Managing Amazon PPC: The Loop, the Numbers, and What to Do When a Fix Doesn't Work

What 'managing' Amazon PPC actually means

Managing Amazon PPC is a loop, not a setup task. You pull the search term and campaign reports, find where money is going to clicks that don't turn into orders, change bids, budgets or negatives, then leave the change alone long enough to see if it worked. Sellers who say they "set it up once" usually mean they ran the loop for a week in month one and never again — that's a launch, not management.

The levers are the same for every account, small or large:

• Bids — what you pay per click, set at the keyword, product target, or placement level.
• Budgets — the daily ceiling that decides whether you keep showing up after 2pm or run out and go dark.
• Negatives — search terms you exclude because they burn spend without converting.
• Match type and harvesting — moving winning terms from broad or auto discovery into exact match campaigns you can bid precisely.
• Placement and dayparting — top-of-search vs. product page vs. rest-of-search, adjusted only where the data supports it.

Everything else — a dashboard, a rules engine, an agency's weekly call — is just a way of running that loop faster or with less of your own time. The loop itself doesn't change.

A worked example: cutting ACoS from 40% to a target of 25%

Say a Sponsored Products campaign on a $30 item spent $2,400 last month and produced $6,000 in sales — a 40% ACoS against a 25% target. That's roughly a $600 gap between what the campaign should have cost at this sales volume and what it actually cost. The fix isn't "cut every bid 10%" — that punishes keywords already hitting target along with the ones that aren't.

The right sequence:

• Pull the search term report and sort by spend, not clicks.
• Isolate terms with spend above roughly $20 and zero orders — in this example that's about $540 of the $2,400, nearly a quarter of the overspend on its own.
• Negative those terms out at the exact level so they stop pulling in through broad or auto campaigns.
• On keywords that did convert but sit above target ACoS, cut bids in the 10-15% range — enough to move down the auction, not enough to fall out of it entirely.
• Leave it. A full click-to-conversion cycle on most products runs 10-14 days; checking daily just means reacting to noise.

A properly run change carries three things before it goes live: the evidence behind it, a measurement plan, and a rollback trigger. In this example, the evidence is the spend and order split above, the measurement plan is the two-week hold, and the rollback trigger is simple — if ACoS is still above 32% at day fourteen, put the bids back and take a second pass at negatives before touching bids again. Most sellers skip that last step and just keep cutting, which is how a campaign ends up bid-starved and invisible three months later.

The four ways to manage it

None of these change the loop described above — they change who runs it and how often it actually gets run.

The mistakes that actually cost money

Some mistakes show up in nearly every account audit:

• Reacting to daily ACoS swings. A single day's ACoS on a low-traffic keyword can swing 20 points either way on three orders. React weekly at minimum, and only once there are enough clicks for the number to mean something.
• Harvesting search terms and never negating them. Pulling the report isn't the fix. The negative has to actually go in.
• Chasing ACoS instead of profit. A campaign at 15% ACoS on a low-margin SKU can lose money faster than one at 35% ACoS on a high-margin one. Set the target from the product's margin, not from a benchmark.
• Cutting new launches too early. This is one we've made ourselves — killing a keyword on day four of a launch because it shows a rough ACoS, before it has enough clicks to be anything but noise. A new ASIN's first two weeks of data are mostly statistical noise, and cutting too early means you never find out what day fourteen would have looked like.
• Treating branded and defensive spend like prospecting spend. A keyword with your own brand name in it isn't there to find new customers. Judging it against the same target ACoS as a broad discovery term misreads what it's for.

What to do when the fix doesn't work

If ACoS is still high after bid cuts and negatives, the problem often isn't in PPC at all. Check the listing's conversion rate first — traffic can be fine while the page isn't converting, and no bid change fixes a listing problem. Check whether CPCs rose category-wide because a competitor raised bids, which looks like "your account got worse" but isn't something a bid cut alone solves. Check whether the budget is capped and Amazon is quietly rationing impressions to part of the day, which shows up as a flat conversion rate but falling sales. And check whether a campaign-level bid strategy override — "dynamic bids, down only," for example — is fighting the keyword-level change you just made.

If the number itself looks wrong — a sudden ACoS spike, or a keyword showing spend with no clicks — check the reporting lag before touching a bid. Amazon's attribution window means a same-day pull is incomplete and often corrects itself within 24-48 hours. Don't make a bid decision off a report that hasn't finished settling.

Where Dr. PPC fits in this

Dr. PPC is the autonomous version of the loop above: it reads the whole ad account, writes a strategy per product against that brand's real economics, and proposes changes with the evidence, measurement plan, and rollback trigger built in — the same three things any well-run account should require, generated and checked automatically instead of by hand. The client sets the autonomy level: every change waiting on a human click, routine changes automatic with the bigger ones queued for approval, or fully autonomous inside agreed guardrails. Inventory risk, pricing, new launches, and creative always go to a human, regardless of setting. It's $300 a month plus 3% of ad spend, capped, month-to-month, with the first 30 days free, and Orbit — the analytics, profitability reporting, and trackers most competitors sell as a separate subscription — is included at no extra cost. It's operated by Full Circle, which has managed more than $500M in revenue across 100+ brands.