Amazon PPC bid and budget rules you can copy, including stock- and margin-aware rules
Amazon PPC bid rules are if-then instructions that change a bid, budget or keyword when a condition is met, such as cutting a bid when ACOS stays above target for 14 days. Each needs a condition with enough data, an action with a capped step, and a guardrail such as a floor bid. Amazon's console covers schedules, events and ROAS goals; stock and margin rules need the Ads API.
- Amazon's schedule bid rules for Sponsored Products (since November 6, 2023; events added January 2, 2025) schedule bid increases, not decreases, on top of placement adjustments.
- Amazon's budget rules only raise daily budgets: on a schedule, or when a metric such as ROAS over the previous 7 days meets a threshold. Active rules are reviewed daily to set the next day's budget.
- Rule-based bidding steers Sponsored Products bids toward a ROAS target and can raise a bid up to 5 times the adjusted bid; schedule bid rules on that campaign go on hold while it is active.
- For sellers, Sponsored Products credits orders to clicks within 7 days (14 for vendors), conversions can take up to 12 hours to appear, and negative keywords take 72 hours to take effect.
- Amazon charges no additional fee to use the Amazon Ads API, and the sponsored ads report pages in the console shut down on December 31, 2026.
Many published Amazon PPC rule lists stop at the condition and the action: "if ACOS is above 40%, lower the bid." The part that decides whether a rule helps or hurts is the third one, the guardrail: how much data the rule needs, how far it may move a bid, and what it must never touch. This library gives all three for every rule, in one table you can copy into a spreadsheet, a PPC tool or your own code. It comes from building rule engines for clients, not from running a single account.
How do bid rules work on Amazon Ads?
A bid rule watches a metric over a lookback window and changes something when a threshold is crossed. It has three parts: a condition (metric, threshold, lookback and minimum data), an action (what changes, and by how much) and a guardrail (limits, timing and exclusions). In TRAKTOR, rule-based automation we built for campaign budgets, criteria were metrics such as click-through rate, conversion rate or ROAS; the lookback was today, yesterday, 7 or 14 days; rules ran hourly, daily or weekly; and each action moved a budget by an amount or a percentage within minimum and maximum limits. Each budget change went into an action log with the rule, the campaign, the old and new budget and the time.
Amazon runs some rules for you. Sponsored Products has had schedule-based bid rules, which raise bids for set hours, days and date ranges, since November 6, 2023, and the console added event-based bid rules for shopping events on January 2, 2025. Rule-based bidding adjusts base bids toward a ROAS goal, up to 5 times the adjusted bid, and budget rules raise daily budgets (all checked September 25, 2026).
Anything beyond those needs the Amazon Ads API, used directly or through a PPC tool built on it. Amazon says the API supports automated bid, keyword and budget optimization, with no additional fee from Amazon Ads (checked September 25, 2026). Your rules still sit underneath Amazon's bidding strategy: with dynamic bids up and down, Amazon may raise or lower a bid by up to 100% in each auction, so the number a rule sets is not the price you pay.
The rule library: condition, action and guardrail
How to read the table. CVR is the ad group's conversion rate (orders ÷ clicks) over the last 60 days, or the campaign's when the ad group has too few clicks. Lookbacks end 7 days ago because Amazon keeps crediting a seller's Sponsored Products orders to a click for 7 days. For vendors, Sponsored Brands and display campaigns the window is 14 days, so end those lookbacks 14 days ago (Ad campaign attribution, checked September 25, 2026). The percentages are starting points to tune per account, not benchmarks.
