Troubleshoot Meta Ads

41 common Meta ads situations, explained at the mechanism level. Find yours, understand what is happening, and go straight to the relevant fix or tool.

⌕
All
Account & Policy
Performance
Creative
Budget
Targeting
Leads & Tracking
Meta's automated enforcement systems disable accounts when they detect a policy violation, unusual activity, or a payment failure. Most restrictions are recoverable, but the path depends on the reason.
  1. Go to business.facebook.com/accountquality to identify restriction type
  2. Check email (including spam) for Meta enforcement notifications
  3. Review recent ads for disapprovals that preceded the restriction
  4. Complete identity verification if required
  5. Submit review request via Account Quality dashboard
Diagnose in tool Read full guide Open problem page
A Business Manager restriction affects everything underneath it — all ad accounts, pages, and pixels. It is more serious than a single ad account restriction and usually requires identity or business document verification to resolve.
  1. Check business.facebook.com/accountquality for Business Manager status
  2. Go to Business Settings > Security Center for pending verifications
  3. Check if any admin users have restricted personal accounts
  4. Complete Business Verification with registration documents
  5. Submit review at the Business Manager level, not ad account level
Diagnose in tool Read full guide Open problem page
Meta's automated review system disapproves ads that match patterns it associates with policy violations. The system works by pattern recognition, not intent — which means legitimate ads get flagged regularly.
  1. Hover over Disapproved status to see the policy category
  2. Check landing page for claims not made in the ad copy
  3. Review ad copy for health claims, before/after, or guaranteed outcomes
  4. Confirm Special Ad Category is set if required (credit, housing, employment)
  5. Use Request Review button after making substantive changes
Diagnose in tool Read full guide Open problem page
Meta asks for two types of verification: personal identity (confirming who you are as an individual) and business verification (confirming your business exists). They have different requirements and different implications for your account.
  1. Go to Business Settings > Security Center to see verification type required
  2. For identity: prepare government-issued photo ID matching your account name
  3. For business: prepare business registration, utility bill, or bank statement
  4. Ensure documents are recent (within 3 months for utility bills)
  5. Upload documents unedited — no cropping or annotations
Diagnose in tool Read full guide Open problem page
A payment failure pauses all campaigns immediately. Meta does not run ads without confirmed payment. Most billing problems resolve within minutes once the payment issue is corrected, but an unresolved failure that sits for more than 48 hours can escalate to account suspension.
  1. Go to Billing > Payment Activity to check failed payments
  2. Check if card is expired or billing address is incorrect
  3. Contact bank to authorize Meta (Facebook) transactions
  4. Check Account Spending Limit in Billing settings
  5. Clear any outstanding balance with a manual payment
Diagnose in tool Read full guide Open problem page
Your account is being forced to re-learn a more expensive conversion path. This usually comes from a change in auction competition, conversion rate, or optimization stability.
  1. Check conversion rate on site (not just CTR)
  2. Check CPM and frequency for spikes
  3. Check if anything changed in budget, attribution, creatives, or pixel
Diagnose in tool Read full guide Open problem page
Overnight drops are usually attribution delay, learning disruption, audience saturation, or a tracking issue. The key is identifying which layer changed.
  1. Check attribution window and reporting delay (24 to 72h)
  2. Check delivery (spend, impressions) to confirm it is real
  3. Check pixel and events for sudden drop in fires
Diagnose in tool Read full guide Open problem page
This is usually conversion friction, tracking breakage, or the algorithm drifting toward cheap clicks due to weak conversion feedback.
  1. Confirm landing page works on mobile and loads fast
  2. Check pixel fires and match quality
  3. Compare click quality: sessions, time on site, add-to-cart, lead start
Diagnose in tool Read full guide Open problem page
CPM spikes are usually auction competition, audience restriction, or a shift to more expensive placements and impressions.
  1. Check audience size and exclusions for over-restriction
  2. Check placement changes or Advantage+ changes
  3. Check seasonality and auction overlap
Diagnose in tool Read full guide Open problem page
Volatility can be normal at low volume. The question is whether the system is stable enough to learn, or constantly being reset.
  1. Check conversion volume (low volume causes noisy CPA)
  2. Check if you are editing too frequently
  3. Check if budget is forcing delivery swings
Diagnose in tool Read full guide Open problem page
Learning requires enough stable conversion signals. If volume is too low or edits are too frequent, the system cannot settle.
