Amazon PPC Launch Strategy: Step-by-Step Guide for Small Sellers

This launch strategy is written for small and mid-size sellers running on real-world budgets – think $500–$5,000 per month of controlled spend, not $30k–$50k. Your ads are funded by cash flow, not venture money. You do not have the luxury of “burning to learn.”

If you’re spending thousands per day, don’t DIY your campaigns. Hire someone who actually enjoys thinking about Amazon PPC at unhealthy levels. That’s my week, every week – I’m the guy who sets up 700 campaigns on a Wednesday just to test a theory.

Table of Contents

• Who This Launch Strategy Is For
• Phase 1: Exact Match – Teaching Amazon What Your Product Is
• Phase 2: Expansion – Auto, Broad, Phrase, ASIN, Category
• Bids vs Budgets – Gas Pedal and Handbrake
• Wrap-Up: The Simple Launch Sequence

Who This Launch Strategy Is For

This guide assumes:

• Your listing is already in good shape (title, bullets, images, reviews) – or at least not terrible.
• You understand the basics of PPC from the previous article: how Sponsored Products work, and what CTR, ROAS, and TaCoS mean.
• You have a limited budget and need your campaigns to be controlled and profitable fairly quickly.
• Your goal is profitable growth of the whole Amazon engine – PPC and organic together – not just flexing on ACoS or ROAS in isolation.

If that’s you, the launch sequence is simple:

1. Phase 1: Exact match only – clean, highly controlled launch.
2. Phase 2: Add Auto, Broad, Phrase, ASIN, and Category – after Phase 1 stabilizes.

Everything else – 16-campaign guru setups, “data gathering” spends you can’t afford – goes in the bin.

Phase 1: Exact Match – Teaching Amazon What Your Product Actually Is

Phase 1 is where small sellers win. You’re not trying to hit every search variation under the sun. You’re doing the opposite: sending Amazon a clean, unambiguous identity signal about your product.

Note: I know every guru on YouTube screams “start with Auto to learn.” You shouldn’t. Not on a small budget. Unless you’re ready to burn hundreds of dollars per day, Auto campaigns don’t teach you anything useful – the data is pure noise, as explained in a later section.

It’s like asking monkeys to write Shakespeare. Sure, if you throw enough monkeys at enough typewriters for long enough, one of them will eventually smash out “to be or not to be.”

Auto campaigns work the same way: with enough budget, eventually Amazon stumbles onto the right keywords. But you’ll burn a mountain of cash while it learns absolutely nothing useful from the noise.

You need to teach Amazon’s search engine (A9/Cosmo) what your product is first. That’s why Phase 1 is Exact-only. Clean, controlled signals. No chaos. No system guessing whether you’re selling a schoolbag, a leather purse, or a backpack.

Here’s the reality: Amazon doesn’t know what your product is.

You created a listing, added a title, bullets, description, and backend keywords – but from the algorithm’s point of view (A9/Cosmo), it’s still guesswork.

A new listing can easily look like several different product types at once. For example, Amazon may see your listing as:

• a schoolbag
• a women’s leather bag
• a genuine leather backpack

Why? Because your copy probably contains enough overlapping keywords that the system can map your product to multiple “concepts.” Amazon doesn’t assume – it tests.

That’s where Phase 1 comes in. Your first PPC campaign must send Amazon a sharp, controlled signal:

“THIS is the exact product I am selling.”

Exact-match keywords do that better than anything else. They strip away ambiguity and tell Amazon which searches your product should be relevant for – and which ones it shouldn’t even bother testing. That’s why Phase 1 is exact-only, low budget, and focused on clean data.

Step 1: Create Your First Exact-Match Campaign

In Seller Central:

1. Go to Advertising > Campaign Manager.
2. Click Create campaign → Sponsored Products.
3. On the first screen:
   • Ad group name: doesn’t matter too much – “exact” is fine.
   • Products: add your product – you can find it by title, sku or asin
   • Targeting: manual targeting
   • Manual Targeting: keyword targeting

Step 2: Set Your Bidding Strategy (Down Only – Before Adding Keywords)

This part is subtle but important.

