If your new Amazon PPC agency changed bids on day one, it destroyed the only baseline it could be judged against. A good first month is mostly reading, writing targets down and cutting waste in a set order, and this is the week-by-week version, with the questions to ask if it is not happening.
Why the first month decides whether you can judge the agency at all
Most owners judge a new agency on the first monthly report. By then the most important decisions have already been made: whether anyone recorded where the account started, whether targets were agreed before changes began, and whether the changes were made in an order that lets you tell which one worked.
An agency that skips those steps can still produce a report that looks fine. What it cannot produce is proof. So the useful question for month one is not whether ACOS went down. It is whether you will be able to tell, three months from now, what the agency actually did.
Days 1 to 7: read the account before touching it
The first week should be diagnosis, not changes. Three exports cover almost everything: the Sponsored Products Search Term report for at least 60 days, the Business report by ASIN, and a bulk operations file with the full campaign structure. What reports an Amazon PPC audit needs explains what each one shows.
From those, a good agency writes down a baseline: spend, sales, ACOS and TACOS by campaign type, branded and non-branded separately, and the conversion rate of each main ASIN. If the listing is the real bottleneck, this is the week you should hear it, before any more money goes into ads pointed at it.
The question to ask at the end of week one: can I see the baseline you recorded? If the answer is a dashboard screenshot from today, nothing was recorded.
Days 7 to 14: agree the targets in writing
Before restructuring, the agency should work out your break-even ACOS per product, which is your margin after every Amazon fee and the cost of goods. A copied industry target means nothing on your numbers. How to calculate break-even ACOS shows the arithmetic.
Targets should then be set by job, not as one number for the whole account. Campaigns buying rank on a launch, campaigns defending your brand and campaigns acquiring new customers are allowed different ACOS levels, and the agency should say which is which, and for how long a higher ACOS is accepted where it is.
Days 14 to 21: restructure and take the first waste out
This is when changes start, in money order, biggest first. The fastest measurable win is usually negating search terms that spend without selling, because it is pure subtraction. Amazon's own guidance is to judge a term after it has received at least 20 clicks, to use phrase or exact match for negatives, and to let negatives run for two weeks or more before changing strategy again.
Alongside that: separating discovery, exact and defensive campaigns so each has a readable signal, removing the same keyword running in several campaigns at different bids, and moving search terms that have proven themselves in automatic campaigns into manual ones, which is how Amazon describes using the search term report for automatic campaigns.
Expect this to show up as lower wasted spend within two to three weeks. It is a real saving, but it is stopping losses, not growth, and a good agency will say so.
Days 21 to 30: the weekly routine and the first report
By the end of the month the routine you are paying for should be visible: search term review, harvesting, negation, bid moves toward target, and checks on budgets and placements. Amazon's own best-practice guide suggests monitoring metrics at least twice a week in the first week of a new campaign, which is a fair minimum for anything newly restructured.
The first report should read as a list of decisions, not a data dump: what changed, why, and what it did, compared against the baseline from week one. If a change has not had time to show a result yet, the report should say that rather than claiming one.
What you should not expect in month one
Rank and organic growth. Gains in organic rank usually take 60 to 90 days, because Amazon's ranking responds to sustained sales on a keyword rather than any single change. How long Amazon PPC takes to work explains the two clocks.
A fair read of the last few days either. Amazon credits a sale to an ad for days after the click, so the most recent week always looks worse than it will once the sales catch up. An agency reacting to yesterday's ACOS is reacting to incomplete data.
Red flags in the first 30 days
Bulk bid changes in the first week, before any baseline exists. A request for your Seller Central password: Amazon Ads lets you invite users with standard roles (admin, editor, viewer) or custom permissions, so nobody needs your login. Reports that show only ACOS, with no TACOS and no branded split, which lets defending your own brand flatter the numbers. And any promise of a guaranteed rank or sales figure, because nobody controls Amazon's algorithm or a competitor's budget.
If you are still choosing who to hire, how to choose an Amazon PPC agency lists the questions to ask before you sign. This first month is exactly how our own Amazon PPC management starts, from the free audit onwards. If you would rather keep the work in-house with a senior plan to follow, that is what Amazon consulting is for.
Sources
- Amazon Ads — A guide to targeting with Sponsored Products (source for evaluating a term after at least 20 clicks, using phrase or exact match for negative keywords, letting negatives run for two weeks or more, and adding search terms from automatic campaigns to manual campaigns)
- Amazon Ads — Best practices for your Sponsored Products ads (source for monitoring metrics at least twice a week in the first week, and creating manual campaigns to supplement automatic ones)
- Amazon Ads — A guide to the Amazon Ads advertiser account (source for inviting users with standardized roles: admin, editor, viewer, or custom permissions)