The Hourly ACoS
Benchmark Report.

This guide is not for everyone. Here is who it is for.

This guide is about hourly performance data for Amazon Sponsored Products campaigns — specifically, how to use that data to make smarter decisions about when your ads run. It is written for sellers and agency operators who are already running ads and want to understand whether the timing of their spend is costing them money.

It assumes you know what ACoS is, that you have campaigns actively running, and that you have at least 30 days of spend history behind you. If you are in your first two weeks of advertising, bookmark this and come back. The data you need to act on any of this simply does not exist yet.

One more thing before we get into it. A lot of content on dayparting presents generic overnight horror stories as if they apply equally to every account. They do not. The patterns in this guide are drawn from Perpetua's 2022 intraday study of anonymized Sponsored Products campaigns across the US marketplace — the most rigorous published research on this topic — and from industry benchmarks published by Atom11, IG PPC, and Adbrew. Your account will differ based on your category, your price point, your customer's shopping behavior, and your product's stage in its lifecycle. The point of this guide is to give you a framework to find your own numbers, not to hand you someone else's averages and call it a day.

Your 7-day ACoS is hiding something.

Amazon's default ACoS reporting is an average. Your 7-day ACoS, your 30-day ACoS, your campaign-level rollup — all averages. That is not a criticism of Amazon's reporting. Averages are useful. They tell you the direction you are trending, whether a campaign is broadly profitable, whether a keyword is working over time. What they cannot tell you is what is happening inside those averages at any given hour of the day.

Here is why that matters. A campaign running at 28% ACoS over 30 days looks healthy. But that 28% is the blended result of some hours where your ACoS runs at 11%, and other hours where it runs at 74% or higher — with almost no orders to justify the spend. When you average those together, the profitable hours mask the wasteful ones. The account looks fine. Nothing triggers a review. And every day, some portion of your budget runs through windows where it generates almost no return.

The 7-day average makes this invisible. The only way to see it is to pull the hourly data — which Amazon gives you, but most sellers never look at.

Before you do anything: how much data is enough?

The most common mistake with dayparting is acting on too little data. You pull one week of hourly performance, see that Monday at 2 AM had a 140% ACoS, and pause your ads every Monday at 2 AM going forward. That is not optimization — that is reacting to noise.

The minimum data window for reliable hourly pattern analysis is 14 days. The better window is 30 days. If you are in a seasonal category — outdoor furniture, holiday decor, anything with meaningful demand peaks — you want 60 to 90 days so you are identifying a pattern, not an anomaly from a slow weekend. Amazon limits each hourly report download to 14 days, so you will need two downloads to build a 30-day view. Combine them in a single spreadsheet before you start.

Here is what you are looking for once you have the data. You want hours where three things are true simultaneously: spend is meaningful, orders are low or zero, and ACoS is significantly higher than your account average. One condition alone is not a dead zone. Two is worth watching. All three is where you act.

You are also looking for the inverse — hours where spend is meaningful, conversion rate is high, and ACoS runs well below your account average. These are your peak windows, and they matter just as much as the dead zones. The goal of dayparting is not just to cut waste. It is to redirect budget from dead hours into peak hours. If you suppress overnight spend without reallocating it, your total sales volume will likely drop. The strategy is to shift, not simply reduce.

How to find your hourly ACoS data.

The report you need is the Hourly Sponsored Products report, made possible by Amazon Marketing Stream — Amazon's hourly data feed that launched in 2022 and has since become the foundation for all serious intraday optimization. It lives in the Amazon Ads console. The correct path as of 2025 is below. Note: some older guides send you to Bulk Operations. That is not where this report lives.

When you open the file, you will see one row per campaign per hour. All times are in Pacific Standard Time — Amazon's baseline for the US marketplace. So 0 = midnight PST, 6 = 6 AM PST, 18 = 6 PM PST. If most of your customers are on the East Coast, add three hours when interpreting the patterns. The four columns that anchor your analysis are:

Do not lead your analysis with CTR or CPC. Clicks are not the same as orders, and a high click-through rate in a dead zone hour just means you are paying for curious browsing that does not convert. Always anchor on orders and ACoS.

Reading the data without fooling yourself.

Once you have built a pivot table grouping by Start Hour, your data will immediately show some hours with alarming ACoS figures — 80%, 120%, sometimes higher. Before you flag every one of these as a dead zone, check the Spend column first.

If an hour shows 120% ACoS but only $0.40 in spend, that is three or four clicks that happened not to convert. That is noise, not a pattern. An hour is only worth acting on if it has meaningful spend behind it. As a general rule: if a specific hour accounts for less than 1% of your total account spend across the analysis period, the data is too thin to be reliable. You are making decisions based on statistical noise. Wait for the data to thicken — more spend, more time — before treating it as actionable.

The hours that are genuinely worth addressing are those with high spend, low orders, and consistently elevated ACoS across your full analysis window. "Consistently" is the key word. You want to see a pattern holding across multiple weeks, not a bad night during a slow period.

One important nuance: fewer people are shopping on Amazon at 3 AM than at 7 PM, which means overnight hours have fewer impressions, clicks, and orders in absolute terms — even when the ratio metrics look bad. A 90% ACoS at 3 AM might represent only $2 in actual spend. The same ACoS at 7 PM could represent $50. The ratio matters, but the volume behind it determines what it actually costs you. This is why you calculate your dead zone cost in dollar terms, not as a percentage. We will walk through that calculation shortly.

Once you have identified your genuinely problematic hours — high spend, low orders, elevated ACoS, consistent across weeks — calculate two numbers. First, the total spend in those hours across your analysis window. Second, the orders generated in those hours. This gives you a real cost-per-order for your dead zone, not a ratio. That absolute number is what you will use to size the opportunity.

