- Facebook Ads data has four levels: Account, Campaign, Ad Set, and Ad. Each shows isolated performance.
- Apply breakdowns for placement, device, age, and action type to find which segments actually convert.
- Meta Pixel must be installed on your site before any conversion data appears in drill-down reports.
- Attribution windows (1-day vs. 7-day click) change how conversions are counted. Compare both before acting.
- Complex queries time out. Split date ranges and limit breakdowns to two per query to avoid data gaps.
- Coinis Advertise surfaces campaign performance without manual Ads Manager exports or API queries.
What Does Drilling Down Your Facebook Ads Performance Mean?
Facebook Ads Manager organizes your data in layers. Understanding which layer to look at, and when, is the foundation of smart optimization.
Hierarchical levels: Campaign, Ad Set, and Ad
Your account holds campaigns. Campaigns hold ad sets. Ad sets hold individual ads. Each level shows aggregated performance from everything beneath it. Drill down by clicking through each level in Ads Manager or querying Meta's Insights API directly.
Why granular analysis matters for optimization
Top-level numbers hide the truth. A campaign with a solid CTR might contain one ad set burning through budget with zero conversions. Drilling down reveals what's actually working. Stop guessing. Start acting.
---
The Four Levels of Facebook Ads Performance Data
Every data request starts at one of four levels. Choosing the right level saves time and surfaces cleaner signals.
Account level: High-level overview of all spending and results
Account-level data shows total spend, impressions, clicks, and conversions across every active campaign. Use it for budget monitoring and period-over-period comparisons. It tells you how much you spent. It doesn't tell you why results moved.
Campaign level: Grouped results from all ad sets and ads in a campaign
At campaign level, you see how a single objective performed overall. Compare campaigns against each other here. Traffic campaigns versus conversion campaigns. Brand awareness versus retargeting. This level reveals which goal is generating results.
Ad Set level: Performance of a specific targeting or budget segment
Ad sets are where targeting lives. Audience A versus Audience B. Placement A versus Placement B. Drill to ad set level to find which targeting configuration drives real results for your budget. This is where most optimization decisions happen.
Ad level: Individual ad creative performance
This is where creative testing plays out. Same audience, different copy. Same copy, different image. Ad-level data shows which creative combination wins. It is the most granular native level available in Ads Manager.
---
Key Breakdowns: Slicing Your Data Further
Breakdowns let you cut any level's data by an additional dimension. Per Meta's Insights API documentation, the API supports 40+ breakdowns to slice data by device, placement, demographics, and action types.
Demographic breakdowns: Age, gender, country, region
Filter results by age range, gender, or geography. A product converting better with 25-34-year-olds than 35-44-year-olds? A demographic breakdown surfaces that fast. Reallocate budget accordingly.
Placement and platform: Where ads appeared (Feed, Stories, Reels, Audience Network)
One of the most actionable breakdowns. Break down by publisher platform (Facebook, Instagram, Audience Network, Messenger) and by platform position (feed, story, reel). You will often find one or two placements driving the majority of conversions at the lowest cost per result.
Device breakdowns: Desktop, mobile, impression device type
The device platform breakdown splits results between mobile and desktop. Conversion rates frequently differ between the two. Knowing this shapes bid strategy and landing page priorities before you scale.
Action breakdowns: Conversion device, destination, action type
Action breakdowns show what users did after seeing your ad. Conversion device reveals whether someone saw the ad on mobile but purchased on desktop. Action type separates link clicks from purchases from video views. These are the breakdowns that connect ad spend to real business outcomes.
Note: As of April 27, 2021, certain breakdowns (region, DMA, hourly aggregations) no longer return off-Meta action metrics. Per Meta's Insights API breakdowns documentation, plan your breakdown strategy around this restriction.
Time-based breakdowns: Hourly, daily aggregation
Day-of-week and hourly patterns often reveal budget waste. If conversions cluster in evenings, running full spend through midday loses money. Time-based breakdowns make this visible and quantifiable.
---
How to Set Up Proper Tracking Before Drilling Down
Granular breakdowns are only as valuable as the data feeding them. Weak tracking means misleading reports.
Why Meta Pixel installation is essential for meaningful data
The Meta Pixel is a JavaScript snippet placed in the `
` tag of every page you want to track. Per Meta's Pixel documentation, this is required for tracking Facebook ad-driven visitor activity and enabling conversion data. Without it, Ads Manager measures only on-Facebook actions: link clicks, video views, engagement. It cannot track purchases, leads, or page events on your website.Install the Pixel before running conversion campaigns. Verify it fires correctly on your key pages using Meta's Pixel Helper browser extension.