1. Locate the Bidding strategy section.
2. Select Dynamic bids – down only.
3. Do this BEFORE you add keywords.

Why? Amazon recalculates and repositions suggested bids and your default bid level when you flip between “fixed”, “up and down”, and “down only.” If you add keywords first and then change the bidding strategy, Amazon can push your bids below where you think they are or generally mangle what you just set. Set “Down only” first, then add your keywords. And yes, I know amazon wants you to do up/down, but it gives you a lot less control. And yes — in general, ignore Amazon’s PPC suggestions. Their ‘recommended’ bids and strategies exist to make Amazon money, not you.

Step 3: Choose Targeting – Manual, Exact Only

1. Under Targeting, choose Manual targeting.
2. Under Manual Targeting, choose Keyword targeting.
3. Under Keyword Targeting: disable Broad and Phrase for this campaign – we only want Exact here.

Step 4: Add 5–6 Core Exact Keywords

Now add your keywords:

• Use your research from Helium 10 (Cerebro/Magnet) and pick the 5–6 most relevant, highest-intent, buyer-ready keywords for your product – the things a ready-to-buy customer would actually type.
• Paste them in and set match type to Exact for each.
• Initial bids: start around the suggested bid or slightly below (e.g. 80–100% of suggested).
• Remove: “Keywords related to your product category” (marked in red). Its not relevant for this campaign.

Do not dump 50+ keywords into this campaign. We’re not doing “spray and pray.” We’re doing controlled, high-signal testing. If these 5–6 are wrong, you’re teaching A9/Cosmo the wrong thing about your product.

Step 5: Settings

• Campaign name: Enter a sensible campaign name.
• Daily budget: Enter $20 (or equivalent in your currency)
• Dates: No need to add a start or end date – today is fine, if you’re ready to launch
• Portfolio: Is a way to categorize various campaigns – e.g. into products or product lines. Not relevant if this is your first campaign.

If your screen looks slightly different, don’t worry. Just find the same settings as described. From time to time amazon changes the layout of the backend, but the terms should remain the same regardless.

Bids and Budgets in Phase 1 – How to Think About Them

From the fundamentals article, two controls matter:

• Bids = gas pedal. How aggressively you’re willing to pay per click. High bids: more auctions, more traffic, more spend. Low bids: fewer auctions, less traffic.
• Budget = handbrake. How much you allow yourself to spend in a day, maximum, across this campaign.

Your starting point in Phase 1:

• Daily budget: ~$20 for this campaign.
• Keyword bids: near suggested, with “Dynamic bids – down only”.

Optional: Top of Search Boost (For Slightly Higher Budgets)
If you’re running a slightly higher budget – say $20–$30/day – and you’re willing to push a bit harder, you can open the Campaign settings (not the ad group), find Bid Adjustments, and increase your Top of Search placement by around 25%.

This does exactly what it says on the tin: Amazon increases your bid by 25% when competing for the very top placements of the search results page (i.e., bidding at 125% of your base bid). Expect a higher cost per click, but also expect a higher CTR – those placements get the most eyeballs.

This won’t fix bad keywords or a bad listing, but it can accelerate Phase 1 by giving Amazon more high-quality click data faster.

When and How to Adjust in Phase 1

Amazon’s learning cycle takes time. If you poke your campaigns every day, you’ll just confuse the system and your data becomes useless.

During the first 3 days:

Only touch bids if one of these is true:

• Problem 1: Budget dies in a few hours.
  Your bids are too high. Lower bids on the worst offenders by 10–20% at a time.
  Example: $1.50 → $1.20 (20% cut), not $1.50 → $1.47.
• Problem 2: Zero or almost no impressions.
  Your bids are too low. Increase by 10–20% per step and wait another 24 hours.

After the first week:

• Review performance once per week (twice at most).
• Keep the same 10–20% adjustment rule:
  • Good ROAS (3 or higher), low impressions → gently raise bids.
  • High spend, weak ROAS (less than 2) → gently lower bids.
  • Keywords with impressions but terrible CTR → usually not a PPC problem; check your main image, title, reviews, and price.

