The Amazon Marketing Stream Playbook: From Daily to Hourly Ads Optimization
Amazon Marketing Stream (AMS) delivers your Sponsored Ads data in near-real time — hourly, not daily. This guide explains what AMS is, why hourly data changes how advertisers bid and pace budgets, and how AMALYZE's 24/7 optimization engine turns the firehose into measurable PPC gains.
What is Amazon Marketing Stream?
Amazon Marketing Stream is a push-based API from Amazon Ads that streams Sponsored Products, Sponsored Brands, and Sponsored Display performance data to advertisers in near-real time. Instead of pulling reports once a day, AMS pushes hourly datasets directly to a subscriber endpoint — usually within minutes of the hour closing.
For Amazon Sellers and Vendors running serious PPC, that single change — daily to hourly — rewires how campaigns can be optimized. Bids, budgets, dayparting, and negatives can now react to what's actually happening on the marketplace today, not to yesterday's averages.
Why hourly data beats daily reports
Traditional Amazon Ads reporting collapses a day into a single row: spend, impressions, clicks, sales, ACOS. That hides the truth — performance inside a 24-hour window swings dramatically by hour, weekday, device, and placement.
- Pacing is broken without hourly dataMany campaigns blow through their daily budget by mid-afternoon and miss the highest-converting evening hours entirely.
- Bids react too slowlyA keyword that suddenly stops converting today won't show up in daily reports until tomorrow — by then you've wasted a full day of spend.
- Promotions and trends are invisibleHour-level data exposes Prime Day spikes, lightning-deal traffic, and viral demand windows that daily totals smooth away.
Datasets AMS delivers
AMS publishes several hourly and event-driven datasets. The ones that matter most for optimization:
A four-step hourly optimization playbook
- 01Detect intra-day pacing risk
Watch budget-usage hour over hour. If a campaign is on track to exhaust budget before 6pm, raise budget for high-converting campaigns and cap losers — before you go dark in the best hours.
- 02Shift bids by daypart and weekday
Cross hourly conversion data with day-of-week to find the 10–20% of hours that drive most of your sales. Lift bids inside those windows and pull them back during graveyard hours.
- 03React to search-term drift in hours, not days
Hourly conversion data surfaces a non-converting search term within hours of it appearing. Harvest converting terms into exact-match campaigns and negate the rest the same day.
- 04Defend during demand spikes
When AMS shows a surge in clicks on a key ASIN (a viral moment, a deal, a competitor going out of stock), raise bids and budgets immediately — and pull them back the moment the surge ends.
How AMALYZE's 24/7 Optimization uses AMS
AMALYZE consumes the Amazon Marketing Stream the moment it's published and feeds it into a continuous optimization loop. Instead of one batch job per night, the engine re-evaluates bids, budgets, and placements every hour, 24/7.
- Budgets are re-paced every hour based on live conversion velocity, not a daily forecast.
- Bid rules read the same hour's CVR and ACOS, so cold keywords aren't punished for one slow morning — and hot ones are pushed immediately.
- Dayparting schedules are learned per ASIN and per marketplace from weeks of hourly data, then re-validated continuously.
- Alerts trigger on intra-day anomalies — a spend spike, a sudden CTR drop, a campaign hitting cap — instead of being discovered the next morning.
Getting started with AMS
You have two paths to use Amazon Marketing Stream: build it yourself, or plug into a tool that already consumes it.
Apply for AMS access via Amazon Ads API, set up AWS SQS subscriptions, build a warehouse for hourly datasets, and write your own optimization rules. Powerful, but a multi-quarter engineering investment.
Connect your Amazon Ads account once. AMALYZE handles the AMS subscription, data pipeline, and hourly optimization rules — you keep full control of strategy and guardrails.
Frequently asked questions
Yes — AMS access itself is free for advertisers eligible for the Amazon Ads API. You only pay for the infrastructure that consumes and stores the stream.
Reports are pull-based and aggregated daily. AMS is push-based and delivers hourly datasets within minutes of the hour closing — enabling intra-day decisions.
If you spend less than a few hundred dollars per month, daily reports are usually enough. Once intra-day pacing matters — typically beyond ~$1k/month per marketplace — hourly data starts paying for itself.
No. AMALYZE manages the AMS subscription and pipeline. You connect your Amazon Ads account, choose your guardrails, and the 24/7 optimization runs on your data.
Tactics from the AMALYZE Sponsored Success episode
Beyond Basic Dayparting: Why Static Hourly On/Off States Fail
General "dayparting"—simply switching campaigns on or off during specific hours—is a relic of legacy API limitations. The speaker emphasizes that this method is fundamentally flawed for the Amazon ecosystem. If you disable your ads at 10:00 AM because historical data suggests low conversion, you effectively "blind" your data set and break the customer discovery funnel (awareness, consideration, purchase).