| # | Rule | Condition | Action | Guardrail | In Amazon's console? |
|---|---|---|---|---|---|
| K1 | Negate a wasted search term | Clicks ≥ 2 ÷ CVR and no orders, in the 30 days ending 7 days ago | Add it as a negative exact keyword in the ad group it came from | Never branded or protected terms; wait 72 hours before judging the result | No: search-term report plus bulk sheets or the API |
| K2 | Harvest a converting search term | 2 or more orders at or under target ACOS in 30 days, from an auto, broad or phrase ad group | Add it as an exact keyword in a manual campaign at the target bid; negate it as exact in the source ad group | At most 10 new keywords per run; skip SKUs below their stock throttle point | No |
| K3 | Trim an expensive keyword | ACOS above target over the 14 days ending 7 days ago, with clicks ≥ 1 ÷ CVR | New bid = bid × target ACOS ÷ actual ACOS | Cut of 20% at most per run; never below the floor bid; one change per 7 days | No (rule-based bidding steers a whole campaign to ROAS instead) |
| K4 | Push an efficient keyword | ACOS under 70% of target with 2 or more orders over 14 days, and the campaign stays in budget all day | New bid = bid × target ACOS ÷ actual ACOS | Raise of 15% at most per run; ceiling on the effective bid; skip low-stock SKUs | No |
| K5 | Wake a silent keyword | Fewer than 50 impressions in 14 days | Raise the bid 10% | Three raises at most, then flag it for a person; never above the ceiling | No |
| K6 | Cut a keyword that clicks but never sells | Clicks ≥ 2 ÷ CVR and no orders, in the 30 days ending 7 days ago | Cut the bid 30%, or pause the keyword if it is already at the floor | Check price, reviews and recent listing changes first; protected terms exempt | No |
| B1 | Feed a winner that runs out of budget | Out of budget on 3 of the last 7 days, with ACOS at or under target | Raise the daily budget 20% | Month-to-date spend stays inside plan; a person approves raises above an amount you set | Partly: performance-based budget rules raise budgets on ROAS |
| B2 | Starve a loser | ACOS above break-even over 14 days | Propose a 20% cut to the daily budget and list its top-spending targets for review | A person approves each cut; never below a floor budget | No: budget rules only raise budgets |
| B3 | Bid up in peak hours | An hour's conversion rate beats the campaign average by 30% or more in each of the last 4 weeks | Schedule a 10 to 20% bid increase for those hours | Pair it with a schedule-based budget rule so the budget lasts the day | Yes: schedule bid rules (Sponsored Products) |
| B4 | Bid down in dead hours | An hour converts at under half the campaign's rate, 4 weeks running | Lower bids for those hours and restore them afterward | Keep hourly data and schedules in one time zone; restore from saved bids even if a run fails | Not directly: lower the base bid and schedule increases for the hours you keep |
| B5 | Prepare for an event | Prime Day, Black Friday or your own sale dates | Schedule-based budget rule and an event bid rule for those dates | Check stock before the event; every rule gets an end date | Yes |
| B6 | Pacing alert | Spend by 2 p.m. (your reporting time zone) above 70% of the daily budget | Alert a person | Changes nothing on its own | No: needs hourly data |
| S1 | Throttle a low-stock SKU | Days of cover below lead time plus safety stock | Cut bids on that SKU's targets 30%; keep its top exact-match terms | Save settings and restore them when the shipment is received | No |
| S2 | Hold bids to break-even | ACOS above break-even (margin ACOS over 100%) for 14 days | Apply K3 with break-even ACOS as the target | Launch SKUs exempt on purpose; watch their TACOS instead | No |
| G1 | Circuit breaker | Account spend today reaches 150% of its average day | Stop all rule changes and alert a person | Only a person restarts the rules | No |
| G2 | Settle a conflict | Two rules want to change the same bid in one run | Apply the lower bid and log both | One change per target per run | No |
Three formulas drive most of the table:
Bid at target ACOS = target ACOS × conversion rate × average order value
New bid = current bid × (target ACOS ÷ actual ACOS), limited to a 20% cut or a 15% raise per run
Clicks before a no-sale rule acts = 2 ÷ conversion rate
The bid formula follows from ACOS = CPC ÷ (conversion rate × average order value), explained on our ACOS page: at a 25% target, a 10% conversion rate and a $30 order, the bid is $0.75. Treat that as the most you can afford to pay per click on average. With dynamic bids up and down, Amazon may raise a bid by up to 100% in an auction, so on those campaigns compare the result with the average CPC you actually pay, not only with the bid you set.
