Amazon Sponsored Products looks simple. Choose keywords, set bids, launch campaigns, and wait for sales.
That simplicity is exactly why many sellers quietly lose money.
Amazon’s advertising platform will continue spending budget whether individual search terms convert or not. A campaign can generate clicks, impressions, and even sales while still producing disappointing profits.
The difference between a profitable account and an unprofitable one usually comes down to structure, measurement, and disciplined optimization.
Sponsored Products works best when you treat it as a system rather than a collection of bidding tricks. You need to know what a click is worth, which searches are producing that traffic, how campaigns should be separated, and when to move a search term from discovery into controlled targeting.
Here’s how to build that system.
Sponsored Products is Amazon’s pay-per-click advertising format for individual products. You choose the products you want to advertise, define how you want to target shoppers, set your bids and budget, and pay when shoppers click your ads. Amazon describes Sponsored Products as an auction-based CPC model in which your bid represents the maximum amount you are willing to pay for a click.
Amazon has historically described its Sponsored Products pricing as a generalized second-price auction. In simple terms, your bid is a ceiling rather than a guaranteed CPC, although the actual auction and pricing mechanics are more nuanced than simply assuming you will always pay one cent more than the next bid. Amazon’s current public guidance emphasizes the maximum bid, CPC pricing, and the role of relevance in determining ad delivery.
The important takeaway for advertisers is straightforward: your maximum bid controls how much you are willing to pay, but the bid alone does not determine whether your ad wins.
Your base bid is not always your final ceiling. Placement adjustments and dynamic bidding can raise the effective bid, so account for both when setting your maximum CPC.
Amazon also considers relevance and predicted performance when determining which ads appear. That means a higher bid does not automatically guarantee the best placement.
There is another eligibility requirement that can stop an otherwise active campaign from serving: the advertised product must be eligible for the Featured Offer, formerly known as the Buy Box. If the advertised SKU is not Featured Offer eligible, Sponsored Products ads will not generate impressions for that SKU.
Sponsored Products Placement Types
| Placement | Typical CPC Premium | Typical Conversion Rate | What the Placement Is Good For |
| Top of Search, First Page | Highest | Often highest | High-intent keywords and proven performers |
| Rest of Search | Moderate | Moderate | Scaling profitable search terms |
| Product Pages | Lower to moderate | Lower to moderate | Competitor and category targeting |
These are directional patterns, not fixed benchmarks. Actual CPC and conversion rates vary significantly by category, competition, product, keyword, price, listing quality, and shopper intent.
The practical lesson is to evaluate each placement against its own economics rather than assuming the most expensive placement is automatically the most profitable.
The best ACoS target is based on your break-even ACoS, not an industry benchmark.
ACoS, or Advertising Cost of Sales, measures advertising spend as a percentage of attributed sales:
ACoS = Ad Spend ÷ Ad Sales × 100
Your break-even ACoS tells you how much of a sale you can spend on advertising before the product stops generating contribution profit.
Break-even ACoS = (Sale Price − COGS − Amazon Fees − Other Variable Costs) ÷ Sale Price × 100
Other variable costs can include shipping, packaging, discounts, and expected returns. For example, suppose a product sells for $40. The table below includes all variable costs in its cost totals.
| Product | Sale Price | Total Variable Costs | Contribution Before Ads | Break-Even ACoS | Actual ACoS | Result |
| Product A | $40 | $20 | $20 | 50% | 30% | Profitable |
| Product B | $40 | $30 | $10 | 25% | 30% | Loss-making |
Both products have exactly the same 30% ACoS.
One is profitable. The other is not.
That’s why statements such as “a good Amazon ACoS is 30%” are almost meaningless without knowing the product’s margins.
Target ACoS vs. Break-Even ACoS
Your break-even ACoS tells you the ceiling for profitability. Your target ACoS is the number you actually want the campaign to achieve.
Those numbers do not always have to be the same.
A new product may intentionally run above its long-term target ACoS while it gathers conversion data and discovers relevant search terms. A mature campaign should generally aim to operate below its break-even ACoS if profitability is the goal.
The mistake is treating every campaign as though it should have the same ACoS target.
ACoS measures advertising efficiency by comparing ad spend with attributed ad sales.
