How to Set Up an Amazon PPC Campaign From Scratch

Setting up an Amazon PPC campaign is technically easy.

The difficult part is setting it up so you can understand what is working later.

A new campaign should not only generate traffic. It should give you enough control to answer:

• Which products are getting the spend?
• Which searches or targets are producing sales?
• What traffic is wasting budget?
• Which bids need to move?
• Which campaigns deserve more budget?
• What should be separated as the account grows?

Amazon's Sponsored Products setup lets advertisers select products, define targeting and bidding, set a daily budget and campaign dates, and launch the campaign. Sponsored Products supports automatic, manual, and negative targeting.

But the campaign builder is only the starting point.

The real work is deciding what to put into it.

Before You Create the Campaign

Do not open Campaign Manager and start clicking immediately.

Check the product first.

Product readiness

Review:

• Product title
• Main image
• Bullet points
• A+ Content, where applicable
• Price
• Reviews
• Rating
• Featured Offer status
• Inventory
• Product variation structure
• Promotions

Advertising can increase traffic.

It cannot fix every retail problem.

If the product is overpriced, frequently unavailable, poorly presented, or converting badly, sending more paid traffic can simply make the problem more expensive.

Amazon also recommends reviewing product detail pages, inventory, and pricing before launching campaigns for new products.

Step 1: Define the Campaign Objective

Before choosing keywords, decide what the campaign is supposed to do.

Common objectives include:

Profitability

You want advertising to produce sales within a defined economic range.

Growth

You are willing to invest more aggressively to capture additional demand.

Product launch

You need traffic and search-term data for a new ASIN.

Brand defense

You want to protect branded search demand or your own product traffic.

Discovery

You want to find search terms, products, or audiences worth targeting more deliberately.

The same product can need different campaigns for different objectives.

That is why "one campaign per product" is not a complete strategy.

Step 2: Choose the Right Campaign Type

For most sellers starting from scratch, Sponsored Products is the logical starting point because it can place individual products in Amazon shopping experiences and gives you automatic, manual, and negative targeting options.

Amazon's current Sponsored Products creation flow is:

1. 1Select Create campaign
2. 2Choose Sponsored Products
3. 3Add the products
4. 4Define targeting and bidding
5. 5Set campaign settings and daily budget
6. 6Launch

Sponsored Brands and display advertising solve different problems and should not simply be added because the account has them available.

Start with the campaign type that matches the objective.

Step 3: Decide Which Products Go Into the Campaign

This is one of the most important decisions.

Do not group products simply because they belong to the same category.

Ask:

• Are the products similar in price?
• Do they have similar conversion rates?
• Do they have similar margins?
• Do they target similar search intent?
• Do they need similar bids?
• Do they have the same business objective?

If the answer is mostly yes, grouping may make sense.

If not, separate them.

For example:

The practical test is:

If I change the bid or budget, am I comfortable applying that decision to every product in this campaign?

If not, create more separation.

For a larger catalog, see How to Manage Amazon PPC for Multiple Products.

Step 4: Name the Campaign for the Operator

Campaign names are not marketing copy.

They are an operating system.

A name should tell you what you are looking at without opening the campaign.

For example:

US | SP | Product-A | Auto | Discovery
US | SP | Product-A | Exact | Core
US | SP | Product-A | Phrase | Research
US | SP | Product-A | PT | Competitors

You can adapt the naming system to your account.

The important part is consistency.

A good naming structure should make it possible to filter and understand:

• Marketplace
• Ad type
• Product or ASIN
• Targeting type
• Campaign role

Do not build a naming system that looks sophisticated but takes five minutes to understand.

Step 5: Choose Automatic or Manual Targeting

Sponsored Products gives you three important targeting approaches:

• Automatic targeting
• Manual targeting
• Negative targeting

Automatic targeting lets Amazon match your product to relevant shopping queries and products. It can also provide search-term and targeting insights for later manual campaigns.

Manual targeting gives you control over the specific keywords or products you want to target.

Negative targeting lets you exclude traffic you do not want.

Amazon notes that the targeting strategy selected when creating a Sponsored Products campaign cannot simply be switched later. If you want to change from automatic to manual, you create a new campaign.

That makes the initial structure worth thinking about.

Step 6: Build an Automatic Campaign for Discovery

If you have little historical PPC data, automatic targeting is a practical way to start collecting information.