Do not add more keywords into this Phase 1 campaign. If you want to test a new keyword, it gets its own campaign later.

Run Phase 1 for approximately 7-10 days.

This window is enough to start teaching A9/Cosmo the most important thing during launch: “this is my product.”

Even better: think in clicks, not just days. Aim for each core keyword to get roughly 30–50 clicks. That’s enough signal for A9/Cosmo to decide whether your product actually converts on that query. In slow niches that might mean more than 14 days; that’s fine. Time doesn’t train the algorithm – data does.

You need enough impressions and clicks for Amazon to form a stable identity profile. Once Amazon has seen users search for a term, click your ad, and interact with your listing, the system effectively concludes:

“Got it! Your product is a genuine leather backpack.”

That’s when Phase 1 has done its job. Amazon now understands your product type and primary keyword alignment. Only at this point is it worth expanding into Auto, Broad, Phrase, ASIN, and Category campaigns.

Optional: Add a Small Brand-Defense Campaign

As you start getting traction, competitors will eventually bid on your brand. Small sellers can’t afford to lose their cheapest, highest-intent traffic.

Set up a tiny brand-defense campaign:

• Create a Sponsored Products campaign like SP-Exact-Brand-Defense.
• One ad group, your own product(s).
• Target your brand name and simple brand+product phrases as Exact match.
• Use low–moderate bids and a modest daily budget.

This keeps you in front of shoppers who already know your brand instead of donating that traffic to your competitors.

Phase 2: Expansion – Auto, Broad, Phrase, ASIN, Category

Phase 2 is where most people blow their budgets. The trick is to expand deliberately, not all at once.

We’ll add:

• A slow Auto campaign.
• Research-based Broad, Phrase, and extra Exact campaigns.
• ASIN targeting (competitors).
• Optional category targeting.

Still with one rule: one ad group per campaign. Multiple ad groups just split data and make optimization harder without giving you any meaningful control.

Important: Do not launch all Phase 2 campaigns at once.
Start with one or two (e.g. Auto + Broad), let them gather data for 3–4 days, then add the next campaign.
If you launch five campaigns on the same day, your budget will explode which is not the goal – we want to produce meaningful data which we can use to optimize the spend.

Step 1: Add a Slow Auto Campaign

This is not a discovery engine at low budget. It’s a background earner and a way to let Amazon “roam” a bit.

Many gurus tell you to launch Auto first and “harvest.” That only works if you can dump $300–$500 per day into an Auto campaign. That level of spend gives you statistically reliable patterns.

But at small budgets? You do not learn anything useful.

You will see garbage like:

• 1 impression → 1 click → maybe 1 sale

That is statistical noise. It’s a random passer-by clicking and buying on a whim, not a repeatable keyword you can build a PPC structure around.

So we run Auto slowly, safely, and for supplementary sales only.

1. Create a new Sponsored Products campaign.
2. Name it something like SP-Auto-ProductName.
3. Daily budget: $15–$20.
4. One ad group. Same product as Phase 1.
5. Targeting: Automatic targeting.
6. Bidding strategy: choose Dynamic bids – down only before you add or adjust bids just as before.
7. Leave the four auto targeting groups (Close match, Loose match, Substitutes, Complements) at low bids.

We’re not expecting magic keywords from this at $15-$20/day. We want a slow, steady stream of “easy” sales where Amazon finds obvious matches you might not have picked manually.

A more advanced strategy is splitting Auto into four separate campaigns (Loose, Close, Substitutes, Complements). It gives incredible control but adds complexity. We’ll cover that in another session.

How to adjust Auto

• Use the same 10–20% bid change rule as in Phase 1.
• If Close match or Substitutes are profitable, nudge those groups up slightly.
• If one group burns money with no sales, lower the bid or pause that group.
• Adjust once per week, not daily.
• Aim for a ROAS of 3x or higher.
• Let auto simmer. Auto campaigns takes a couple weeks to dial themselves in.

Only bother “harvesting” search terms into new campaigns if a specific term has at least 10–15 clicks or 3 or more sales and looks promising. Under that threshold, the data is just noise and not statistically meaningful, especially at small budgets.

