Every account we take over has the same fossil layer: a set of Amazon PPC daily budgets that were chosen once, eighteen months ago, by someone who picked a round number that felt comfortable. $50. $75. $100. Nobody has touched them since, and nobody thinks of them as a strategic setting.
They are. And not in the way most sellers assume.
The obvious cost of a capped campaign is the sales you didn’t make after it went dark. That’s real, and it’s the version everyone talks about. The expensive version is quieter: a campaign that runs out of budget stops spending, but it also stops sampling. Every number that campaign reports to you afterward — ACOS, CVR, CPC, conversion rate by keyword — was measured on a biased slice of the day. You then use those numbers to make decisions.
That’s the part that costs real money, because a bad decision made on contaminated data doesn’t stay in one campaign.
What Actually Happens When a Campaign Caps Out
Amazon doesn’t hand you a clean shutoff at a predictable hour. Delivery smoothing means spend gets throttled as you approach the cap, then serving degrades, then the campaign shows “out of budget” and stops entirely until midnight in your account’s time zone.
Three consequences follow, and only the first one is widely understood.
One: you lose the hours you didn’t choose. A campaign that exhausts at 1pm gave away the entire afternoon and evening. In most consumer categories that’s where conversion rate peaks — the buyer who browsed at 9am from a desk comes back at 8pm from the couch and buys. You didn’t decide to skip that window. Your budget decided for you.
Two: the throttling happens before the shutoff. As the cap approaches, you’re not competing at full strength in auctions you’d otherwise win. You aren’t out of the market, you’re just weaker in it, and that shows up as declining impression share that looks like a competitive problem rather than a budget problem.
Three — the one nobody accounts for: your data is now a morning sample. If your campaign only ran from midnight to 1pm, then your ACOS for that campaign is the ACOS of the first thirteen hours of the day. Not your ACOS. Your morning ACOS.
The Contamination Problem
Here’s why point three matters more than the missed sales.
Mornings and evenings on Amazon are not the same market. CPCs, competitive density, and conversion rates all move through the day, and they move differently by category. A campaign that only ever runs before 1pm is reporting performance from a systematically different auction environment than an uncapped campaign running the full 24 hours.
Now put those two campaigns next to each other in a weekly report and compare their ACOS.
That comparison is meaningless, and we watch teams make decisions on it constantly. Consider two Sponsored Products campaigns in the same account:
- Campaign A, uncapped, runs all day, reports 31% ACOS
- Campaign B, capped at $60, exhausts at 12:30pm daily, reports 24% ACOS
The obvious read is that B is the efficient one and deserves more money. Sometimes that’s true. But it’s equally possible that B looks efficient because it only ever competed in the cheaper morning auctions, and that the moment you fund it properly and it starts running into the expensive afternoon, its ACOS converges on A’s.
You cannot tell which from the report. The capped campaign has never been observed running a full day, so you have no read on its actual all-day economics. You’re grading a student who only ever sat the easy half of the exam.
The reverse case is just as common and more damaging. A capped campaign that burns its budget in the morning at genuinely poor efficiency gets paused for “high ACOS,” when what it needed was a bigger denominator and a full day of data.
To be clear about scope: the fix for when you spend is dayparting, and that’s a separate lever with its own logic. This post is about how big the number is and what it does to your ability to read the account at all. Dayparting on top of a cap you set arbitrarily is optimizing a distorted picture.
The Paradox We See in Audits
Raise a capped campaign’s budget by 25% and ACOS improves.
It happens often enough that new account managers treat it as a fluke. It isn’t magic and it isn’t always true, but the mechanism is straightforward: the campaign was previously buying only its cheapest, earliest clicks and stopping — and those early clicks were disproportionately low-intent browse traffic. Funding the full day let it reach the evening buyers who convert.
The pattern doesn’t hold everywhere. In some categories the morning genuinely is the efficient window and raising the cap makes ACOS worse. That’s a legitimate finding and it’s what dayparting is for. The point is that you can’t know which case you’re in until the campaign has been observed running unconstrained for at least a week.
Which gives you a clean diagnostic: any campaign that has never run a full day at full strength is a campaign you don’t have real numbers on.
The Metric Almost Nobody Pulls
Amazon will tell you how much this is costing you and most sellers never look.
In the campaign manager, the budget column flags out-of-budget status, and Amazon surfaces estimated missed impressions, clicks, and sales for budget-constrained campaigns along with a recommended budget. Treat the recommendation itself with suspicion — Amazon’s incentive is for you to spend more, and the recommended number reliably runs hot. Treat the constraint flag as reliable, because that one is just a fact about your account.
The number we actually work from is out-of-budget hours per week.
Our working thresholds, from managing hundreds of accounts:
- 0-3 hours/week: healthy. The cap is a genuine safety rail and it’s barely binding.
- 4-14 hours/week: worth investigating. Usually fine on a deliberately capped discovery campaign, a problem anywhere else.
- 15-40 hours/week: the campaign’s reported metrics are unreliable. Fix the budget before you touch a single bid.
- 40+ hours/week: you are running a part-time campaign and calling it a full-time one. Every conclusion drawn from it is suspect.