Amazon's automatic targeting uses four targeting groups:

• Close match
• Loose match
• Substitutes
• Complements

The important point is not to treat the automatic campaign as a permanent dumping ground.

Use it as a discovery engine.

You want to learn:

• Which search terms generate clicks?
• Which terms generate orders?
• Which product targets produce sales?
• Which traffic consumes spend without converting?

Then use those findings to build more controlled campaigns.

Amazon specifically positions automatic targeting as a way to discover search trends and identify targets that can later be added to manual campaigns.

Step 7: Build Manual Targeting When You Have a Target List

Manual targeting is where you decide exactly what you want to buy.

You can target:

Keywords

Useful when you know the search terms relevant to the product.

Products

Useful when you want to target specific ASINs or product categories.

For keyword targeting, think about the role of each match type rather than simply adding the same keyword everywhere.

For example:

Exact
"leather office bag"

Phrase
"leather office bag"

Broad
"leather office bag"

The actual structure should depend on how much control you need and how much search-term data you have.

The point is to make your targeting architecture understandable.

Step 8: Build the Initial Keyword List

Do not start with 500 keywords just because a keyword tool produced them.

Start with terms that have a clear relationship to the product.

A useful keyword set can include:

Core product terms

What the product actually is.

Feature terms

What makes the product different.

Use-case terms

What the shopper wants to accomplish.

Audience terms

Who the product is designed for.

High-intent variations

Searches that indicate a stronger purchase intent.

Then separate:

• Core terms
• Research terms
• Brand terms
• Competitor terms
• Irrelevant terms

The goal is not maximum keyword count.

The goal is relevant traffic with enough control to learn from the results.

Step 9: Set the Initial Bids

A bid is not your target CPC.

Your base bid is the amount you are willing to bid for the click. Amazon may adjust the bid depending on the bidding strategy and auction context.

Amazon provides suggested bids and bid ranges in the campaign creation workflow. These are useful starting references, not a guarantee of profitability.

Start with the economics.

Suppose, using USD as an example:

• Selling price = $50
• Contribution before advertising = $20
• Conversion rate = 10%

At a simplified level, a $1 CPC means approximately:

10 clicks × $1 = $10 spend

At a 10% conversion rate, those 10 clicks produce about one order.

The advertising cost per order would therefore be approximately $10 before accounting for variation and attribution.

If conversion falls from 10% to 5%, that same $1 CPC would cost roughly $20 per order.

This does not give you a perfect launch bid.

It gives you a way to understand what CPC your economics can tolerate.

For more on bid decisions, see What Is Bid Optimization on Amazon.

Step 10: Choose the Bidding Strategy

Do not select a bidding strategy simply because the name sounds aggressive or conservative.

Think about what you are trying to control.

You may be trying to:

• Capture more traffic
• Protect efficiency
• Test demand
• Scale a proven target
• Reduce downside on weak traffic

Amazon offers different dynamic bidding options for Sponsored Products, allowing bids to be adjusted based on the likelihood of a conversion. The available options and settings can vary by campaign type and marketplace.

For a new campaign, keep the setup understandable.

You need to know:

What bid did I set?

What traffic did it buy?

What happened after the click?

That is more important than creating a complicated bidding system on day one.

Step 11: Set the Daily Budget

Your budget determines how much opportunity the campaign has to participate in traffic.

Do not set it only from the question:

"How much am I comfortable spending?"

Also ask:

"How much traffic could this campaign realistically use?"

And:

"What happens if the campaign performs well?"

For example, a campaign with a $10 daily budget can become difficult to evaluate if it regularly reaches the limit before collecting enough traffic.

At the same time, a $500 budget does not make sense simply because the account can afford it.

Budget should reflect:

• Product economics
• Expected traffic
• Campaign objective
• Search demand
• Conversion rate
• Current performance
• Scaling capacity

Amazon publishes marketplace-specific recommended minimum daily budgets. Treat those figures as starting references, not profitability targets.

Step 12: Decide Whether to Add Negative Targets

Do not wait months to think about negative targeting.

If you already know a search term or product is irrelevant, exclude it.

Examples:

• Wrong product type
• Wrong size
• Wrong audience
• Incompatible use case
• Irrelevant brand
• Clearly unrelated competitor product

For new campaigns, avoid building an enormous negative list based on assumptions.

Start with obvious exclusions.

