Facebook Ads Reporting: Custom Reports That Actually Drive Decisions

Meta's default reports show you everything except what matters. The ROAS column you see by default is platform-attributed, not blended. The "Today" view is misleading because attribution windows haven't matured. The campaign-level view hides where the money actually goes.

Here's the reporting setup that lets operators make actual decisions.

The reports Meta gives you

Default Ads Manager dashboard:

• Campaign / Ad Set / Ad views.
• Date range and breakdown by demographic, placement, etc.
• Pre-configured columns (Performance, Engagement, Conversions, etc.).

Useful but limited. The defaults bury what matters and surface vanity metrics.

What actually matters

For most e-commerce accounts, you should be tracking:

Performance metrics

• Spend. Per campaign, ad set, ad.
• Purchases (Pixel + CAPI deduplicated). The actual count.
• Purchase ROAS. Revenue ÷ spend. Caveat: platform-attributed.
• Cost per purchase. Spend ÷ purchases.
• CTR (link). Quality of click-through.
• Hook rate. 3-second video views ÷ impressions.
• Hold rate. ThruPlays or 75% video views ÷ impressions.

Funnel metrics

• CPM. Cost per 1,000 impressions. Watch for spikes (often signal fatigue).
• Frequency. Average impressions per user. Above 5/week often signals fatigue.
• Outbound CTR. Clicks that land outside Meta — true site traffic.
• Landing page views. Meta-reported actual landing arrivals.

Funnel ratios

• CTR → CVR. Of people who clicked, how many converted? Reveals landing page issues.
• Add to cart → Purchase. Cart-to-purchase ratio. Reveals checkout friction.
• CPM trend. Rising CPM often indicates audience fatigue.

Build a custom column set

In Ads Manager, click "Columns" → "Customize Columns." Build a set with these:

• Spend.
• Impressions.
• CPM.
• Frequency.
• Link Clicks.
• Outbound CTR.
• Landing Page Views.
• Adds to Cart.
• Initiated Checkouts.
• Purchases.
• Purchase ROAS.
• Cost per Purchase.
• Hook Rate (custom calculation: 3-Second Video Views / Impressions).
• Hold Rate.

Save this column set as your default. Skip the default Performance view — it's noise for most decisions.

Attribution settings

Default attribution: 7-day click + 1-day view. Reasonable.

When to change:

• Pure click attribution (7-day click only). If view-throughs feel inflated. Stricter, lower reported numbers.
• Click + view 1-day. For quick read on recent performance.
• 28-day window. Used to be available; deprecated for most accounts.

Choose one and stick with it. Constantly switching attribution windows confuses analysis.

Date range strategy

• Today. Useful for spotting catastrophic issues; misleading for performance assessment (attribution not yet mature).
• Last 3 days. Better for daily monitoring.
• Last 7 days. Standard for weekly performance review.
• Last 14 days. Best for ad-set and creative-level evaluation.
• Last 30 days. Strategic review. Spotting trends.

For decision-making, default to last 7 days. For week-over-week comparisons, lock to recent 7 days vs prior 7 days.

Breakdowns that reveal truth

In the Ads Manager breakdown menu, useful slices:

• By Age. Spot which age groups perform best.
• By Gender.
• By Region/Country. International performance variance.
• By Placement. IG Feed vs IG Reels vs FB Feed vs Stories. Often reveals dramatic variance.
• By Time of Day. Less actionable since auto-bid handles this, but useful for awareness.
• By Device. iOS vs Android vs Desktop.

A common finding: 30-50% of spend goes to placements producing 10-15% of conversions. Reallocate.

The reporting cadence

Daily glance (5 min)

• Total spend today.
• Total purchases.
• Yesterday's ROAS vs prior 7-day avg.
• Any campaign with anomalous CPM or frequency.

If something looks broken (ROAS at 0, purchases stopped), investigate. Otherwise, no action.

Weekly review (45 min)

• Campaign-level performance week-over-week.
• Ad set-level performance.
• Creative-level performance and fatigue.
• Audience overlap check.
• Placement breakdown.

Decide: what to scale, what to cut, what to test next week.

Monthly deep dive (3 hrs)

• Cohort analysis (LTV by acquisition month).
• New vs returning customer ROAS.
• Creative concept performance.
• Trend analysis (CPM, frequency, ROAS over 90 days).
• Attribution reconciliation against MER.

Platform attribution vs reality

Meta's reported ROAS is usually inflated 20-40% compared to actual blended performance. Why:

• Click and view attribution credits Meta for conversions where another channel was the real driver.
• Cross-channel attribution gaps (the same conversion is credited by multiple platforms).
• iOS attribution gaps.

Don't trust Meta's reported ROAS as the full truth. Reconcile against:

• MER (total revenue ÷ total marketing spend).
• Triple Whale or Northbeam for blended attribution.
• Direct match of Pixel-reported orders to Shopify orders.

Use Meta's ROAS for relative comparisons (which campaign is better than another). Use MER for absolute decisions (am I scaling profitably).

Custom reports for different stakeholders

For the operator (you)

• Campaign-level table with ROAS, CPM, frequency.
• Top 10 ads by spend with creative thumbnails.
• Audience overlap matrix.
• Placement breakdown.

For executives

• Total spend, total revenue.
• Blended ROAS / MER trend.
• New vs returning customer split.
• Brief commentary on significant changes.

Don't show executives the campaign-level details. They'll over-rotate on noise.

For the creative team

• Hook rate, hold rate, CTR by ad.
• Creative concept performance (group by concept tag).
• Fatigue alerts (creatives crossing CPM thresholds).
• New creative pipeline (what's tested next week).

Tools to extend Meta reporting

When Meta's native reports aren't enough:

• Triple Whale. Blended attribution across Meta, Google, TikTok. Strong for e-commerce decision-making.
• Northbeam. More sophisticated attribution (MMM-leaning).
• Looker Studio (Google Data Studio). Custom dashboards combining Meta, Google, GA4, and Shopify data.
• Google Sheets + Meta API. Bespoke pulls into a custom dashboard.

For most accounts under $50K/month spend, Meta + Sheets is enough. Above that, Triple Whale or similar pays for itself in better decisions.

Common reporting mistakes

• Using "Today" for evaluation. Attribution hasn't matured.
• Comparing different attribution windows across decisions. Inconsistency creates noise.
• Reading platform ROAS as truth. Always cross-check with MER.
• Looking at the wrong level. Scaling at campaign level when ad-level fatigue is the actual issue.
• No documentation of changes. Meta's "Activity Log" helps but build your own log of major changes.

What "good" reporting looks like

• A custom column set saved and used consistently.
• Weekly review documented in a shared doc/sheet.
• Monthly cohort review tied to LTV data.
• Reconciliation between Meta reporting and actual P&L.
• Decisions logged with rationale.

The goal of reporting isn't beautiful dashboards. It's faster, better decisions. If your reporting setup doesn't help you decide what to scale, kill, or test next — it's not doing its job.