The 7-day conversion heatmap.

The heatmap below shows general conversion patterns by hour and day across the US Amazon marketplace, based on Perpetua's intraday study. These are directional starting points — not your account data. Use them as a first hypothesis to validate against your own hourly report. Categories, price points, and customer bases all shift the picture meaningfully.

Two patterns hold broadly across the Perpetua study data. Weekday evenings — specifically 6 to 9 PM Pacific — show the lowest ACoS and highest conversion rates, ranging 20–23% in the study. The overnight window from midnight to roughly 5 AM shows the highest ACoS and lowest conversion rates. Saturday morning is the strongest single window of the week for most general merchandise categories.

What the study also shows — and this is the part most guides skip — is that CTR is actually higher overnight than during peak hours. More people click per impression. But they do not convert. This browsing-without-buying behavior is why CTR is the wrong metric to anchor dayparting decisions on. Always anchor on orders and ACoS. CTR can mislead you into thinking an hour is performing when it is not.

ACoS performance by time block

The chart below shows ACoS performance indexed against the day's average, based on the Perpetua study data. A score of 100 means that block performs at the daily average. Below 100 is more efficient than average. Above 100 is less efficient.

What your dead zone is actually costing you.

Most sellers think about dayparting in percentage terms — "my overnight ACoS is 80%, that seems bad." The percentage matters, but the number that actually changes behavior is the absolute dollar figure. Pull your spend and order data from your dead zone hours and plug your numbers in below.

Once you have your dead zone cost in dollar terms, you have a clear size for the opportunity. If your dead zone hours are generating $3,000/year in spend with almost no orders behind them, that is the budget available to shift into your peak windows. That is the number to bring into a conversation with a client, a finance team, or your own decision-making process about whether to invest time in setting up dayparting rules.

Will pausing ads overnight hurt my organic rank?

This is the most common objection to dayparting, and it deserves a straight answer rather than the hand-wavy reassurance most guides offer.

The short answer is: for established products with stable ranking, done correctly, dayparting does not hurt organic rank and can improve it. Here is why. Amazon's A9 algorithm factors in sales velocity and conversion rate as ranking signals. When you concentrate your ad spend in the hours where your conversion rate is highest, you are generating the same number of ad-attributed orders but with a higher conversion rate per impression. A higher conversion rate is a positive ranking signal. By not running ads in hours where you are getting clicks but no orders, you are removing the negative drag of a poor conversion rate from your account's performance signal.

The longer answer has an important caveat: this logic applies to established products. For products still in launch phase, overnight conversions — even expensive ones — contribute to the sales velocity and keyword ranking history you are still building. Pausing overnight too early in a product's life can slow that momentum. The cut-off is roughly 60 days of active advertising history, though this varies by category competitiveness.

The metric to watch alongside ACoS is TACoS — Total ACoS, which includes organic revenue in the denominator. As dayparting concentrates ad conversions in peak hours, conversion rate improves, ranking signals strengthen, and organic sales grow. As organic sales grow, TACoS falls even if ACoS stays flat. This compounding effect is what makes dayparting a longer-term play. Expect TACoS to lag by one to two weeks before showing improvement — it takes time for the 30-day rolling averages to reflect the new pattern.

What Amazon's native tools let you do — and where they stop.

Before you act on your hourly data, you need to understand what Amazon lets you control natively — because there is a critical limitation that most sellers discover only after trying to set up dayparting themselves.

Amazon introduced Bid Schedule Rules in late 2023. These allow you to set time-based bid adjustments for Sponsored Products campaigns. Here is what the rule actually does and does not support:

The practical implication of the native limitation is this: using Amazon's tools alone, you can amplify your peak hours but you cannot suppress your dead zone hours. You can push harder when things are working, but you cannot pull back when they are not. For accounts where the dead zone represents meaningful spend, this is an incomplete solution — and it is the core gap that third-party dayparting tools exist to fill.

What to do with the data. In this order.

Assuming you have at least 30 days of data, your product is past the launch phase, and you have identified meaningful dead zone hours, here is how to act. The sequence matters — doing action two before action one is how sellers accidentally cut total revenue while improving efficiency on paper.

What to watch each week

When not to daypart — and when to be careful.

A guide that does not cover the failure modes is incomplete. There are accounts and situations where dayparting is the wrong move, and applying it anyway will cost you.

Launch-phase products (first 60 days). Do not fully pause overnight on a product still building ranking. Overnight conversions — even expensive ones — contribute to sales velocity and keyword ranking history. If you want to reduce overnight waste during launch, use a 20–30% bid reduction rather than a pause. You stay in the auction for the rare overnight conversion without paying full price for it.

Thin data (under 14 days, or under 5 orders in a given hour block). If your analysis window is short or your hour-level order counts are in the single digits, you are making decisions on statistical noise. Wait for the data to thicken. This is especially true for lower-volume accounts where individual hours may have zero orders for days at a time by chance, not because of a structural pattern.

Highly seasonal categories. If you sell Halloween costumes or Christmas ornaments, your customer's shopping behavior in August looks nothing like October. Dayparting rules built on off-season data will be wrong during your peak selling period. Revisit your hourly analysis at the start of each major selling window and update rules accordingly rather than leaving a single rule set running year-round.

Over-fragmenting your rules. It is tempting, once you have hourly data, to build a different rule for every hour of every day. This level of complexity creates management overhead without proportional return. Start with one dead zone suppression rule and one peak hour budget rule. Add complexity only when you have data showing it will improve performance.

Dayparting done well is not about squeezing every possible optimization out of a single day's schedule. It is about finding the structural inefficiencies in your spend pattern — the hours where your budget reliably runs without returning — and systematically redirecting that budget to where your customers are actually buying. The data is available to every Amazon advertiser. Most of them never look at it.