Event tracking: Ensuring conversions are attributed correctly
Standard events (Purchase, Lead, AddToCart, CompleteRegistration) tell Meta exactly what happened on your site. Each event feeds the conversion columns you will analyze in drill-down reports. No events means no conversion data. Custom conversions work too but require additional configuration.
Conversion attribution windows: 1-day vs. 7-day click and view
Meta attributes conversions differently depending on the window you select. A 1-day click window counts only conversions within 24 hours of a click. A 7-day click window extends that to a full week. View-through attribution adds conversions from people who saw your ad but did not click.
Different windows tell different stories. Compare them in your reporting to understand the true attribution impact of each campaign before drawing conclusions.
---
Common Drill-Down Scenarios and What to Look For
Finding your best-performing placement to reallocate budget
Apply publisher platform and platform position breakdowns to your campaign data. Sort by cost per result. Feed placements almost always outperform Audience Network for direct-response goals. Move budget to what converts. Exclude what does not.
Identifying underperforming audience segments
Break down by age and gender. Find segments with high spend and low conversion rate. Exclude them or create a separate ad set with adjusted bids. Small adjustments here move the overall cost per acquisition meaningfully.
Comparing mobile vs. desktop conversion rates
Use the device platform breakdown. If mobile drives 80% of clicks but only 20% of purchases, your mobile landing page is likely the issue, not the ad creative. Drilling down isolates the problem so you fix the right thing.
Detecting which action types drive real business value
Not all actions are equal. Link clicks cost far less than purchases. Use the action type breakdown to separate vanity metrics from business results. Focus optimization on the action type that maps directly to revenue.
---
Limitations and Best Practices When Drilling Deep
Data per-call limits and when breakdowns become unavailable
Per Meta's Insights API best practices documentation, data per-call limits prevent overly complex queries from completing. The more breakdowns you stack and the wider your date range, the more likely a query times out. Break wide date ranges into weekly chunks. Limit breakdowns to one or two per query. Use asynchronous jobs for large reports.
High-cardinality breakdowns like action_target_id or product_id at the ad level are the most common timeout triggers. Keep those queries narrow and targeted.
Rate limiting on historical reach data (13+ months old)
Reach, frequency, and cost-per-thousand-people queries with breakdowns for data older than 13 months are rate-limited to 10 asynchronous requests per ad account per day, per Meta's best practices documentation. Plan historical analysis accordingly. Export data regularly so you are not chasing old data under tight rate limits.
Balancing granularity with query performance
More breakdowns reveal more. But the more you slice, the smaller each data cell becomes. Small sample sizes produce noisy results. A placement breakdown with 50 impressions is not statistically meaningful. Give breakdowns enough data volume before drawing conclusions or shifting budget.
---
Or let Coinis do it.
From a product URL to a live Meta campaign. AI-generated creatives. On-brand copy. Direct publish to Facebook and Instagram. Real performance reporting. All in one platform.
Start free. Upgrade when you're ready.
15 AI tokens a month. No credit card.
Frequently Asked Questions
What is the difference between campaign, ad set, and ad level data in Facebook Ads Manager?
Campaign level shows aggregated results for all ad sets and ads under one objective. Ad set level shows performance for a specific targeting or budget segment. Ad level shows individual creative performance. Drilling through each level reveals where results are actually coming from.
How do breakdowns work in Facebook Ads?
Breakdowns slice any level's data by an additional dimension, such as age, gender, placement, device, or action type. Per Meta's Insights API documentation, 40+ breakdowns are available. Apply one or two per query to avoid timeouts and keep results readable.
Why is my conversion data missing when I drill down in Facebook Ads?
Conversion data requires the Meta Pixel to be installed on your website. Without it, Ads Manager can only track on-Facebook actions like clicks and video views. Install the Pixel in the head tag of your key pages and set up standard events for the conversions you want to measure.
What are data per-call limits and how do they affect drill-down reports?
Meta's Insights API enforces data per-call limits to prevent overly complex queries from timing out. Queries with wide date ranges, many breakdowns, or high-cardinality dimensions like product_id are most at risk. Break queries into smaller date chunks and limit breakdowns per query to avoid incomplete data.