The first three months of an Amazon PPC account decide most of what comes after. Accounts that start without a plan tend to end up with dozens of overlapping campaigns, no negative keywords and no idea which targets actually sell. Fixing that later takes longer than building it right the first time.
This roadmap splits the first 90 days into three phases. Each phase has a short list of tasks and a test for when you are ready to move on.
Before day one: check the listing
Ads send shoppers to your listing. If the listing does not convert, the ads will not either. Before launching, check that the main image is clear, the title says what the product is, the price is competitive for the category and the product has at least a few reviews if possible. Listing optimization before ads covers the checklist.
Also work out your break-even ACoS: roughly your margin before ad costs. It is the line every decision in the next 90 days is measured against.
Days 1 to 30: build and collect data
The goal of the first month is data, not profit. You want to learn which search terms and products convert for your listing.
Launch an auto campaign. Let Amazon find search terms based on the listing. Keep the bid moderate and the budget large enough to run through the day.
Launch a manual broad or phrase campaign. Use 15 to 30 keywords from your research. This tests the terms you believe in.
Launch a manual exact campaign on your most obvious terms. A handful of keywords you are confident about, with a slightly higher bid.
Name everything consistently. You will thank yourself in month three. Campaign naming has a simple convention.
Check daily for problems, not performance. Is each campaign serving? Is any running out of budget before evening? Are any search terms clearly irrelevant? Add those as negatives immediately. Do not cut keywords for low sales yet; most have too few clicks to judge.
Ready to move on when: each campaign has enough clicks per target that a few targets clearly stand out, good or bad. For many accounts that takes three to four weeks.
Days 31 to 60: cut and promote
The second month turns data into structure.
Harvest search terms. Pull the search term report from the auto and broad campaigns. Terms that produced orders at an acceptable ACoS move into the exact campaign as keywords. Add them as negative exact in the source campaign so the two do not compete.
Cut waste. Search terms with many clicks and no orders become negatives. Use a click threshold that makes sense for your conversion rate rather than a fixed number. Negative keywords covers how to choose it.
Adjust bids. Lower bids on targets well above break-even ACoS. Raise bids modestly on targets well below it that are not getting many impressions.
Check placements. The placement report shows whether top of search, rest of search or product pages perform best. Add a placement adjustment only where the data clearly supports it.
Ready to move on when: the exact campaign holds most of your proven terms, the auto and broad campaigns are mainly discovering, and weekly changes are getting smaller.
Days 61 to 90: tighten and automate
The third month is about making the account run without you watching it every day.
Look at time patterns. With two months of data, hourly and day-of-week patterns start to show. If some hours consistently convert far below average, scheduling becomes worth testing. Reading hourly performance data explains how to find them.
Set budgets from data. Campaigns that convert below break-even and run out of budget deserve more. Campaigns that never spend their budget do not need it.
Add guardrails. Set alerts for spend spikes and campaigns that stop serving, so a problem does not run for days before you notice.
Decide what to automate. List the tasks you repeated every week in month two. Budget changes, schedules and simple pause rules are good first candidates. What to automate first covers the order.
Common mistakes in the first 90 days
Changing too much too often. Bids changed every day never collect enough data to judge. Make a change, then wait a week.
Judging on the last few days. Sales take time to attribute. The last two or three days in any report are incomplete.
Too many campaigns. More campaigns means less data per campaign. Start with three and add structure only when the data asks for it.
Ignoring organic sales. Ads can lift organic rank. Watch total sales and TACoS, not only ACoS.
What day 90 should look like
A clean structure of discovery and performance campaigns. A list of proven keywords in exact match. A growing negative list. Budgets sized to what each campaign actually needs. A weekly routine you can run in an hour or two. And a clear view of whether the product can be advertised profitably, which is the most useful thing the first 90 days can tell you.
Frequently asked questions
How long before Amazon PPC becomes profitable?
It varies by product, price and competition. Many accounts spend the first month or two buying data at a higher ACoS than they want long term, then improve as waste is cut and winners are promoted. Set a break-even target and track the trend rather than expecting a profitable first week.
How much should I spend in my first 90 days of Amazon PPC?
Enough that each core campaign gets a meaningful number of clicks every week. A budget so small that campaigns stop by mid-morning produces too little data to make decisions. Start with a modest daily budget per campaign, watch whether they run out early, and raise budgets on campaigns that convert.
When should I start automating Amazon PPC?
Once the structure is stable and you know which tasks you repeat every week. Budget and schedule rules can go in early. Bid automation works better after you have a few weeks of data on each target, usually in the second or third month.
Off Hours takes over the month-three work, running dayparting, budget and performance rules plus free spend alerts on a 15-minute cadence. Start a free 14-day trial.