Cost cap and bid cap are Meta's two primary tools for controlling acquisition costs in CBO campaigns. Both limit what you pay for conversions, but they work fundamentally differently — and choosing the wrong one can tank your campaign performance or strangle your delivery.
This guide explains exactly how each bid strategy functions within CBO, when to use each one, and how to set optimal limits without killing your reach.
How Cost Cap Works
Cost cap tells Meta: "Try to get me conversions at or below this average cost." It's a soft limit that allows individual conversion costs to vary while maintaining your target on average.
The Mechanism
When you set a cost cap, Meta's algorithm adjusts bids dynamically to achieve your target average CPA:
- Some conversions may cost more than your cap
- Others cost less, balancing the average
- The system prioritizes volume while respecting average cost
- Learning is required to calibrate bidding effectively
Behavior During Learning Phase
Cost cap campaigns often exceed targets during learning as the algorithm tests different bid levels. According to Meta's bid strategy documentation, expect higher-than-target CPAs until the campaign exits learning phase (typically 50+ conversions).
Delivery Characteristics
Cost cap tends to deliver more consistently than bid cap:
- More stable daily spend patterns
- Fewer dramatic fluctuations in delivery
- Better suited for campaigns needing consistent volume
- May overspend during peak competition periods
How Bid Cap Works
Bid cap tells Meta: "Never bid more than this amount for any single conversion." It's a hard limit that prevents any bid above your specified maximum.
The Mechanism
With bid cap, Meta will not enter auctions where winning requires a bid above your cap:
- Every conversion costs at or below your cap
- No averaging — strict per-conversion limit
- Fewer auction opportunities when competition is high
- More predictable per-conversion cost but less predictable volume
Behavior During Learning Phase
Bid cap campaigns may struggle to exit learning phase if the cap is set too low. The algorithm has less room to experiment with bids, which can extend or permanently stall learning.
Delivery Characteristics
Bid cap delivery is more volatile:
- Spend varies significantly day-to-day
- Low-competition periods see high delivery
- High-competition periods may see near-zero delivery
- Excellent cost control but potentially inconsistent volume
Direct Comparison
Cost Predictability
- Cost cap: Average CPA predictable; individual conversions vary
- Bid cap: Maximum CPA guaranteed; but may not spend full budget
Volume Predictability
- Cost cap: More consistent daily volume
- Bid cap: Volume fluctuates with competition levels
Learning Phase Success
- Cost cap: More likely to exit learning successfully
- Bid cap: May stall in learning if cap too restrictive
Scaling Behavior
- Cost cap: Scales more smoothly with budget increases
- Bid cap: Scaling limited by auction availability at cap level
When to Use Cost Cap
Stable Volume Requirements
Use cost cap when you need consistent conversion volume. Ecommerce businesses with fulfillment capacity, lead gen with sales team capacity, or any business planning around predictable lead flow benefits from cost cap's delivery stability.
Target CPA with Flexibility
If your margins allow for some CPA variance as long as the average stays acceptable, cost cap provides that flexibility while meeting your targets.
During Scaling Phases
When scaling CBO campaigns, cost cap adapts better to increased budgets. The algorithm has more auction flexibility to find conversions at higher spend levels.
New Campaign Launch
Cost cap's flexibility during learning phase makes it better for new campaigns. Set it 20-30% above your target CPA initially, then tighten after learning completes.
When to Use Bid Cap
Strict Margin Requirements
When every conversion must be profitable — common in affiliate marketing — bid cap prevents unprofitable conversions entirely. You'd rather miss volume than acquire customers at a loss.
Mature, Stable Campaigns
Bid cap works best on campaigns with predictable performance history. The algorithm knows what works and can operate effectively within stricter constraints.
High-Competition Windows
During seasonal peaks (Black Friday, holiday shopping), bid cap prevents runaway costs when auction prices spike. You accept lower volume rather than overpaying.
Budget-Constrained Accounts
For low-budget accounts where every dollar matters, bid cap ensures no wasted spend on expensive conversions.
Setting Optimal Cap Levels
Cost Cap Guidelines
Start with these benchmarks:
- New campaigns: Target CPA x 1.3 (30% buffer for learning)
- Established campaigns: Historical average CPA x 1.15
- Scaling campaigns: Current CPA x 1.2 (buffer for increased competition)
Tighten gradually after exiting learning phase — reduce by 5-10% every 3-5 days as performance stabilizes.
Bid Cap Guidelines
Set bid caps based on unit economics:
- Breakeven cap: Revenue per conversion minus desired profit
- Conservative cap: Breakeven x 0.8 for profit margin
- Growth cap: Breakeven x 0.9 for volume with some profit
Never set bid cap below your historical average CPA — you'll get zero delivery.