TACoS, or Total Advertising Cost of Sales, compares ad spend with total sales, including organic sales.
TACoS = Ad Spend ÷ Total Sales × 100
ACoS tells you how efficiently your advertising generated attributed sales. TACoS helps you understand how advertising spend relates to the broader business.
A declining TACoS can be a positive signal when total sales are growing faster than ad spend, particularly when organic sales are growing alongside paid sales. It is not automatically positive on its own, though. Always look at total revenue and profit alongside it.
Neither should be the goal in isolation.
Think about Sponsored Products in three stages:
Launch: Accept more expensive traffic when the goal is to gather search-term and conversion data.
Growth: Increase spend on search terms and placements that convert within an acceptable target ACoS.
Harvest: Concentrate budget on proven terms, control waste aggressively, and protect contribution profit.
The right target changes as the product and campaign mature.
Sponsored Products gives you several ways to find and target shoppers.
| Campaign Type | What It Does | Best Used For | Typical Behavior |
|---|---|---|---|
| Automatic | Amazon matches your product to relevant shopping queries | Discovery and search-term research | Broad reach, less control |
| Manual Keyword | You choose broad, phrase, or exact keywords | Controlled search targeting | More control over reach and bids |
| Manual Product Targeting | You target ASINs or categories | Competitor and category traffic | More contextual, product-focused traffic |
The goal is not to choose one type and ignore the others. The strongest accounts typically use different campaign types for different jobs.
Use automatic campaigns primarily for discovery.
Amazon’s automatic targeting can match your product to shopping queries through four targeting groups:
Amazon recommends reviewing the performance of these targeting groups separately and adjusting bids based on their results.
Do not treat an automatic campaign as a set-and-forget sales machine. Its biggest value is the search and product data it generates.
Once you know which searches convert, move those opportunities into more controlled campaigns.
A broad match gives you the widest reach. The shopper’s query can contain related terms, variations, synonyms, and words in different combinations. For example, a broad keyword such as running shoes could potentially match searches such as “athletic sneakers” or “tennis trainers.” It would not be limited to searches containing the exact words “running shoes.”
Phrase match is more restrictive. The keyword’s words generally need to appear in the same order, although shoppers can add words before or after them. For example, running shoes could match “women’s running shoes” or “running shoes for men,” but would not normally capture a search such as “shoes for running.”
Exact match gives you the tightest control. The shopper’s search needs to closely match the keyword, including the same words and order, with Amazon allowing certain close variations. Exact match tends to produce higher conversion rates because the targeting is more precise, but it also limits reach.
A useful way to think about them is:
Broad discovers → Phrase refines → Exact controls
Product targeting lets you show Sponsored Products ads against specific ASINs or categories rather than relying on keywords.
It can work particularly well when your product gives shoppers a clear reason to choose it over the product they are currently viewing.
That advantage might come from a more competitive price, stronger reviews, a useful feature, a better value proposition, or a better fit for the shopper’s needs.
Product targeting is not necessarily a volume play. It is often about finding specific product pages where your offer can compete effectively.
Campaign structure determines how useful your data will be.
If you combine unrelated products, match types, and targeting strategies in one campaign, you make it harder to understand what your budget is actually buying.
Use One Match Type Per Campaign
For a clean optimization architecture, keep broad, phrase, and exact match in separate campaigns.
This is not an Amazon requirement. Amazon allows advertisers to test multiple match types for the same keyword, and its own guidance notes that advertisers will not bid against themselves simply by using different match types.
Separating them is an optimization choice.
It gives you cleaner control over budgets, bids, negatives, and performance data. When you see that a campaign generated $500 in sales, you know whether that result came from broad discovery, phrase targeting, or exact targeting.
Group Campaigns by Product, Not Convenience
Avoid putting a dozen unrelated SKUs into one campaign simply because it is easier to manage.
Ideally, each campaign should contain one product or a tightly related variation family with similar economics.
If one product has a 50% contribution margin and another has a 20% contribution margin, they should not share the same profitability assumptions.
Build a Three-Tier Campaign Architecture
A simple structure looks like this:
Discovery → Validation → Performance
Discovery: Automatic + broad match
Find new search terms and product targets.
Validation: Phrase match
Test promising terms with more control.
Performance: Exact match
Concentrate budget on proven searches.