Then let real search-term data tell you where additional negatives are needed.

Step 13: Check the Product Before Launching

Before clicking launch, run this checklist.

Product

• Correct ASIN selected
• Product is available
• Price is competitive
• Featured Offer is eligible
• Listing is ready
• Inventory is sufficient

Campaign

• Correct marketplace
• Correct product
• Correct campaign type
• Clear campaign name
• Correct targeting
• Correct bids
• Correct budget
• Correct start date
• No accidental end date

Targeting

• Relevant keywords or products
• Match types make sense
• Obvious negatives added
• No accidental duplicate structure

Then launch.

Step 14: Do Not Optimize the Campaign After 10 Clicks

This is where many new advertisers create noise.

A campaign launches.

A few clicks arrive.

No order appears.

The bid gets reduced.

Then another five clicks arrive.

The bid gets increased again.

Now the campaign has changed before enough evidence existed to understand it.

Instead, establish a review threshold based on the product and expected traffic.

You want enough information to evaluate:

• Impressions
• CTR
• Clicks
• CPC
• Spend
• Orders
• Conversion rate
• Sales
• ACoS
• Search terms

There is no universal number of clicks that makes every decision valid.

The right threshold depends on the product, price, conversion rate, traffic volume, and objective.

Step 15: Review Search Terms Before Changing Everything

Once the campaign has enough traffic, search-term data becomes one of your most useful inputs.

Look for:

Winners

Search terms generating:

• Orders
• Strong conversion
• Acceptable ACoS
• Meaningful sales

Waste

Search terms generating:

• Significant spend
• No orders
• Poor conversion
• Irrelevant traffic

New opportunities

Search terms that:

• Were not in your original keyword list
• Have strong relevance
• Are generating clicks
• Show early signs of conversion

The workflow becomes:

Discover → Validate → Isolate → Scale

For example:

Auto campaign
     ↓
Search term generates sales
     ↓
Validate economics
     ↓
Add to controlled manual campaign
     ↓
Give it an appropriate bid
     ↓
Monitor separately

This is more useful than treating automatic targeting as a campaign you simply leave running forever.

Step 16: Separate Traffic Problems From Conversion Problems

Suppose a campaign spends $500 and produces no sales.

The first question should not be:

"Should I lower the bid?"

Ask:

Are we getting the wrong traffic?

Check:

• Search terms
• Targets
• CTR
• Product relevance

Is the traffic relevant but not converting?

Check:

• Price
• Reviews
• Rating
• Product page
• Featured Offer
• Inventory
• Promotions

Is CPC too expensive?

Check:

• Bid
• Competition
• Placement
• Targeting

Different causes require different actions.

A lower bid does not fix a product page that cannot convert.

Step 17: Build a Simple First-Campaign Structure

For a single product, a practical starting structure can be:

Product A

SP | Auto | Discovery

SP | Exact | Core

SP | Phrase | Research

SP | Product Targeting | Competitors

You do not need every campaign on day one.

If you have no meaningful keyword or competitor data, start simpler.

The structure can expand as the account produces evidence.

The goal is to avoid two extremes:

Too simple

One campaign contains everything, making optimization difficult.

Too complicated

Twenty campaigns are created before there is enough data to justify them.

Step 18: Create a Change Log

This is a small operational habit that becomes extremely useful later.

Record:

Without a change log, you eventually look at a performance chart and ask:

"Why did this change?"

A campaign history gives you context.

Common Setup Mistakes

1. Starting with keywords instead of the product

The campaign exists to advertise a product.

Make sure the product can convert before trying to maximize traffic.

2. Putting unrelated products together

Shared budgets and bids can hide product-level differences.

3. Using one bid for everything

Different targets can have very different economics.

4. Setting a tiny budget on a high-demand campaign

You may never collect enough data to evaluate the campaign properly.

5. Setting a huge budget because you want sales

Budget does not create profitable demand.

6. Adding hundreds of keywords immediately

More targets do not automatically mean better traffic.

7. Changing bids every day

You can end up optimizing noise rather than performance.

8. Ignoring search terms

The keyword list tells you what you intended to target.

Search-term data tells you what shoppers actually searched.

9. Treating ACoS as the only KPI

ACoS does not explain traffic quality, conversion, product margin, inventory, or business priority.

10. Building an account that cannot be managed later

Campaign structure should make reporting and optimization easier as the account grows.

