Amazon PPC for Beginners: Your First 90 Days, Without the Expensive Mistakes

I learned Amazon PPC the expensive way: on my own FBA product, a lighting brand I launched in 2023, with my own money disappearing every time I got it wrong. In the first month, I burned hundreds in auto campaign spend on search terms I never harvested — money that fed Amazon's engine and nothing else. I checked the account every morning and changed bids based on three-click sample sizes. I didn't know my break-even ACoS until week three, which meant I had no idea whether a 65% ACoS was a disaster or just normal launch noise. Three years and 500+ campaigns across six brands later, here's the plan I wish someone had handed me on day one.

Before your first campaign: know two numbers

Your break-even ACoS. (Price − COGS − Amazon fees) ÷ price. This is the ACoS above which ads lose money. If you don't know it, you can't tell a good campaign from a bad one — calculate it here.

Your realistic budget. Work backwards from a sales goal with the PPC budget calculator. If the implied ACoS comes out above your break-even, fix your listing or your costs before funding ads.

Nothing I say after this matters if you skip those two steps. Every panic bid-change and every wrong conclusion I drew in month one came from not having a baseline to compare against.

Six terms every beginner must understand

You'll see these in every report, every tutorial, and every Seller Central screen. Learn them now so you stop guessing at what the numbers mean.

• ACoS (Advertising Cost of Sale): Ad spend divided by ad revenue, as a percentage. The core efficiency metric for PPC campaigns.
• ROAS (Return on Ad Spend): Revenue divided by spend. The inverse of ACoS. Some sellers think in ROAS, others in ACoS — pick one and stay consistent.
• CPC (Cost Per Click): What you pay each time a shopper clicks your ad. Your bid sets the ceiling; the auction sets the actual cost.
• CTR (Click-Through Rate): Clicks divided by impressions. Low CTR usually means the main image or title isn't compelling enough at a glance.
• CVR (Conversion Rate): Orders divided by clicks. If CVR is low, people are clicking and leaving — that's a listing problem, not a bid problem.
• TACoS (Total ACoS): Total ad spend divided by total revenue (organic plus paid). The metric that tells you whether PPC is building the business or just subsidising it.

The starting structure: two campaigns, two jobs

Don't launch ten campaigns. Launch two:

1. Auto campaign — Amazon matches your listing to search terms automatically. This is your paid research engine. Modest budget, dynamic bids down only.
2. Manual campaign — the 5–15 obvious keywords you already know describe your product, in broad or phrase match.

That's genuinely it for month one. Complexity comes later, and it comes from data, not from guessing. (The full system is in Amazon PPC campaign structure.)

Day 1 setup checklist

When you open Seller Central to build these two campaigns, these are the settings that matter:

• Bidding strategy: Dynamic bids, down only. Don't use "up and down" in the early weeks — you'll overspend before you have conversion data to justify it.
• Daily budget: Set a floor of $20–50 per campaign. Low budgets throttle impressions mid-morning and you never accumulate enough data to make real decisions.
• Auto campaign targeting groups: Enable all four groups (close match, loose match, substitutes, complements) at launch. You want broad discovery. Narrow it later based on performance.
• Manual campaign match type: Start with phrase match, not exact. Phrase gives range without the scatter of broad.
• Bid starting point: Use Amazon's suggested bid as a reference, then set yours 10–15% below it. You can raise bids on proven performers — overpaying on day one just burns your data budget.
• Negative keywords: Add obvious irrelevant terms before you even go live — unrelated use cases, generic category terms, or competitor brand names. Start clean.

Weeks 2–6: let data, not anxiety, make the calls

The beginner instinct is to check hourly and react to everything. Resist it. The rules that matter:

• Judge nothing under 10 clicks. One sale on two clicks isn't a winner; five clicks with no sale isn't a loser.
• Weekly, negative the proven losers: search terms with 10+ clicks and zero orders get a negative exact (how negatives work).
• Expect launch ACoS to look scary. Early ACoS runs high while data accumulates and your listing has no review base. What matters is trend, not day three.

The weekly rhythm is the discipline. Not daily. Not hourly. Set a recurring time each week to review search terms and adjust bids — then close the tab and leave it alone.

Weeks 6–12: the harvest

By now your auto campaign has done its job — its search term report shows which real customer searches convert. Now:

1. Take terms with orders (and 10+ clicks of history) into a new exact match campaign with precise bids.
2. Negative those terms in the source campaign so they have one home.
3. Shift budget toward exact — it's where your profit lives.

This harvest loop — discover in auto, prove in broad, profit in exact — is the entire engine. Everything advanced (bid strategy, placements, scaling) is refinement on top of it.

What 90 days actually looks like

Most beginners expect a smooth ramp. The reality is messier. That's normal.

Weeks 1–2: ACoS will likely be above your break-even. You have no review base, no purchase history for the algorithm to learn from, and you're paying full price for discovery. Don't touch bids. Let data accumulate.

Weeks 3–6: ACoS starts trending down as you negative junk terms and the algorithm learns what converts. You might see your first profitable days. Resist the urge to scale budget here — wait for consistency, not a single good week.

Weeks 7–12: If the product and listing are solid, ACoS converges toward a sustainable range. Organic rank starts lifting from the purchase velocity your PPC generates. TACoS becomes the number to watch — if your total ACoS is falling even as ad ACoS holds steady, that's the sign the flywheel is spinning.

After 90 days: You have real data. You know which keywords drive orders, what your conversion rate is, and whether your unit economics work at scale. That's the moment to push budget — not before.

The three mistakes that burn beginner budgets

1. No negatives. The account bleeds through hundreds of junk search terms while the seller stares at bids. My auto campaign in the first month was serving on searches for unrelated product categories — ceiling fixtures, outdoor lighting, things I wasn't even selling. Every click was wasted spend. Build a negative list from day one and add to it weekly.
2. Bidding by panic. Slashing everything after a bad week, doubling everything after a good one. Weekly cadence, 10-click minimum, always. I once cut bids on a keyword after three bad days in a row — it turned out to be a top performer over the prior month. Took two weeks to recover the impression share.
3. Blaming ads for a listing problem. If people click and don't buy, the traffic isn't the issue — price, images, or reviews are. PPC amplifies your listing; it can't fix it.

When to get help

If you're doing real volume and PPC has grown past what you can manage in a spreadsheet and an hour a week — that's the point where structure pays for itself. The free audit is a good test: I'll look at what you've built, find the wasted spend, and send you the three highest-leverage fixes in a short video. If you're early and small, the fixes above are honestly most of what I'd tell you anyway.