This structure turns your campaigns into a progression rather than a collection of disconnected targets.
Use Consistent Campaign Names
A campaign name should tell you what is inside without opening it.
A useful naming format is:
Product | Targeting | Match Type | Goal
For example:
Coffee Grinder | Manual | Exact | Performance
or
Coffee Grinder | Auto | Discovery
Good naming becomes increasingly important as your catalog and ad account grow.
Keep ad groups focused enough that the budget and data remain useful.
A 200-keyword ad group may technically work, but it becomes difficult to determine which targets are actually responsible for spend and sales. High-performing keywords can effectively subsidize weak ones without making the problem obvious.
Amazon’s own getting-started guidance has recommended starting manual campaigns with at least 30 keywords, but that does not mean every campaign needs hundreds.
For ongoing optimization, prioritize a manageable set of tightly related keywords rather than filling an ad group simply to hit a number.
It depends on your catalog size, budget, and how different your products are.
For a small catalog, one product per campaign often gives you useful control without creating an overwhelming number of campaigns.
For a large catalog, creating a separate campaign for every SKU can create unnecessary management overhead.
The real question is:
Do these products need different budgets, bids, profitability targets, or optimization decisions?
If the answer is yes, separate them.
The best keyword research often comes from your own advertising data.
Start with these sources:
The important distinction is between keywords that look relevant and search terms that actually produce profitable sales.
Your account data should eventually win that argument.
The Search Term Report shows the customer searches that generated at least one ad click. Amazon specifically recommends using it to identify high-performing searches and create negative keywords or product targets for searches that do not meet your goals.
Review it in two directions.
First, find spend with no orders.
Sort by spend and look for search terms that have accumulated meaningful click volume without generating a sale.
Second, find orders with acceptable ACoS.
Look for search terms that generated sales at or below your target ACoS.
Then take action:
The Search Term Report should drive your optimization decisions rather than simply confirm what you already believe.
Keyword harvesting means taking a search term that has demonstrated performance in a discovery campaign and promoting it into a more controlled campaign.
For example:
Automatic campaign → “stainless steel coffee grinder” → 3 orders at target ACoS → Exact campaign
A practical harvesting rule is to look for a search term with at least two or three attributed orders, sufficient click volume, and ACoS at or below your target.
The exact threshold should reflect your product price and conversion rate. A $20 product and a $200 product do not generate enough evidence at the same spend level to use identical rules.
The point is to base the move on actual conversion evidence rather than promoting every search term that happens to receive a click.
Start with Amazon’s suggested bid range as a reference, not as an instruction.
A suggested bid tells you something about the competitive environment. It does not know your contribution margin, target ACoS, or maximum profitable CPC.
Your economics should determine your ceiling.
Sponsored Products Bidding Strategies
| Strategy | What It Does | When It Can Make Sense | Main Risk |
| Dynamic bids, down only | Amazon can lower bids when a conversion is less likely | New campaigns, tighter budgets, discovery | May limit reach |
| Dynamic bids, up and down | Amazon can raise or lower bids based on predicted conversion likelihood | Proven campaigns with enough conversion data | Can increase spend quickly on high-opportunity auctions |
| Fixed bids | Keeps the bid you set rather than dynamically adjusting it | When you want tighter control for testing | Less responsive to auction conditions |
Amazon says dynamic bids up and down can increase bids when an auction appears more likely to convert and lower them when conversion is less likely. Amazon also recommends reviewing these campaigns about every two weeks and limiting changes during testing so you can identify which change affected performance.
Use Placement Adjustments Carefully
Amazon allows placement adjustments, including top-of-search adjustments.
This can be a better lever than increasing the keyword bid across every placement.
For example, if an exact-match keyword converts profitably at the top of search but performs poorly on product pages, you may want to concentrate additional bid pressure where the keyword already works rather than increasing its base bid everywhere.
The decision should come from placement-level performance.
A useful starting formula is:
Maximum CPC = Target ACoS × Conversion Rate × Sale Price
Suppose:
Then:
0.25 × 0.10 × $40 = $1.00
Your starting maximum CPC would be approximately $1.00.
This is a planning formula, not a guarantee of actual CPC or profitability. Your conversion rate can change as traffic changes, and your target ACoS must still sit within your product’s break-even economics.