That last tier is more common than it sounds. A campaign going dark at 1pm every day is out of budget roughly 77 hours a week.
The Three Legitimate Uses of a Budget Cap
Budget caps aren’t bad. They’re just misassigned. There are exactly three jobs a cap does well:
A research ceiling on discovery. Auto and broad campaigns exist to buy information. Capping them is how you decide what you’re willing to pay for that information. Here, running out of budget is the system working — you bought your data allowance and stopped.
A hard stop on conquest. ASIN-targeting and competitor conquest is the tier most likely to run a 40% ACOS while feeling strategic. A cap is the discipline that keeps an ambitious tactic from quietly becoming a line item.
A risk cap on anything brand new. A campaign with no history gets a small budget until it has enough data to earn a bigger one. Standard practice, no argument.
And one illegitimate use, which is where most of the damage lives:
Using budget to control ACOS. This is the big one. When ACOS climbs and someone cuts the daily budget in response, they haven’t made the campaign more efficient — they’ve made it shorter. Efficiency is the bid’s job, and the keyword’s, and the listing’s. Budget controls exposure. Using it as an efficiency lever produces a campaign that looks better and sells less, and it contaminates the data you’d need to actually diagnose the problem.
The 30-Minute Weekly Routine
This is the cadence we run, and it’s deliberately boring.
1. Pull the budget-constrained list. Filter for campaigns flagged out of budget in the last 7 days. Sort by spend.
2. Sort them into two piles. Deliberately capped (discovery, conquest, new) versus accidentally capped (your core non-branded and branded campaigns, which should almost never be constrained). The second pile is your work.
3. For each accidentally capped campaign, raise the budget by 20-30%. Not 100%. You want the campaign to run a full day, not to blow the account’s spend profile in a week. If it’s still capping after seven days, raise it again.
4. Wait a full week before judging. Then compare ACOS to the previous week — but only for campaigns that actually ran unconstrained. This is the first honest read you’ve had on them.
5. Change one lever at a time. Do not raise the budget and adjust bids in the same week. You’ll get a number that moved and no idea why, and you’ll spend the following month arguing about it.
6. Fund from somewhere real. Budget increases should come out of campaigns that are genuinely underperforming on a full-day sample, not out of thin air. If the account has a total spend ceiling, this is a reallocation exercise, not an increase.
The one caveat worth stating: if a campaign is capping out and running at genuinely terrible efficiency on a full-day sample, raising the budget makes it worse faster. The sequence matters — establish the honest read first, then decide whether to fund or fix.
A Note on Q4 Budgets Set in July
One timing point, because we’re at the end of July and this decision is being made now for a lot of brands.
Q4 CPCs run 30-50% above baseline in most categories. A daily budget is a dollar figure, not a click figure. Which means a budget you set in July silently becomes a much tighter cap in November — same dollars, fewer clicks, more hours dark, on the highest-intent traffic of the year.
Combine that with the peak fulfillment surcharge running October 15 through January 14, and both sides of your contribution math move against you at once.
Two things to do about it now: set your Q4 budgets as a percentage of contribution margin rather than as a carried-forward dollar figure, and diary a budget review for early October rather than discovering the problem on Black Friday morning when your hero campaign goes dark at noon.
Frequently Asked Questions
Should I just uncap everything?
No. Uncapped campaigns with loose bids are how accounts lose money quickly. The correct posture is that budget controls exposure and bids control efficiency — set budgets so your core campaigns can run all day, and manage cost through bids, match types, and negatives.
Does going out of budget hurt my campaign long-term?
The direct effect is the lost hours and the sampling bias. The indirect effect is more interesting: a campaign that goes dark daily on rank-feeding keywords surrenders sales velocity, and velocity is a ranking input. Chronic capping on the terms you’re trying to rank for costs you organic position on top of the paid sales.
What about budget rules?
Rule-based budget adjustments are useful for scheduled events and performance-triggered increases, and they’re worth setting up before Q4. They don’t fix an undersized baseline budget — a rule that raises a bad number by 20% on Prime Day still leaves you with a bad number for the other 360 days.
How do I know if my budget is right?
The simplest test: is the campaign spending its full budget while staying inside its target efficiency? If it caps out and stays efficient, it’s underfunded. If it never reaches its cap, the budget isn’t binding and isn’t your problem. If it caps out and runs inefficiently, the budget isn’t the issue — the bids, terms, or listing are.
Is this the same as dayparting?
No. Dayparting decides which hours you compete in on purpose. Budget sizing decides whether you get to compete at all past a certain point. Get the budgets honest first, so that when you do daypart, you’re making the decision from real full-day data instead of from a morning-only sample.
—
Most accounts we audit are carrying five or six core campaigns that have never once been observed running a full, unconstrained day. Every optimization decision made on those campaigns for the last year was made on partial-day data — and the team making them had no idea, because nothing in the interface tells you that your ACOS is a morning number.
Fix the budgets first. Then the rest of your optimization work starts producing conclusions you can actually trust.
If you’re looking for a team that manages every lever — creative, advertising, and operations — Velocity Sellers works with brands doing $100K+/month on Amazon. Contact us for a free account audit.