  1. Check weekly conversion volume per ad set
  2. Check for frequent edits resetting learning
  3. Check optimization event is achievable (too deep means too few signals)
Diagnose in tool Read full guide Open problem page
Learning Limited means insufficient conversion volume for the system to reliably optimize. Fix it by increasing signals or simplifying structure.
  1. Check conversions per week per ad set
  2. Consolidate ad sets and reduce fragmentation
  3. Consider optimizing for a higher funnel event temporarily
Diagnose in tool Read full guide Open problem page
An active campaign that is not spending means Meta cannot win auctions for it. The campaign exists and is technically eligible to run, but something structural is preventing delivery.
  1. Check delivery status at ad set level for specific reason
  2. Verify all ads are approved and active
  3. Check audience size (below 50,000 is a problem)
  4. Check if bid cap or cost cap is set below market price
  5. Confirm campaign schedule and check for billing issues
Diagnose in tool Read full guide Open problem page
Most ads complete review within 24 hours. Reviews that extend beyond 48 hours indicate either a policy flag requiring human review, a high-volume queue period, or an account history that triggers additional scrutiny.
  1. Check exact ad status: In Review vs Pending Review
  2. Note how long ad has been in review (under 24h is normal)
  3. Check if campaign is in a Special Ad Category
  4. Check account's recent disapproval history
  5. Contact support after 48+ hours with specific ad ID
Diagnose in tool Read full guide Open problem page
Limited Delivery means the campaign cannot reach its delivery potential. Meta uses this status when structural issues prevent the algorithm from finding enough conversion events to optimise effectively.
  1. Calculate budget floor: target CPA x 50 / 7
  2. Check if too many ad sets are splitting budget
  3. Check if conversion event fires frequently enough
  4. Check audience size estimate in ad set
  5. Check if bid cap or cost cap is too restrictive
Diagnose in tool Read full guide Open problem page
Rising frequency and falling reach mean you have exhausted the readily available audience. Meta has reached everyone it can reach efficiently and is cycling back to show the same people the same ad again.
  1. Check frequency trend (above 6-7 on cold audience is saturation)
  2. Check reach curve over past 7-14 days
  3. Compare audience size to daily budget ratio
  4. Check CTR week over week for decline alongside rising frequency
Diagnose in tool Read full guide Open problem page
Editing a campaign resets or disrupts the learning phase. During learning, CPMs are elevated and delivery is unstable. The more significant the edit, the larger the disruption.
  1. Check campaign edit history for recent changes
  2. Compare CPM from week before edit to week after
  3. Check if delivery shows Learning or Learning Limited status
  4. Note how many days since the edit (learning resolves in ~7 days)
  5. Avoid making additional edits during learning phase
Diagnose in tool Read full guide Open problem page
ASC is Meta's fully automated campaign type for e-commerce product sales. It works well when the account has sufficient purchase history. Manual campaigns remain the right choice for new accounts, lead generation, B2B, and any objective that is not a direct product purchase.
  1. Check monthly purchase event volume (ASC needs 50+ monthly)
  2. Check account age and pixel history (3+ months ideal)
  3. Confirm objective is product purchase, not leads
  4. Count available creative assets (need 10+ for ASC)
  5. For borderline accounts: run both in parallel for 4 weeks
Diagnose in tool Read full guide Open problem page
ASC performance drops trace most often to creative exhaustion, another campaign competing in the same auction, or a structural edit that triggered re-calibration.
  1. Check creative performance breakdown — which are getting spend?
  2. Check for overlapping manual campaigns competing in auction
  3. Check edit history for changes in past 14 days
  4. Check Audience Segment breakdown for existing vs new customer split
  5. Compare CPM week over week for seasonal competition
Diagnose in tool Read full guide Open problem page
Low CTR usually means weak hook, unclear offer, or mismatched audience intent. It is a message and creative problem first.
  1. Check the first 1 second and first line of copy
  2. Check if the offer is explicit and specific
  3. Compare CTR by creative, not by campaign
Diagnose in tool Read full guide Open problem page
Fatigue is when the same audience has seen the same creative too many times, lowering response. It is visible in frequency and declining CTR and CVR.
  1. Check frequency trend and CTR trend together
  2. Check if CPM is rising while CTR is falling
  3. Rotate new angles, not just new edits
Diagnose in tool Read full guide Open problem page
If you are attracting clickers instead of buyers, your creative and landing page promise are misaligned. The algorithm optimizes toward the signals you feed it.
