique identifiers further down the hierarchy — campaign IDs and names — as the join key. The ad account dimension then aggregates across all child campaigns from those IDs.
Why don't blended rows merge when the campaign or ad account looks the same?
Why don't blended rows merge when the campaign or ad account looks the same?
Blend joins are literal string matches on every mapped key. Common breakers:
- A leading or trailing space on a Reddit campaign name.
- An
act_prefix on a Facebook ad account ID that doesn’t appear on the other side. - A typo in a Google Sheet ID.
- A differing column name inside a custom-metric formula.
- A mapping that includes both Campaign ID and Campaign Name when one side only has the name.
Limits and defaults
How many mapping rows and pairs are allowed?
How many mapping rows and pairs are allowed?
Up to 15 widget-dimension rows per blend configuration. Each row supports up to 5 target-to-source pairs. The same widget dimension can only be selected in one row.
A data source was already blended at creation time
A data source was already blended at creation time
Several connectors ship with default blend settings — GA4 uses UTM-based defaults (forced Left Join), and HubSpot Deals uses first-touch UTM defaults. The defaults are visible in the Data integration tab and can be edited or replaced anytime.
Why did the widget return an empty result?
Why did the widget return an empty result?
The most common cause is No grouping as Total Mapping combined with a widget dimension that has no explicit mapping. The default fallback row is intentionally ignored in this case. Add a row that matches the widget’s current dimension, or pick a different Total Mapping.
Filtering
Can blended data be filtered?
Can blended data be filtered?
Yes. Two paths work:
- Filter directly using a dimension from the non-ad source.
- Filter using a marketing dimension that is included in at least one breakdown-mapping condition (Step 3) or in the metric mapping (Step 4).
Data Conversion
Which sources can be converted into ad channels?
Which sources can be converted into ad channels?
Non-ad sources where columns are user-defined: Google Sheets, CSV File Import, the database connectors (for example, BigQuery, PostgreSQL, Redshift, MySQL), and Amazon S3 file imports. Marketing connectors (Meta Ads, Google Ads, etc.) are already ad channels and cannot be converted.
Can a converted source be filtered, grouped, or charted the same way as Meta Ads?
Can a converted source be filtered, grouped, or charted the same way as Meta Ads?
Yes. Once a source is converted, it appears in the channel filter and contributes to channel-level totals (spend, impressions, clicks). The same breakdowns and metrics apply, provided the mapping covers them.
Can other data sources be blended onto a converted source?
Can other data sources be blended onto a converted source?
Yes — that is the main reason to convert rather than blend. After conversion, treat the source like any ad channel: blend HubSpot Deals onto it for attribution, blend a goal-tracking sheet onto it for budget comparison, and so on.
Blending with GA4
Why do GA4 blended rows appear duplicated across campaigns or drop out of the widget?
Why do GA4 blended rows appear duplicated across campaigns or drop out of the widget?
GA4 blend data aggregates at the ad-account level and groups only by UTM Source, UTM Campaign, and UTM Medium. Campaigns that share those three values (differing only in UTM Content) return the same GA4 numbers under each custom breakdown row.When GA4 Session Campaign Name is mapped to a marketing breakdown, the campaign name is applied as a filter — but a Facebook UTM value doesn’t equal Facebook’s own campaign name, so GA4 rows fall out as separate rows named by the session campaign name.Fixes:
- Add a lower-level condition such as UTM Content to separate the campaigns.
- Reconsider the campaign-name mapping if it’s forcing unwanted rows.