Finding Your Starting Point
If you don't have historical data:
- Run lowest cost (no cap) for 2-3 days
- Note the average CPA achieved
- Set cost cap at that CPA level
- Or set bid cap at that CPA x 0.9 for stricter control
Common Mistakes
Setting Caps Too Low
The most common error. Caps set 20%+ below achievable CPA result in minimal or zero delivery. Your ads simply don't enter enough auctions to generate meaningful volume.
Frequent Cap Adjustments
Changing caps daily disrupts algorithmic learning. Each change triggers recalibration, preventing the system from stabilizing. Wait 3-5 days between adjustments.
Ignoring Delivery Signals
If your campaign isn't spending its budget with caps enabled, the cap is too restrictive. A campaign spending 50% of budget needs a higher cap or different strategy.
Same Cap for All Ad Sets
Different audiences have different acquisition costs. A retargeting ad set might achieve $15 CPA while prospecting costs $45. Setting a single $30 cap kills retargeting (over-constrained) and allows prospecting to overspend.
Hybrid Approaches
Cost Cap with Bid Cap Backstop
Some advertisers run cost cap campaigns with bid cap on specific ad sets that need stricter control. This provides volume optimization overall while protecting sensitive segments.
Different Caps for Different Stages
Apply different strategies based on funnel position:
- Prospecting: Cost cap for volume discovery
- Retargeting: Bid cap for efficient conversion
- Retention: Cost cap with aggressive target
Minimum ROAS Alternative
For revenue-focused campaigns, minimum ROAS can be more effective than either cost cap or bid cap. It optimizes for value rather than just volume.
Monitoring Cap Performance
Key Metrics to Track
- Actual vs target CPA: How close are you to your cap?
- Delivery rate: What percentage of budget is spent?
- CPA variance: How much do individual conversions vary?
- Learning status: Did you exit learning phase?
Warning Signs
Adjust your cap strategy when you see:
- Less than 70% budget delivery for 3+ days
- CPA consistently 30%+ below cap (cap too high)
- Stuck in learning phase beyond 7 days
- Wide variance in daily conversion volume (bid cap issue)
When to Remove Caps
Sometimes no cap delivers best results:
- During heavy testing phases when learning is priority
- When scaling aggressively and CPA can flex
- If caps consistently prevent full budget delivery
- For retargeting audiences with very high conversion rates
How ROASPIG Helps
Effective bid strategy depends on strong creative that converts efficiently. ROASPIG supports CBO optimization through:
- Creative Testing: Find high-performing creative that allows for tighter caps
- Conversion Optimization: Better creative means lower natural CPA
- Variant Production: Test multiple creative approaches to find efficiency
- Performance Tracking: Monitor creative contribution to CPA goals
- Fatigue Prevention: Fresh creative maintains efficiency over time
The Bottom Line
Neither cost cap nor bid cap is universally better — each serves different strategic needs:
- Cost cap for volume stability, learning flexibility, and scaling
- Bid cap for strict cost control, mature campaigns, and margin protection
Most advertisers should start with cost cap to build campaign learnings, then consider bid cap for segments requiring tighter control. The key is matching your bid strategy to your business constraints — not copying what worked for someone else.
Frequently Asked Questions About Cost Cap vs Bid Cap
Cost cap controls your average CPA across all conversions, allowing individual conversions to vary. Bid cap sets a hard maximum for any single conversion, preventing bids above that amount. Cost cap prioritizes volume with average cost control; bid cap prioritizes strict cost limits even if it means less volume.
Cost cap is generally better for new campaigns because it gives the algorithm more flexibility during learning phase. Set it 20-30% above your target CPA initially. Once the campaign exits learning and stabilizes, you can tighten the cap or consider switching to bid cap if you need stricter control.
Your bid cap is set too low for the current auction environment. The algorithm can't find enough auctions where it can win at your bid level. Increase your bid cap by 10-20% or switch to cost cap for more delivery flexibility. Check your historical CPA — your cap shouldn't be below this level.
Yes, you can set different bid strategies at the ad set level within a CBO campaign. This allows you to use cost cap for ad sets where you want volume flexibility and bid cap for ad sets requiring strict cost control. Just be aware that this adds complexity to optimization.
Wait at least 3-5 days between cap adjustments to allow the algorithm to stabilize. Frequent changes disrupt learning and prevent optimization. When you do adjust, make incremental changes (5-10%) rather than dramatic shifts. Only remove or significantly change caps when you have clear performance data supporting the change.