Step 2: Build Broad / Phrase / Extra Exact Campaigns (Real Research)

Now we bring in your keyword research properly.

For each match type, create its own campaign:

• SP-Broad-ProductName-Set1
• SP-Phrase-ProductName-Set1
• SP-Exact-ProductName-Set2 (for secondary exact keywords)

Never mix Broad, Phrase, and Exact in the same campaign.
Amazon will prioritize the cheapest-to-serve match type and bury the rest,
so you’ll never get clean data and half your keywords will never show.

Inside each campaign:

• One ad group.
• 5–6 tightly related keywords per campaign – not 50.
• Targeting: keyword targeting.
• Match type: only Broad or only Phrase or only Exact (depending on the campaign).
• Daily budget: around $10–$15 per campaign to start.
• Bidding strategy: set to Dynamic bids – down only before adding keywords.

How to treat each match type

• Broad: Exploration but still controlled. Good for finding cheaper variants and long-tail phrasing.
• Phrase: Middle ground. Your phrase stays intact, but words can appear before/after. Good for intent-driven variations.
• Exact (Phase 2 set): Extra terms you didn’t include in Phase 1, like variants and long-tails.

Adjusting these campaigns

Same principles:

• Look at data weekly.
• Kill or reduce bids on keywords with:
  • High spend, no sales.
  • Terrible CTR (clearly not relevant).
• Raise bids slowly on:
  • Keywords with good ROAS but low impressions.
• Always in 10–20% increments, up or down.

Practical rules: If a keyword has spent more than the cost of your product? Disable it. There are exceptions to this if you have a product with a high LTV (for example subscriptions or strong repeat purchase), but this is a general, easy-to-remember rule of thumb for most single-purchase products.
Don’t wait for miracles. Amazon already told you it isn’t relevant enough to convert, or simply too expensive.
And yes – even if it’s your absolute main keyword.
Competition for that term may simply be too brutal, so you need to shift focus toward the “smaller” and less expensive keywords where you can actually win. Don’t chase. Go after what works.

Negative Keywords (for Broad, Phrase, and Auto)

Negative keywords prevent Amazon from wasting money on obviously irrelevant traffic.

Use Negative Phrase only – simple, safe, and very effective.

For example, if you’re selling a genuine leather backpack, add negatives like:

• pleather
• fake leather
• handbag
• purse

Negative Phrase means: if the phrase appears anywhere inside a shopper’s search term, your ad will NOT show.

This keeps your PPC focused on real shoppers and blocks people hunting for cheap alternatives. It also protects Broad, Phrase, and Auto from drifting into garbage traffic.

Negative Exact exists, but it’s a scalpel, not a hammer. If you add “leather handbag” as a negative exact term, it blocks only that exact query. Nothing more. So when shoppers search “leather handbag for women”, it slips straight through — because exact match means exact.

Step 3: Add ASIN and Category Targeting

Once Exact/Broad/Phrase and Auto are stable, you can add product and category targeting.

ASIN targeting (competitors)

1. Create a new Sponsored Products campaign: SP-ProductTargeting-ProductName.
2. One ad group.
3. Choose Product targeting.
4. Add:
   • Direct competitors’ ASINs (similar price/quality).
   • Weaker listings where your offer clearly looks better.
5. Set low–moderate bids and “Dynamic bids – down only”.

ASIN targeting can give you cheaper clicks and let you sit on your competitors’ pages. Great when your images/offer are stronger.

Tip: Avoid targeting huge brands like Nike, Herschel, or Coach.
Their listings attract a completely different audience and your CTR will crater.
Target competitors your own size first – similar price, similar quality, similar positioning.

Category targeting (use with care)

1. Create another Sponsored Products campaign if you want category-level coverage.
2. Target the main category or a subcategory that fits your product.
3. Filter by price, star rating, or brand if helpful.
4. Set low bids – category targeting can get wasteful fast if you’re not careful.

If Category targeting starts spending money without conversions, cut bids or pause it. It’s a nice-to-have, not a must-have for small budgets.