  1. Check landing page message match with ad promise
  2. Check on-site conversion rate by device
  3. Check if your objective and optimization event match the business goal
Diagnose in tool Read full guide Open problem page
Many tests win on a small pocket of audience. When scaled, the system expands to less responsive users and performance normalizes.
  1. Check audience size and overlap between test and scale setup
  2. Check if the test was under-delivering (small sample)
  3. Scale by keeping structure stable and adding new creative angles
Diagnose in tool Read full guide Open problem page
Minimum budget is defined by how many conversion signals you need per week for stable optimization. Budget depends on CPA and optimization depth.
  1. Estimate target conversions per week
  2. Estimate expected CPA or CPL
  3. Back-calculate daily budget needed for stability
Diagnose in tool Read full guide Open problem page
Scaling pushes the system into less efficient inventory and broader audience. If your conversion loop is weak, CPA rises fast.
  1. Check whether CVR drops after scaling or CPM rises first
  2. Increase budget gradually and keep edits minimal
  3. Scale by creative supply, not only by budget
Diagnose in tool Read full guide Open problem page
Budget increases preserve learning but can destabilize if too large. Duplication resets learning but can isolate audiences. Choose based on stability and signal volume.
  1. If stable and high volume, increase budget gradually
  2. If limited by audience or structure, duplicate for controlled tests
  3. Avoid duplicating into identical overlap setups
Diagnose in tool Read full guide Open problem page
Bid strategy controls how aggressively Meta bids in auctions on your behalf. Each option has a specific use case. Using manual bidding without adequate historical data is the most common cause of campaigns that do not deliver.
  1. Check conversion history (50+ conversions needed for cost cap)
  2. If using cost/bid cap, check if campaign is actually delivering
  3. Verify CPA target is based on historical data, not wishful thinking
  4. Use Highest Volume when starting out or during learning phase
  5. Switch to Cost Cap only after reliable CPA target is established
Diagnose in tool Read full guide Open problem page
When budget is below the learning threshold and can't move, you have two levers: reduce what you're asking the algorithm to do, or change what you're optimising for. Neither fully replaces more budget, but both shift the constraint.
  1. Consolidate to one ad set. Split budgets multiply the problem — each ad set needs 50 conversions/week independently. One ad set at your full budget is always better than two at half.
  2. Move up the funnel. If Purchase isn't hitting 50/week, switch to Add to Cart or Initiate Checkout. These fire 3–5x more often and give the algorithm enough signal to stabilise.
  3. Raise your CPA target temporarily. If your break-even CPA is €40 and you're targeting €25, the math doesn't close at any budget. Test at break-even CPA for 2 weeks before judging.
  4. Accept longer evaluation windows. Below threshold, week-over-week results are noise. Judge performance over 3–4 weeks minimum, not 7 days.
  5. Kill underperforming creatives faster. At low volume every impression matters. One weak creative dragging down the ad set costs more proportionally than at scale.
Diagnose in tool Read full guide Open problem page
Broad works when conversion signals are strong and creative is clear. Interests work as a training wheel when signals are weak or the message is ambiguous.
  1. Check conversion signal strength and volume
  2. Check if your creative clearly defines the buyer
  3. Use interests to constrain early learning only
Diagnose in tool Read full guide Open problem page
Separate retargeting is useful when you need tight control and budgets. Otherwise, modern delivery can self-retarget if you feed it strong signals.
  1. Check how much traffic you actually have to retarget
  2. If low volume, retargeting can starve and be noisy
  3. If high volume and varied funnel, separate retargeting can help
Diagnose in tool Read full guide Open problem page
Most accounts run too many ad sets and split signals. Fewer ad sets usually improves learning speed and stability.
  1. Start with 1 to 2 ad sets unless you have strong reasons
  2. Avoid splitting by small interests or tiny geo segments
  3. Consolidate once you have signal
Diagnose in tool Read full guide Open problem page
Advantage+ Audience is Meta's automated targeting that uses your inputs as suggestions and expands beyond them when the algorithm finds better conversion opportunities. For most accounts with clean conversion history, it outperforms manual interest and lookalike targeting.
  1. Check monthly conversion volume (below 50 limits expansion signal)
  2. Check pixel and CAPI Event Match Quality (below 5 degrades automation)
  3. Assess ICP size: above 100k people suits expansion, below 5k needs manual
  4. For B2B with narrow ICP: use manual targeting with precise filters
  5. To test: run both in parallel for 14 days, compare CPA not CTR
Diagnose in tool Read full guide Open problem page
The default answer is consolidate. Fragmented campaign structures keep each ad set below the conversion volume threshold the algorithm needs to optimise effectively.