A Practical 7-Day Launch Review

The first review should focus on whether the campaign is collecting useful information.

Check traffic

• Are impressions coming in?
• Are clicks coming in?
• Is CTR reasonable for the product and placement?

Check cost

• Is CPC within a sensible range?
• Is spend concentrated in a few targets?
• Is the campaign reaching its budget?

Check conversion

• Are orders arriving?
• Is conversion materially below what the product normally achieves?
• Are relevant search terms converting?

Check search terms

• What searches generated clicks?
• What searches generated sales?
• What searches look irrelevant?
• Are new opportunities emerging?

Do not force a major optimization just because the seven-day review exists.

The purpose is to identify whether the campaign has enough evidence and whether something obviously needs attention.

When to Create a Second Campaign

Do not create another campaign simply because the first one feels messy.

Create a new campaign when you need a different control layer.

Examples:

Different bid strategy

A group of proven exact terms may need different bids from discovery traffic.

Different budget

A profitable product campaign may need its own budget.

Different objective

Brand defense and non-brand acquisition should not necessarily share the same management logic.

Different product

A materially different ASIN may need independent economics.

Different targeting

Keyword targeting and product targeting can have different performance patterns.

The question is:

What decision will this new campaign allow me to make that I cannot make cleanly today?

If there is no good answer, you may not need another campaign.

Your First Amazon PPC Campaign Is a Learning System

A new campaign should answer questions.

You want to learn:

• Which searches matter?
• Which targets convert?
• What CPC can the product support?
• Where is Amazon finding relevant traffic?
• Which traffic should be excluded?
• Which terms deserve more control?
• How much budget can the product actually use?
• Which parts of the campaign deserve to scale?

This is why campaign structure matters so much.

You are not just buying clicks.

You are creating a system that turns Amazon traffic into data you can use for the next decision.

A Simple Amazon PPC Launch Checklist

Before launch:

• Product is retail-ready
• Price is competitive
• Inventory is available
• Featured Offer eligibility checked
• Campaign objective defined
• Product selection confirmed
• Campaign naming convention applied
• Automatic or manual targeting selected
• Keywords or product targets reviewed
• Match types selected
• Initial bids set
• Bidding strategy selected
• Daily budget set
• Obvious negative targets added
• Start date checked
• End date checked
• Campaign structure documented
• First review date planned

After launch:

• Monitor spend
• Monitor impressions and clicks
• Review CPC
• Review CTR
• Review conversion
• Review search terms
• Identify wasted spend
• Identify winning targets
• Record changes
• Avoid premature bid changes

Final Takeaways

Setting up an Amazon PPC campaign from scratch is not difficult because the campaign builder is complicated.

It is difficult because the decisions you make during setup determine how easy the campaign will be to manage later.

Start with the product.

Define the objective.

Choose the right campaign type.

Group products only when their economics and objectives make sense together.

Use automatic targeting when discovery is useful.

Use manual targeting when you need more control.

Set bids based on economics and expected traffic.

Set budgets based on opportunity rather than arbitrary numbers.

Add obvious negatives.

Then let the campaign collect enough data to tell you what needs to change.

The best first campaign is not the one with the most keywords or the most complicated structure.

It is the one that gives you clean data, controllable spend, and a clear path to the next optimization decision.

For the next stage, see How to Track Amazon PPC Performance.

Turn Amazon PPC Data Into Action

Once you manage multiple campaigns, the difficult part is no longer creating them.

It is finding which campaigns, targets, and search terms actually need attention.

SellerRoot brings campaign, target, search-term, placement, and account performance into one view so you can move from reporting to action faster.

Use it to identify:

• High-spend targets that need investigation
• Search terms generating strong results
• Campaigns where ACoS or ROAS has changed materially
• Budget-constrained campaigns
• Waste that is easy to miss in manual reviews
• Product-level opportunities across the account

See what changed. Find the reason. Decide what to do next.

Explore SellerRoot

About the author

Yogendra Kashyap

Amazon Ads operators

Yogendra Kashyap is the Founder of SellerRoot and an Amazon Ads expert with 9+ years of experience helping brands grow through data-driven advertising. His expertise spans Amazon PPC, campaign optimization, search term analysis, and marketplace growth. Together with the SellerRoot team, he is building AI-powered tools for Amazon advertisers while sharing practical, experience-backed insights to help brands improve profitability and scale on Amazon.