Bids vs Budgets – Gas Pedal and Handbrake (In Practice)

Quick recap of how to think about the controls across all campaigns:

• Bids = Gas Pedal
  • High bids → more impressions and clicks → more spend.
  • Low bids → fewer impressions and clicks → less spend.
  • You adjust bids per keyword or per target.
  • Always change in 10–20% steps, not pennies.
• Budget = Handbrake
  • Daily budget controls maximum damage.
  • If a campaign constantly hits its daily cap by noon, either:
    • Lower bids (preferred, performance-focused), or
    • Increase budget if ROAS is strong and you can afford more sales.
  • Don’t use budget changes as your main optimization tool. Performance lives at the bid level.

Cadence:

• First days of a brand-new campaign: adjust only if:
  • Budget is burning in a few hours (bids too high), or
  • No impressions at all (bids too low).
• Once stable: review and adjust weekly. Twice per week at most.
• Daily tinkering = noisy data and a confused algorithm.

ROAS expectations during launch:
ROAS 3–4 = Excellent for launch
ROAS 2–3 = Still good if organic ranking moves
ROAS < 2 = Something needs fixing (bids, keywords, listing, or images)

How to Adjust Your Bids and Diagnose Performance (Simplified)

Use this simple diagnostic table to understand what your PPC data is actually telling you, and which lever you should pull.

What “Good” Looks Like During Launch

The real goal of this launch is profit from the whole Amazon engine, not just winning at PPC screenshots.

Think of it like this:

• PPC brings in controlled, targeted traffic and feeds A9/Cosmo clean click + conversion data.
• As A9/Cosmo figures your product out, your organic rankings improve on the right searches.
• Over a few weeks, a bigger share of your orders come from organic, your TaCoS drops, and you keep the parts of PPC that are actually profitable.

If PPC ROAS looks a bit worse during launch but organic rank and total profit are improving, you’re winning.

How far back to look when adjusting: Always use a delayed window. Check performance from 2 to 16 days ago (roughly the last 2 weeks, skipping today and yesterday) and also from 2 to 30 days ago (the last 4 weeks, skipping today and yesterday). Never look at today or yesterday. Shoppers often add to cart or wishlist and buy days later — only about 80% of attribution shows up in the first 24 hours, and Amazon’s network is slow to fully sync. Most sellers adjust PPC like a caffeinated squirrel. Don’t. Amazon’s data needs time to settle before it tells you anything real. Fresh data lies; delayed data gets you closer to the truth.

Wrap-Up: The Simple Launch Sequence

If you’re a small or mid-size seller, you don’t need complexity. You need a launch strategy that respects your budget and still feeds Amazon the right signals.

The sequence:

1. Phase 1 – Exact only
   • One Sponsored Products campaign.
   • One ad group.
   • 5–6 exact-match, buyer-intent core keywords.
   • Dynamic bids – down only, set before adding keywords.
   • $20/day budget.
   • Run 10–14 days and/or until each core keyword has ~30–50 clicks.
   • Adjust bids only when really needed, in 10–20% steps.
   • Optional: a small brand-defense exact campaign for your own brand terms.
2. Phase 2 – Expansion
   • Add a slow Auto campaign.
   • Add separate Broad, Phrase, and secondary Exact campaigns (5–6 keywords each).
   • Add ASIN targeting on competitor products.
   • Optionally add Category targeting with low bids.
   • Always one ad group per campaign.
   • Adjust bids 10–20% at a time, weekly.
   • Only “harvest” terms that have at least 10–15 clicks and consistent performance.

That’s it. No 20-campaign monstrosities, no random keyword floods, no “let’s just see what happens” burn phases. Just a clean, controlled launch made for sellers who actually care if the money coming out is larger than the money going in – from PPC and from organic.

Always keep in mind: the goal here is profit, not vanity ACoS/ROAS numbers. PPC is a lever you use to buy data, accelerate ranking, and control your sales velocity on Amazon – and yes, in the beginning it will feel like an expense – but the goal is to turn that into a profitable engine. Know your numbers end-to-end (landed cost, FBA fees, prep, shipping, VAT, everything) so you can see when the whole system (PPC + organic) is actually making you money.