For the second, take a keyword bid at $1.20 running at 40% ACOS against a 25% target. The ratio is 0.625, a 37.5% cut, so the step limit holds it to 20% and the new bid is $0.96; the rule looks again a week later. For the third, at a 10% conversion rate K1 waits for 20 clicks without an order, and at 5% it waits for 40. Even a term that truly converts at 10% goes 20 clicks without an order about 12% of the time (0.9 to the 20th power), which is why the protected list matters.
Which rules should run on search terms and keywords?
Start with search terms. Negating obvious waste (K1) and moving proven terms into exact match (K2) are low-risk and easy to audit. Amazon allows negative phrase and negative exact keywords, up to 4 and 10 words respectively, and says negative keywords take 72 hours to take effect and negative products 96 hours (checked September 25, 2026). So K1 should leave a newly negated term alone for three days; checked sooner, it sees clicks that arrived before the negative took effect and flags a negative that is working as failed.
Sun Company, also rule-based automation we built, added a signal the ad reports do not carry: organic rank. Its rules moved a search term to negative when the product already ranked organically for it, and to positive when it did not. You can copy that rule if you have an organic rank feed. Try it on a handful of terms first and compare total sales, not ad sales alone, because the aim is fewer paid clicks without losing orders.
Keyword bid rules (K3 to K6) need patience. Amazon suggests letting a campaign run at least 14 days before reviewing it, and two weeks of data before setting a ROAS target for rule-based bidding. Conversions can take up to 12 hours to appear, and a day's attribution stays incomplete until its lookback window ends, so a rule reading the last three days will cut bids on keywords whose sales have not been credited yet. That is why K3 and K4 read the 14 days ending 7 days ago and change each target once a week at most.
How should budget and dayparting rules work?
Budgets behave differently from bids. Amazon treats a daily budget as a monthly average: by default a campaign can use money left over from quiet days to spend up to 100% over its daily budget on a busy day (25% if you change the setting), but never more than the daily budget times the days in the month. Amazon also says budgets are spent as quickly as possible, not paced through the day (checked September 25, 2026). So a spend cap in your own rules should watch month-to-date spend, and a pacing alert (B6) should watch the clock.
Amazon's budget rules raise budgets; none lowers one. Schedule-based rules add a percentage for an event or date range, and in the US, Canada, the UK, India and Japan they can run for set hours. Performance-based rules fire when a metric such as ROAS, measured over the previous 7 days, meets a threshold, and need a daily budget of at least $10. Rules are reviewed daily to set the next day's budget, and overlapping increases add up (checked September 25, 2026). Cutting a losing campaign's budget (B2) takes your own rule.
For B1, Amazon's average time in budget metric is a clean trigger: 100% means the campaign ran all day, and anything lower means it ran out of budget during the period (checked September 25, 2026). Pair it with ACOS, because a campaign that runs out of budget while losing money needs B2, not more budget. The same budgets tab shows Amazon's estimate of the sales, clicks and impressions a campaign missed while out of budget, which helps size the raise, as long as you treat it as Amazon's estimate rather than a forecast.
Dayparting starts with hourly data. The hourly campaign report covers the last 30 days, 14 days per download, and Amazon Marketing Stream pushes hourly metrics, and changes such as a campaign going out of budget, through the Ads API in near real time (checked September 25, 2026). Schedule bid rules raise bids on top of placement adjustments, and Amazon's own tip is to pair them with a schedule-based budget rule so the higher bids do not drain the budget. To spend less at night with Amazon's tools alone, lower the base bid and schedule increases for the hours you keep. Sun Company automated keyword bids at night in its own software instead.
What about rules that react to stock and margin?
Here are two short examples; the full playbook, with every stock tier, is on our page about inventory-aware ad spend. For S1, take a SKU with 20 days of cover and a 35-day wait before new stock is sellable. The rule cuts bids on its targets by 30%, keeps its top exact-match terms and restores the saved settings once the shipment is received. None of Amazon's built-in bid or budget rules reads stock levels, so this one needs your own software or a tool that reads inventory. To see what running out would cost that SKU, try the stockout cost calculator.
