Blend Data: FAQs

What does Data Blending mean?

A: Data blending is the process of bringing together information from various sources and merging it into a single set of data. This allows you to create charts and analyses using data from different sources, as long as they have a common dimension.

Why are blend settings required?

A: Marketing data sources and other data sources have different data structures, requiring different blending methods. Non-marketing data needs to be matched to the format of the Adriel dataset to integrate seamlessly.

How can I blend my data in Adriel?

A: To blend generic data with marketing data, you need to create a blend data configuration on your generic data source. You can access the data source’s blend data configuration in different ways:

1. Via Widget Settings:
    - Open the Widget Settings.
    - Scroll down and click the Data Blend Settings button at the very bottom, then select your data source.

2. Via Data Source List:
    - Open the Widget Settings.
    - Go to the Data Source Settings panel.
    - Click the Pencil Icon on the data source list.
  
3. Via Connections Menu:
    - Go to the Connections menu.
    - Click the Settings button.
    - This opens the right panel; click on Blend Data Settings.

What is the difference between left join and full join?

A: Left join allows you to map rows from the blended data source with the marketing rows when some conditions are fulfilled. Full join allows you to map rows from the blended data source with the marketing rows when conditions are fulfilled, just like left join, but it also adds the rows when the conditions are not fulfilled.

To sum up: left join allows you to add columns to your data set while full join will allow you to add columns and rows to your data set.

For example, we have two datasets:
1. Marketing Data with columns `Campaign ID` and `Link Click`.
2. Google Spreadsheet (GSS) with columns `GSS: Campaign ID` and `GSS: Link Click`.

 

 

This is the Google Spreadsheet’s blend data settings:

• Breakdown mapping: `Campaign ID` <> `GSS: Campaign ID`
• Metric mapping: `Link Click` <> `GSS: Link Click`

Result of Left Join:
- If an ID exists in Marketing Data but not in the GSS, the `Link Click` value from the GSS will not be shown.
    

 

Result of Full Join:
- All IDs from both datasets are included.
    
    

 

Are there any recommended blend data settings?

A: Yes, here are some recommended settings based on different scenarios:

Case 1: ID mapping

If possible, we recommend using these settings as they are the simplest settings possible. It is also very accurate as IDs are unique across different ad accounts or channels.

Case 2: Name mapping

When using name mapping, you might encounter data duplication issues if you use the same names across different ad accounts or channels. To avoid duplication, it is recommended to include the parent levels’ names.

Case 3: UTM mapping

Using UTMs to blend your data sources is convenient but you might encounter data duplication if your UTM configuration is not flawless across all of your marketing data sources. This configuration works in most cases; however, it might need to be refined to fit your exact way of using UTMs.

Case 4: Matching on date breakdowns or total data only

If your goal is to only visualize data from different sources together, you might not need a complex configuration. To visualize your generic data alongside your marketing data using the date breakdowns or the total, you can use this configuration:
- Step 1: Select your date dimension.
- Step 2: Select full join.
- Step 3: Leave the widget dimension mappings empty and select “no Grouping” as a total mapping.

 

My data source doesn’t have any information able to map with my ad account. How can I map the ad account dimension?

A: Ad account names are often missing from tracking channels. To allow mapping for the ad account dimension, unique identifiers like campaign IDs and names can be used. If you choose this condition, the ad account data will be composed of the aggregation of all campaigns.

I just created my data source but it is already blended. Why?

A: When possible, we assign default blend data settings to your data source so that you can start using it right away! You can access the blend data settings of this data source, review it, and update it if needed.

Can blended data be filtered?

A: Yes. You can filter by using a dimension of your generic data source directly, or if you filter by using a marketing dimension, make sure that this dimension is included in at least one condition in your breakdown mappings (step 3) or that it is included in your metric mapping (step 4).

 

I would like to add a filter on a channel included in my blended data source but I can’t find it in the channel list.

A: The channels in the list will not include the channels that are in blended data sources. You will need to add the channel name manually. In the search section, type the channel name you would like to add to your filter and then click on “apply data including <channelName>”. Make sure that the name is the same.

I’ve set the left join but it seems like it is full join. Why?

A: Only full join is supported other than for these breakdowns:
- Channel
- Ad account
- Campaign
- Campaign name
- Ad set name
- Ad
- Ad name
- Keyword
- Keyword text

In this case, you will see the message like:
⚠️ [Note] Except for marketing breakdowns, it defaults to a full join regardless of the settings. This means some rows might be added if the dimension values do not exactly match existing rows.

I’ve set the full join but it seems like it is left join. Why?

A: This issue occurs because some breakdowns involved in filters are not mapped in your blend data settings, causing those filters to be totally or partially ignored. To prevent this, we recommend refining your blend data settings or filter settings.

In this case, you will see the message like:
⚠️ [Note] Some dimensions involved in filters are not mapped in your blend data settings resulting in those filters being totally or partially ignored. To prevent this issue, we recommend refining your blend data settings.

It seems like I cannot see the full set of data. Why?

A: To ensure optimal performance, we limit the amount of data that can be displayed. To view the full dataset, we recommend adjusting the widget settings or changing the date filter.

In this case, you will see the message like:

⚠️ Your data table is too big. It has been trimmed down to 200 rows to keep things running smoothly. It's recommended to adjust the widget settings or select a shorter date range to prevent this issue.

Why does it seem like the data has been duplicated?

A: This issue occurs when some data matches multiple rows. To resolve this, adjust the blend data settings to make the breakdown as unique as possible. You can do this by changing the condition to use an ID that acts as a unique identifier or by adding more conditions to better define the breakdowns.

In this case, you will see the message like:
⚠️ [Note] Some blended data can be attributed to several rows at once, resulting in duplicated data. To prevent this issue, we recommend refining your blend data settings.

 

Can I map the specific data source with the other selected one?

A: Currently, we support mapping marketing and non-standard marketing data only. This feature will be available in the near future.


For more detailed information or additional assistance, please contact our support team.