Use it when you have enough conversion data to justify giving Amazon more room to adjust your bids.
It can be useful for proven campaigns where additional exposure is valuable and the economics support it.
The failure mode is using aggressive bidding on a campaign that has not demonstrated which traffic converts. If Amazon increases bids on marginal opportunities, you can spend faster without improving profitability.
For discovery campaigns, tighter bid control is often more useful while you are still learning.
Profitable advertising is partly about finding good traffic.
It is equally about stopping bad traffic.
Use negative keywords when a search term is irrelevant or has accumulated enough evidence that it is not meeting your campaign objective.
Amazon recommends evaluating a keyword after it has received at least 20 clicks before deciding whether to add it as a negative target.
A practical profitability rule is to also consider your target CPA.
For example, if your target CPA is $30, a search term that has received 20 or more clicks and has spent roughly $30 without an order deserves a serious review.
That does not mean every term should be automatically negated at exactly $30. Product price, conversion rate, margins, and click volume all matter. Check the attribution window, too: orders can be credited after the click date, so the newest clicks may still generate attributed sales.
You have two primary negative keyword options:
Negative exact: Blocks the specific search term and close variations defined by Amazon’s matching rules.
Negative phrase: Blocks searches containing the complete negative phrase or close variations.
Use negative exact when the problem is a specific search. Use negative phrase when the unwanted concept appears across multiple searches.
A campaign with spend and no sales can have several different problems.
The search terms are irrelevant.
Check the Search Term Report and add appropriate negatives.
The listing does not convert.
Check the main image, title, bullets, price, reviews, offer, and product detail page.
The match type is too broad.
A broad keyword can generate traffic that is related to the keyword but poorly aligned with your actual product.
The price is uncompetitive.
Advertising cannot fix an offer that shoppers consistently reject.
The product is not Featured Offer eligible.
If the advertised SKU loses Featured Offer eligibility, Sponsored Products will stop generating impressions.
The answer is not always “lower the bid.”
First identify where the funnel is breaking.
Ad cannibalization happens when paid advertising captures shoppers who may already have found and purchased your product organically.
That does not mean every branded ad or organic-ranked keyword is wasteful. Paid visibility can still defend branded searches, capture incremental traffic, or improve total sales.
The question is whether advertising is creating incremental value or simply replacing sales you would have received anyway.
Compare:
If ad spend increases while total sales barely change, investigate whether paid traffic is replacing organic demand rather than adding to it.
A simple weekly optimization routine can prevent most accounts from drifting.
1. Search Term Report: 15 Minutes
Find:
2. Budget-Capped Campaigns: 5 Minutes
Look for campaigns regularly hitting their daily budgets.
A budget cap is not automatically a problem. First check whether the campaign is profitable enough to justify additional spend.
3. Placement Performance: 10 Minutes
Compare:
Look for meaningful differences in CPC, conversion rate, sales, and ACoS.
4. Proven-Term Bids: 10 Minutes
Review bids on high-performing keywords and targets.
Review them weekly, but do not automatically change them every week. Amazon recommends reviewing dynamic bids about once every two weeks and limiting changes while testing.
5. Negation Pass: 10 Minutes
Look for irrelevant or persistently unprofitable search terms and decide whether they should be negated, bid down, or investigated further.
Judge campaigns based on click and conversion volume, not simply the number of days they have been running.
A campaign that receives 10 clicks has very little evidence. A campaign that receives 100 or more clicks gives you substantially more information, although there is no universal click threshold that makes a campaign statistically conclusive.
As a practical operating guideline, many advertisers want roughly 100 or more clicks before making a major judgment. The appropriate threshold still depends on product price, expected conversion rate, target CPA, and the amount of spend you can reasonably tolerate while testing.
Amazon’s own guidance also emphasizes allowing campaigns enough time to gather data before making major targeting decisions.
The important rule is simple:
Do not pause a keyword because it did not convert after three clicks.
Less often than most advertisers do.
Daily bid changes make it difficult to understand whether a campaign is improving because of your optimization or because of normal auction and demand fluctuations.
Review performance weekly.
Make meaningful bid changes when the data supports them.
For dynamic bidding campaigns, Amazon recommends reviewing bids about every two weeks and limiting changes during testing.