  1. Count active ad sets and calculate average budget per ad set
  2. Check how many ad sets are in Learning or Learning Limited
  3. Calculate budget floor per ad set: target CPA x 50 / 7
  4. Check for auction overlap between campaigns
  5. Only split for genuine structural reasons: different objectives or conversion events
Diagnose in tool Read full guide Open problem page
The right optimization event depends on where your conversion happens. If the form is on Meta, optimize for Leads. If the conversion happens on your website, optimize for Conversions with a website conversion event.
  1. Determine where conversion happens: Meta instant form or website
  2. Check if pixel conversion event fires correctly on confirmation page
  3. Check current lead quality if using Lead Ads
  4. For Lead Ads: disable autofill, add qualifying question, use Higher Intent form
  5. For B2B: use website conversion in almost all cases
Diagnose in tool Read full guide Open problem page
Lead quality is a system problem. Most low-quality lead issues come from weak qualification, wrong optimization event, or missing feedback loops.
  1. Check form design and qualification questions
  2. Check optimization event and CRM feedback loop
  3. Check if creative is attracting the wrong intent
Diagnose in tool Read full guide Open problem page
Spam leads usually come from frictionless forms and incentives that attract non-buyers. Add friction and qualification signals.
  1. Switch to Higher Intent forms or add review step
  2. Add qualifying questions that buyers can answer
  3. Use reCAPTCHA and block obvious spam patterns
Diagnose in tool Read full guide Open problem page
Attribution, delays, deduplication, and tracking gaps create differences. The goal is consistency in decision-making, not identical numbers.
  1. Align attribution windows across systems
  2. Check deduplication rules and event priority
  3. Check CAPI setup and match quality
Diagnose in tool Read full guide Open problem page
Tracking issues are usually missing events, wrong domain configuration, consent issues, or CAPI and pixel not deduplicating correctly.
  1. Check Pixel Helper for event fires on key pages
  2. Verify Aggregated Event Measurement and domain verification
  3. Confirm CAPI deduplication and Event IDs
Diagnose in tool Read full guide Open problem page
Event Match Quality affects how well Meta can connect conversions to people. Better match improves optimization and measurement reliability.
  1. Check which parameters you send (email, phone, external_id)
  2. Ensure hashing and formatting are correct
  3. Improve CAPI coverage and deduplication
Diagnose in tool Read full guide Open problem page
The attribution window determines how long after an ad interaction Meta claims credit for a conversion. The default setting (7-day click, 1-day view) consistently over-reports conversions. For most advertisers, 7-day click only is more accurate.
  1. Check current attribution window in Ads Manager column settings
  2. Compare Meta-reported conversions to CRM data for same period
  3. Run breakdown showing click-through vs view-through conversions
  4. Assess typical purchase cycle length for your product
  5. For impulse purchases: use 1-day click; for B2B: use 28-day click
Diagnose in tool Read full guide Open problem page
When both the browser pixel and CAPI send the same conversion event without a deduplication mechanism, Meta counts the event twice. This inflates reported conversions and feeds the algorithm incorrect signal.
  1. Check Events Manager > Deduplication Diagnostics section
  2. Server events matched with browser should be above 80%
  3. Check event log for same event appearing twice
  4. Compare All Conversions to CRM order count (gap above 50% is unusual)
  5. Check for multiple pixel IDs or integration apps sending events
Diagnose in tool Read full guide Open problem page
More Meta guides
advantage plus underperforming
Meta guide
clicks not converting
Meta guide
cost cap vs bid cap
Meta guide
cpm spike
Meta guide
creative fatigue
Meta guide
how to scale meta ads
Meta guide
how to write ad copy
Meta guide
ios14 impact meta ads
Meta guide
learning limited
Meta guide
low ctr
Meta guide
low quality leads
Meta guide
meta ads attribution explained
Meta guide
meta ads for ecommerce
Meta guide
meta vs crm
Meta guide
performance drop overnight
Meta guide
scaling kills performance
Meta guide
stuck in learning phase
Meta guide
video vs static ads
Meta guide
what is capi
Meta guide
why cpa increased
Meta guide
Not finding your situation?
Some problems need account context.
The list covers the common cases. If yours isn't here, or the answer isn't enough, that's what the free intake call is for. If it's not fixable, I'll say that too.
Book the free intake