I see this question come up constantly in ecommerce communities: "I switched from cost cap to bid cap on Meta Ads and my CPA went through the roof. What happened?"
It happened to a client recently. They had a solid cost cap campaign running at about $40 CPA, $80 AOV, spending $1,000 a day. Decent numbers. But they wanted to push for even better results (we've all been there), so they switched to bid cap.
Here's what the first three days looked like:
Day 1: $27 CPA. Brilliant. Except they hadn't excluded existing customers, which likely inflated the numbers.
Day 2: $45 CPA after excluding existing audiences. Still reasonable, close to the bid cap.
Day 3: $255 CPA. More than four times the bid cap. Budget blown.
Naturally, panic set in.
The whole point of a bid cap is supposed to be that it acts as a hard limit, right? You tell Meta "I will not pay more than X for a conversion" and that should be the end of it.
Except it doesn't always work that way.
The Difference Between Cost Cap and Bid Cap
Before we address why bid caps can overspend, let's quickly clarify what each strategy actually does.
Cost cap tells Meta: "Try to keep my average cost per acquisition around this number, while maximising volume." It's a target, not a ceiling. Meta will attempt to find conversions at or below your cap, but it has flexibility to go over sometimes if it believes the overall average will balance out.
Bid cap is supposed to be stricter. It tells Meta: "This is the maximum I'm willing to bid for any single conversion." In theory, it's a hard limit on each individual bid rather than an average target.
The key word there is "supposed to be."
Why Bid Cap Campaigns Can Still Overspend
Here's what actually happens: during the learning phase, Meta's algorithm will sometimes push beyond your bid cap.
I know that sounds contradictory. You set a hard limit and Meta ignores it? But there's logic behind it, even if it's frustrating.
When you launch a new bid cap campaign (or make significant changes to an existing one, like adjusting your audience), Meta enters a learning phase. During this time, the algorithm is gathering data to understand which audience segments convert best, which placements work, and what combinations deliver results.
To gather this data quickly, the algorithm sometimes bids more aggressively than your cap. It's essentially overpaying for conversions in the short term to learn faster and (in theory) deliver better long-term results.
The frustrating part? You're the one paying for that learning.
What the Data Actually Shows
From experience running bid cap campaigns across multiple accounts, here's what I've observed:
Yes, CPAs can spike dramatically in the first few days. Seeing a $50 bid cap result in $100+ CPAs during the learning phase isn't unusual.
However, there's an important distinction: when bid cap campaigns overspend on a per-conversion basis, they typically don't spend your entire daily budget. The algorithm seems to recognise that something is off and pulls back on volume.
Over about two weeks, the CPA usually falls back within the bid cap range. The algorithm learns which opportunities are genuinely convertible at your target price and focuses spend there.
In the case I mentioned earlier, the Day 3 spike was alarming but expected. The client had also adjusted their audience on Day 2, which essentially reset the learning phase. So Day 3 was really only the second day of learning with the new targeting.
The Two-Week Rule
My general advice for bid cap campaigns: give them a full two weeks before making major judgements.
This doesn't mean you should ignore wild overspending. If your CPA is 10x your bid cap for days on end, something is fundamentally wrong, likely with your targeting, creative, or the bid cap itself being unrealistic for your offer.
But moderate overspending in the first three to five days? That's normal. The algorithm needs time to optimise, and cutting it off too early means you never find out whether bid cap would have worked for you.
When to Use Bid Cap vs Cost Cap vs Highest Volume
This brings up the bigger question: when should you use each bidding strategy?
Highest Volume (Lowest Cost) is Meta's default. You set a budget and Meta spends it all, aiming to maximise conversions regardless of cost. This requires the most attention because costs can creep up without warning, but it can deliver slightly higher ROAS in some accounts because there's no artificial ceiling limiting Meta's ability to compete for conversions.
Cost Cap is a middle ground. You're telling Meta your target CPA, but it has flexibility to exceed that on individual conversions as long as the average stays in line. This works well for accounts that need some cost control but don't want to sacrifice too much volume.
Bid Cap is the strictest option. It's more hands-off once it's working because Meta won't bid above your limit (post-learning phase). But it can also severely limit your volume if your cap is too aggressive for your market.
My preference? Running a mix of strategies in most accounts. Bid cap campaigns provide predictability and peace of mind. Highest volume campaigns (with close monitoring) often deliver the best overall returns but need more babysitting.
How to Make the Transition Smoother
If you're planning to move from cost cap to bid cap, here are some practical steps:
Don't change everything at once. If you switch to bid cap and simultaneously adjust your audience, creative, or offer, you have no idea what's causing any changes you see. Switch the bid strategy first, let it stabilise, then make other adjustments.
Set a realistic bid cap. If your cost cap campaigns were averaging $40 CPA, don't set a $25 bid cap and expect miracles. Start at or slightly above your current average and tighten later if results allow.
Monitor daily spend, not just CPA. During the learning phase, watch whether Meta is spending your full budget. If CPA is high but spend is low, the algorithm is being cautious. If CPA is high and you're hitting full budget, that's a bigger concern.
Exclude existing customers from the start. The Day 1 results in our example were artificially good because existing customers weren't excluded. These people already know you, so converting them is easier and cheaper. Including them skews your data and makes the "real" CPA on Day 2 onwards look worse by comparison.
Run bid cap and cost cap in parallel. You don't need to go all-in on one strategy. Test bid cap with a portion of your budget while maintaining your cost cap campaigns. Compare results over a month, then allocate more budget to whichever performs better.
What to Do If Overspending Continues
If you're past the two-week mark and your bid cap campaign is still consistently exceeding your cap, you have a few options:
Lower the bid cap further. This might seem counterintuitive since you're already overspending, but a lower cap forces Meta to be more selective. Volume will drop, but the conversions you do receive should be closer to your target.
Check your audience size. Bid caps struggle with audiences that are too narrow. If there aren't enough people who match your targeting and are likely to convert at your price, Meta has no choice but to either overspend or deliver almost nothing.
Review your creative. Poor-performing ads mean Meta needs to show them to more people to find conversions, which drives up costs. Refreshing creative can sometimes bring CPAs back in line without touching bid settings.
Accept that your bid cap might be unrealistic. Sometimes the market simply doesn't support the CPA you want. If competitors are willing to pay $60 for a customer and you're capped at $40, you'll struggle to win auctions for quality prospects.
The Bottom Line
Switching from cost cap to bid cap isn't a magic solution for lowering your CPA. It's a different approach to bidding that offers more predictability once optimised, but requires patience during the learning phase.
Expect some overspending in the first week. Give it two weeks before making major changes. And consider running multiple bid strategies simultaneously so you can compare results and reduce risk.
Bid cap can absolutely work, but it needs time to learn just like any other Meta campaign.
Conclusion
The transition from cost cap to bid cap campaigns on Meta Ads often catches advertisers off guard when CPAs spike dramatically in the first few days. The key points to remember:
Bid cap is meant to be a hard limit, but Meta's algorithm can exceed it during the learning phase as it gathers optimisation data
Allow at least two weeks for a bid cap campaign to stabilise before judging performance
Making audience changes resets the learning phase, so avoid stacking multiple changes at once
Consider running bid cap and cost cap campaigns in parallel to compare results and reduce risk
If overspending persists beyond two weeks, investigate audience size, creative quality, and whether your bid cap is realistic for your market