Amazon’s advertising console gives you several reports and management tools, but you do not need to stare at every metric every day.
| Report | Question It Answers |
|---|---|
| Search Term Report | What did shoppers actually search before clicking? |
| Targeting Report | Which keywords, products, or categories are performing? |
| Placement Report | Where are my ads converting? |
| Advertised Product Report | Which advertised products are generating sales? |
| Budget Information | Which campaigns are constrained by budget? |
| Bulk Operations | How can I make large-scale campaign changes efficiently? |
The Search Term Report is particularly important because it connects your targeting decisions to actual customer searches. Amazon notes that the report contains search terms that generated at least one ad click and can be used to identify both high-performing searches and negative-targeting opportunities.
These ad formats solve different problems.
| Ad Type | Primary Focus | Typical Use |
| Sponsored Products | Individual product visibility | Capture high-intent shoppers and drive product sales |
| Sponsored Brands | Brand and product portfolio visibility | Build brand presence and promote multiple products |
| Display ads (formerly known as Sponsored Display) | Audience and product-context targeting | Reach shoppers based on browsing and shopping behavior |
Sponsored Products is usually the most direct place to establish whether your individual products can acquire customers profitably.
Once you understand your product economics and have a reliable Sponsored Products structure, you can make a more informed decision about expanding into other ad formats.
1. Mixing Match Types Without a Clear Reason
You lose control over budgets and optimization when broad, phrase, and exact targeting are thrown together without a clear architecture.
2. Never Adding Negative Keywords
Discovery without cleanup eventually becomes expensive.
3. Judging Campaigns Too Early
A handful of clicks is not enough evidence to declare a keyword a winner or loser.
4. Bidding Without Knowing Break-Even ACoS
A bid is only sensible when you know what the resulting traffic can afford to cost.
5. Advertising a Listing That Does Not Convert
Ads bring shoppers to the product page. They cannot fix poor pricing, weak creative, low trust, or a weak offer.
6. Letting Automatic Campaigns Run Unmanaged
Automatic campaigns are useful for discovery. They become expensive when you collect the search-term data but never act on it.
There is no universal starting budget.
Your budget should be large enough to generate meaningful click and conversion data without putting your product’s profitability at unreasonable risk.
Start with the amount you can afford to use for testing, then increase the budget when profitable campaigns demonstrate that they can absorb additional spend.
No. Sponsored Products is available to eligible sellers and vendors without requiring Brand Registry, although eligibility requirements apply. Amazon states that advertised products must be eligible for the Featured Offer and meet other marketplace and category requirements.
Sponsored Products can increase visibility and generate sales, but you should not treat paid advertising as a guaranteed direct ranking mechanism.
The useful business question is whether advertising generates incremental sales and profitable demand.
Track organic sales, paid sales, total sales, and TACoS together rather than assuming every paid order directly translates into an organic-ranking benefit.
CPC can change as auction conditions, competition, placement mix, shopper demand, and your own bidding strategy change.
Check:
Do not respond automatically by raising every bid.
First determine why the CPC changed and whether the additional traffic is actually converting profitably.
Yes, an eligible product can generally be advertised even if it has no reviews.
The bigger issue is conversion.
If shoppers click your ad and see a product with little social proof, they may be less likely to purchase. That can make your advertising economics difficult even though the campaign itself is technically eligible.
A profitable Sponsored Products account is not built by finding one perfect bid.
It is built by creating a system where every part of the account answers a useful question like:
“What is this traffic worth?”
“Where did it come from?”
“Did it convert?”
“Can I buy more of it profitably?”
“Should I stop paying for it?”
Start with your break-even ACoS. Build campaigns that keep targeting and products separated enough to produce useful data. Use automatic and broad campaigns to discover opportunities, then move proven searches into tighter targeting. Review the Search Term Report consistently. Negate waste. Adjust bids based on evidence rather than daily fluctuations.
Most Sponsored Products accounts do not have a bidding problem.
They have a visibility problem inside their own data.
A free 30-minute Amazon advertising audit can show you which search terms are costing you money, where your campaigns are wasting spend, and what your actual break-even ACoS is. We’ll show you where the waste is coming from and which changes can help eliminate unprofitable traffic before it consumes more of your advertising budget.