For S2, remember that break-even ACOS equals margin before ad spend. A $30 SKU that keeps $9 after Amazon fees and landed cost breaks even at 30% ACOS; a $30 SKU that keeps $6 breaks even at 20%, so the same 30% ACOS loses money on every ad-driven order. When margin ACOS stays above 100% for 14 days, the rule applies K3 with break-even as the target. Launch SKUs are exempt on purpose and judged on their TACOS trend. Teams that think in ROAS can use the same targets, since ROAS is 100 divided by ACOS.
This is the job of the margin-aware advertising module in Ecomsellertool Growth OS: "Ads that follow inventory and margin: budget moves to deep-stock SKUs and away from products about to stock out." Every row in this library is a rule: fixed thresholds and fixed actions. Some steps need a model instead, such as a sales forecast behind days of cover or a relevance check before negating a search term. Our advertising agent page shows which of its steps follow rules and which use a model.
What guardrails does every rule need?
- Enough data. A click minimum tied to conversion rate (1 ÷ CVR before a bid change, 2 ÷ CVR before a no-sale rule acts) and a lookback that ends after the attribution window closes.
- A step limit and a cooldown. A capped percentage per run, one change per target per 7 days, and 72 hours before judging a new negative keyword.
- A floor and a ceiling on the effective bid. Amazon allows placement adjustments up to 900%, and in its own bid adjustment example a $1 bid with a 50% top-of-search adjustment becomes $1.50, which dynamic bids up and down can take to $3. Audience and video adjustments stack on top (checked September 25, 2026). Cap the bid after every adjustment, not the base bid.
- A protected list. Branded terms, launch SKUs and campaigns a person manages by hand are never touched.
- One winner per conflict. Amazon adds schedule bid rules to each other and to placement adjustments; your own engine should pick one change per target per run, the safer one.
- An account-wide cap on month-to-date spend, with a circuit breaker (G1) that stops every rule.
- A log and an undo. Rule, target, old value, new value and time for every change, so anyone can see why a bid moved and reverse it.
- A trial period. New rules propose their changes to a person for a week or two before they act on their own.
Data plumbing is a guardrail too. Besides the attribution delays above, the data itself arrives late and in pieces. When we built TRAKTOR, Amazon's processing times ranged from 5 to 120 minutes, so we used staggered data fetching and prioritized critical updates. Any rule engine should record when each report arrived and skip a day or an hour whose data is still coming in, so that no rule acts on half a picture.
If your rules read reports scheduled in the console, Amazon is shutting down the sponsored ads report pages on December 31, 2026. Saved or scheduled reports not migrated to unified reporting will be deleted, and new reports and edits stop on December 17; Amazon says the dates may change. Its list of report types unified reporting does not support yet includes the Sponsored Products budget report, and top-of-search impression share is reported there at campaign level only (all checked September 25, 2026). Check that every field a rule reads survives the move.
Built-in Amazon rules or your own?
| What you need | Amazon's built-in tools | Custom rules through the Ads API |
|---|---|---|
| Raise bids at set hours, days or events | Schedule and event bid rules (Sponsored Products) | Yes |
| Lower bids at set hours | Not directly: lower the base bid and schedule increases for other hours | Yes, with saved bids restored afterward |
| Steer bids to a ROAS goal | Rule-based bidding, one ROAS target per campaign | Yes, with a target per SKU from its margin |
| Raise budgets | Schedule and performance budget rules | Yes |
| Cut budgets on campaigns above break-even | No: budget rules only raise budgets | Yes, with a person approving |
| Negate and harvest search terms | By hand or with bulk sheets | Yes (K1, K2) |
| React to stock, lead time and landed cost | No | Yes, with SP-API and purchasing data |
| What it costs | Free | The build; Amazon charges no API fee |
Amazon's tools are the better choice when your catalog is small, stock is steady and margins are similar across SKUs: its rules plus a weekly search-term review by hand cover the basics for free. Amazon's free Ads Agent, in beta, summarizes proposed changes for approval, but Amazon lists it for advertisers with access to Amazon Marketing Cloud and Multimedia Solutions with Amazon DSP, and the optimization skill it lists is for those DSP campaigns (checked September 25, 2026).
