Facebook Ads | Top Product Dataset
Brendon Reimer
Below are the fields that ChannelMix considers standard for the Facebook Ads Top Product dataset. By default, this dataset is scheduled to pull yesterday and the previous 28 days of data to align with Facebook's default conversion window.
Since the amount of product data available can be very overwhelming, this dataset is designed to only pull the top performing products in terms of click-through ratio in order to provide you with the most meaningful data. This dataset will pull all product ads that had impressions >= 100 and clicks >= 2 (in order to achieve a 2% CTR) on a given day.
If this filtering is too high or low for your needs, feel free to send a ticket to our Platform Support team to adjust the filters for you after you have activated the dataset.
The conversions included in this dataset will be anything that you see under "action_type" with the word "conversion", "view", "wishlist", "cart", "purchase", or "checkout" in its name from this link. Please feel free to send a ticket to the Platform Support team if you would like to include other conversions.
Facebook Ads data also includes Instagram Ads if you are advertising on both platforms.
This dataset can be automatically activated without a ticket through Dataset Activation.
| Field | Type | Description |
| Report Date | DATE | The date the activity occurred. |
| Account ID | STRING | The account ID. |
| Account Name | STRING | The account name. |
| Campaign ID | STRING | The campaign ID. |
| Campaign | STRING | The campaign name. |
| Ad Set ID | STRING | The ad set ID. |
| Ad Set Name | STRING | A group of ads. |
| Ad ID | STRING | The ad ID. |
| Ad Name | STRING | The ad name. |
| Product ID | STRING | This field usually includes an ID, product name, product category, and product sub-category. |
| Objective | STRING | The objective given to the campaign |
| Impressions | FLOAT | The number of times the ad was served. |
| Clicks | FLOAT | The number of clicks on the ad. |
| Cost | FLOAT | The amount spent on the ad. |
| Link Clicks | FLOAT | The number of link clicks attributed to the ad. |
| Video Views | FLOAT | The number of 3 second video views the ad received. |
| Video p25 Watched Actions | FLOAT | The number of times 25% video was watched. |
| Video p50 Watched Actions | FLOAT | The number of times 50% video was watched. |
| Video p75 Watched Actions | FLOAT | The number of times 75% video was watched. |
| Video p95 Watched Actions | FLOAT | The number of times 95% video was watched. |
| Video p100 Watched Actions | FLOAT | The number of times the entire video was watched. |
| Conversion Name | STRING | The name of the conversion. |
| Conversions | FLOAT | The number of conversions. |
| Conversion Value | FLOAT | The value associated to the conversions. |
| Insert Date | TIMESTAMP | ChannelMix Field: when the row was inserted into the dataset. |
| ChannelMix Profile | STRING | ChannelMix Field: the ChannelMix Profile associated to the data. |
NOTES:
- In some edge cases where the amount of data requested from Facebook is very large, it is possible that Facebook will not return data for a small percentage of ads. Unfortunately, this is an issue within the Facebook Ads API and is outside of our control.
- Due to the process applied by Facebook that allows users to opt in and out of tracking, the conversion numbers in legacy datasets might not match the data that you see in the Facebook user interface (more information available in this help center article). If you are seeing this type of discrepancy in your Facebook Ads dataset, you can submit a ticket to our Data Support team to update your dataset so that the statistical model applied in the Facebook user interface is also applied to the data that we receive from the API. All Facebook conversion datasets that were activated from August 2021 forward will automatically have the statistical model applied to the conversion numbers. If your dataset was activated on or after August 1, 2021 and you are still seeing a discrepancy, please submit a ticket and we will be happy to look into it for you.