Platform & Device

Cross-platform and device performance — for comparing Facebook, Instagram, Messenger, and Audience Network across placements and devices.

This is the richest Meta Ads data model. It carries nearly every available metric, plus exclusives like Reach, Frequency, Hook Rate, Hold Rate, and per-funnel-step cost. Start here when you don’t know which data model to pick.

What this data model represents

Grain: one row per date × ad_account × campaign × adset × ad × ad_platform × placement × impression_device. Metrics, grouped by category — every number you can compute on this data model. Expand below for examples in each.

Show sample metrics in each category

Dimensions, grouped by category — every way you can split, filter, or group those metrics. Expand below for examples in each.

Show sample dimensions in each category

Source: Meta Ads API, daily refresh.

Slice by

Unique to this data model:

Dimension	Type	Values
`ad_platform`	String	`Facebook`, `Instagram`, `Messenger`, `Audience Network`
`placement`	String	`Feed`, `Stories`, `Reels`, `Right Column`, `In-Stream`, `Search Results`, `Marketplace`, `Instant Articles`
`impression_device`	String	`Mobile`, `Desktop`, `Tablet`, `Other`

Plus the standard campaign hierarchy, time, account, and attribution dimensions — documented once on the Meta Ads Overview.

Querying all three slice-by dimensions at once is highly granular and can produce sparse rows. Start with one or two and add the third only when you need that level of detail.

Use it to answer

How does ROAS compare across Facebook vs. Instagram vs. Messenger vs. Audience Network?
Which placement × device combinations are winning, and which are losing money?
Does creative format need to change by platform? (vertical for Reels, square for Feed)
How do Hook Rate and Hold Rate vary by platform to inform video strategy?

Available metrics

Everything you can compute on this data model. Need a metric not listed? See the Meta Ads metric availability matrix.

Metric	Formula
Spend & revenue
Ad Spend	Sum of daily spend
Conversion Value	Sum of reported revenue
ROAS	Conversion Value ÷ Ad Spend
Conversions	Sum of reported orders
AOV	Conversion Value ÷ Conversions
Clicks & impressions
Clicks (All)	Sum of clicks (inline + outbound)
Impressions	Sum of impressions
CTR (All)	Clicks ÷ Impressions × 100
Outbound Clicks	Sum of clicks leading off Meta
Inline Clicks	Sum of on-platform engagement clicks
Unique Clicks	Avg of deduplicated user clicks
Link Click-Through Rate	Link Clicks ÷ Impressions × 100
Cost efficiency
CPC (All)	Ad Spend ÷ Clicks
CPM	Ad Spend ÷ Impressions × 1,000
CPA	Ad Spend ÷ Conversions
Cost Per Outbound Click	Ad Spend ÷ Outbound Clicks
Cost Per Link Click	Ad Spend ÷ Link Clicks
Revenue Per Link Click	Conversion Value ÷ Outbound Clicks
Engagement
Landing Page Views	Sum
Post Engagement	Sum (reactions + comments + shares + clicks)
Inline Post Engagement	Sum (on-platform-only)
Page Engagement	Sum (Facebook Page likes, follows)
Post Reactions	Sum (Like, Love, Haha, Wow, Sad, Angry)
Post Shares	Sum
Cost Per Post Engagement	Ad Spend ÷ Post Engagements
Cost Per Inline Post Engagement	Ad Spend ÷ Inline Post Engagements
Cost Per Page Engagement	Ad Spend ÷ Page Engagements
Cost Per Landing Page View	Ad Spend ÷ Landing Page Views
Conversion funnel
View Content	Sum
Add to Cart	Sum
Checkouts Initiated	Sum
Add Payment Info	Sum
Cost Per View Content	Ad Spend ÷ View Content
Cost Per Add to Cart	Ad Spend ÷ Add to Cart
Cost Per Checkout Initiated	Ad Spend ÷ Checkouts Initiated
Cost Per Add Payment Info	Ad Spend ÷ Add Payment Info
Video
Video Plays	Sum
Video 25% Watched	Sum
Video 50% Watched	Sum
Video 75% Watched	Sum
Video 100% Watched	Sum
Cost Per Video Play	Ad Spend ÷ Video Plays
Hook Rate	Video Plays ÷ Impressions × 100
Hold Rate	Video 100% ÷ Video Plays × 100
Reach & frequency
Reach	Sum of unique people who saw ads
Frequency	Impressions ÷ Reach

Hook Rate and Hold Rate are exclusive to this data model. If you need them at audience or creative level, derive from raw fields (video_play_actions, video_views_pct_100).

Raw fields (for custom calculations)

This data model exposes a near-complete raw field set (one of the richest, alongside Age & Gender):ad_spend · clicks · impressions · reported_orders · reported_revenue · unique_clicks · inline_clicks · inline_post_engagement · outbound_clicks · landing_page_views · add_to_cart · checkouts_initiated · view_content · post_engagement · page_engagement · post_reaction · add_payment_info · post_share · video_play_actions · video_view_avg_time · video_views_pct_25 · video_views_pct_50 · video_views_pct_75 · video_views_pct_100

​What this data model represents

​Slice by

​Use it to answer

​Available metrics

What this data model represents

Slice by

Use it to answer

Available metrics