A PPC tool is the next step. Helium 10's pricing page lists rules-based advertising and dayparting in Helium 10 Ads on its Diamond plan, with a 2% fee on managed ad spend (checked September 25, 2026). A tool is the better fit when you want a rule builder with a screen and your rules use only ad data. Custom rules make sense when a rule needs data from outside Amazon Ads, such as stock across FBA and your own warehouse, lead times or landed cost: that is custom Amazon Advertising API automation.
Ecomsellertool builds these rules on your own Amazon Ads and seller accounts through the official APIs, starting from the advertising and reporting modules in Growth OS. We have built on Amazon's seller APIs since 2017. We write the rules and limits with your team, and each change can be logged with its old and new value. You keep your accounts, your data and the custom code we build; the Growth OS base is licensed to you. Below is one account's ad performance screen in Growth OS; it shows what the software tracks, not a result.
Ad performance per marketplace, against the previous period
See Growth OS ↗Where should you start?
Copy K1, B1 and G1 first. For two weeks, have each rule list what it would change, compare that with what your team did, then let it act within its limits. If you are not sure where your ad spend leaks, start with the free 24-hour diagnostic. You connect Amazon with Login with Amazon, no password is shared, and the report arrives within 24 hours of connecting, on business days. We only read data; we never change listings, prices, stock or ads. If you already know which rules you want built, schedule a call.
Frequently asked questions
How often should Amazon PPC bid rules run?
Daily is enough for bid and search-term rules, because Amazon credits a seller's Sponsored Products orders to a click for up to 7 days and conversions can take up to 12 hours to appear. Run hourly only for pacing alerts and dayparting checks, and change any one bid at most once a week so each change is judged on settled data.
How many clicks should a keyword get before a rule cuts its bid?
Tie it to conversion rate rather than a fixed number. One divided by the conversion rate is the number of clicks you would expect per order, so wait for at least that many before a bid change and twice that many before a no-sale rule negates or pauses anything. At a 10% conversion rate, that is 10 and 20 clicks. Even then, a term that truly converts at 10% goes 20 clicks without an order about 12% of the time, so keep branded and proven terms on a protected list.
Can Amazon's schedule bid rules lower bids at night?
No. Amazon describes them as scheduled bid increases for set hours, days, date ranges or events. To spend less at night with Amazon's tools alone, lower the base bid and schedule increases for the hours you want to keep. A custom rule through the Amazon Ads API can lower bids for set hours directly and restore them afterward.
Do bid rules conflict with Amazon's dynamic bidding?
They stack, which is why guardrails matter. Your rule sets the base bid, placement and audience adjustments raise it where they apply, and dynamic bids up and down may then raise or lower the result by up to 100% per auction. In Amazon's own example, a $1 bid with a 50% top-of-search adjustment becomes $1.50, and dynamic bidding can take it to $3. Rule-based bidding is different: while it is active on a campaign, that campaign's schedule bid rules are put on hold.
Should bid rules target ACOS or ROAS?
Either works, since ROAS is 100 divided by ACOS. What matters more is where the target comes from. One account-wide target treats a SKU with a 40% margin and a SKU with a 15% margin the same way. Set each SKU's target from its break-even ACOS, which is its margin before ad spend, and judge launch SKUs on TACOS instead.
What happens to rules that read downloaded Amazon Ads reports?
Amazon is shutting down the sponsored ads and Amazon DSP report pages on December 31, 2026, and saved or scheduled reports that are not moved to unified reporting will be deleted. Amazon plans to stop new reports and edits on those pages on December 17, 2026, and says the dates may change. If a spreadsheet or tool reads those scheduled reports, migrate the schedules now, check that every field the rule reads came across, or move the rules to the Amazon Ads API.
How we research, fact-check and compare: our editorial standards. Spot an error? Email hello@ecomsellertool.com and we will correct it.