A static hourly schedule fails to account for:
- Time Zone Discrepancies: Server-side updates during daylight savings transitions (summer vs. winter) often cause automation to activate at the wrong hour if specific offsets aren't handled manually.
- Event-Driven Spikes: Disabling ads on a Tuesday evening because it is "traditionally slow" ignores external triggers like major sporting events or local holidays (e.g., "Tanz in den Mai") where shopping behavior shifts suddenly.
- Data Sufficiency: Optimizing purely by the hour can lead to "bid gambling." If a target has high impressions but no sales in a specific 60-minute window, a basic tool might crash the bid. However, if those sales are attributed later, you’ve lost a high-performing slot.
The Three Axis Optimization Framework
Effective utilization of Amazon Marketing Stream data requires moving away from one-dimensional bid changes. AMALYZE utilizes a three-axis model to ensure bids are mathematically sound rather than reactive.
1. Retention & Attribution Windows (The X-Axis)
The Marketing Stream returns four distinct data points: 1-day, 7-day, 14-day, and 30-day attribution.
- Fast-Moving Goods: For products like diapers, a 1-day attribution model is often superior. If the customer needs diapers, they buy now; including 30-day "re-purchase" data in your bid calculation can artificially inflate your target ACOS.
- High-Involvement Goods: For vacuum robots or expensive tools, customers often research for days. Narrowing the window too much results in "invisible sales" where you pay for the click but don't credit the ad, leading to incorrectly lowered bids.
2. Observation Periods (The Y-Axis)
Instead of daily reports, data flows hourly. This allows for "sliding" observation windows. If you set a 14-day observation period, the system doesn't just evaluate once a day; every hour, it takes the last 14 days + 1 hour. This prevents the "hockey stick" effect (extreme outliers in bid changes) by smoothing data across hourly increments.
3. Granularity at the Target Level (The Z-Axis)
Optimization must happen at the individual target level (the specific keyword, ASIN, or auto-match type), not at the campaign or portfolio level. Managing every bid individually based on its specific hourly performance avoids the trap of "successful" campaigns subsidizing "wasted" ad spend on failing keywords.
Solving the "Hockey Stick" Problem in Search Term Reports
A common phenomenon in Amazon PPC is the "Hockey Stick" graph: a few keywords performing at the target ACOS, while a large tail of targets has either exorbitant ACOS or wasted spend with zero conversions.
To audit this using Marketing Stream data:
- Export your search term report.
- Isolate individual targets and map their hourly performance.
- Look for "Infinite ACOS" zones—hours where click spend accumulates consistently without ever converting.
- Tactic: Instead of pausing these keywords, use the Marketing Stream's granular data to lower the bid only during the hours where the "waste" occurs, or extend the observation window to 90 days for slow-turning products to see if a conversion eventually justifies a lower, steady bid.
Strategic Attribution for Hybrid Sellers
For Sellers and Vendors operating hybrid accounts, Amazon Marketing Stream provides a critical point of synchronization. Traditionally, Sponsor Products use different default attribution windows (7 days vs. 14 days).
By leveraging the Stream’s specific attribution return values, advertisers can:
- Standardize reporting across both Seller and Vendor Central.
- Set a "Unified ACOS" goal based on the same attribution timeframe (e.g., 14 days) across all accounts.
- Actionable Workflow: If you notice a 20% discrepancy between 7-day and 14-day attribution in certain categories, adjust your bid ceiling to account for those "late" conversions that standard Seller Central reports might miss.
The Prime Day "Peak & Valley" Workflow
Static bidding is most dangerous during high-traffic events like Prime Day. The speaker notes that the real value of hourly data is not just in "bidding up" when traffic rises, but in the "cool down" phase.
- Real-Time Reaction: As conversion rates climb during the first hour of Prime Day, the system detects the shift and raises bids to match the increased probability of a sale.
- The Exit Strategy: Most manual advertisers leave high bids active for 24 hours after the event ends, wasting budget on "traffic hangovers." Hourly stream data allows the system to detect the moment the peak ends, immediately pulling bids back down to baseline levels to protect the ACOS of the following day.
- Leveraging Historical Peaks: For seasonal events, you can apply "static" timeframes from previous years (e.g., Easter data from 2023) to your current hourly bid strategy as a baseline for the upcoming season, provided your API connection has been active long enough to warehouse that data.
Sponsored Success: Amazon Marketing Stream
The full AMALYZE Sponsored Success episode on Amazon Marketing Stream (German).
Turn Amazon Marketing Stream into hourly wins.
See how AMALYZE's 24/7 optimization engine uses AMS to re-pace budgets, adjust bids, and harvest search terms every hour — across every campaign you run.