Why Your PPC Tool Raised Your ACOS (and What to Check)

You signed up for a PPC automation tool, connected your Amazon account, and within two weeks your ACOS jumped from 25 percent to 40 percent. The tool's dashboard says it is "optimising," but your margin says otherwise. Before you cancel, there are several mechanical reasons this happens, and most of them are expected behaviour rather than a sign that the tool is broken.

This article walks through the honest mechanics of why ACOS rises after switching to automation, what numbers to check at each stage, and when a rising ACOS genuinely is a problem you should act on. Rufusly does not run ads, so there is nothing to sell you here. This is a reference piece.

The learning period is real, not a stalling tactic

Every PPC tool needs data before it can make informed bid decisions. When you hand over a campaign that you have been managing manually, the tool inherits your historical performance but not your intuition. It does not know which search terms you tried and abandoned, which placements you avoided on purpose, or which ASINs convert better on weekends.

To fill those gaps, the tool runs what amounts to a controlled experiment. It tests bids at different levels, enables match types you may have kept paused, and tries placements (top of search, rest of search, product pages) that you might not have explored. Each of these tests costs money and generates clicks that may not convert immediately. That spend shows up as a higher ACOS.

The learning period is not indefinite. Most tools converge within two to four weeks for products that sell at least a few units per day. Products with very low velocity take longer because the tool needs enough conversions to distinguish signal from noise. If your tool vendor cannot tell you roughly how long the learning phase lasts for your catalogue size, that is worth asking about before you start.

Bid exploration pushes spend into unproven territory

Manual PPC management tends to settle into a comfort zone. You find keywords that work, set bids that feel right, and leave them alone unless something breaks. The result is a stable ACOS, but also a ceiling: you are only capturing the demand you already know about.

An automation tool is designed to push past that ceiling. It raises bids on keywords where it suspects there is untapped volume, and it lowers bids on keywords where the current spend is not justified by conversions. The raises happen immediately (higher bids win more auctions, so you see the spend straight away). The payoff from those new impressions takes days or weeks to materialise as conversions, because Amazon's attribution window is 7 or 14 days depending on the ad type.

During this lag, your ACOS looks worse than it is. You are seeing today's spend against yesterday's conversions. If you check again after the attribution window closes, the numbers often look different.

Expanded match types bring in unfamiliar traffic

If you were running mostly exact-match keywords, a PPC tool may switch some to broad or phrase match to discover new search terms. Broad match casts a wide net: your ad appears for queries you never explicitly targeted. Some of those queries convert well and become new exact-match winners. Others waste spend and get negated.

The waste is front-loaded. You pay for the bad queries before the tool identifies and blocks them. This is the single biggest reason ACOS spikes in the first week after a tool takes over, and it is also the fastest to resolve. Check your search term reports daily during this phase and flag any obviously irrelevant terms to the tool (or add them as negatives yourself if the tool allows manual overrides).

Budget reallocation can look like overspend

Some tools redistribute your total budget across campaigns based on performance signals. A campaign that was spending five pounds a day might get bumped to fifteen if the tool calculates that its conversion rate justifies the increase. Meanwhile, another campaign gets scaled back.

The total spend may stay the same, but the campaign-level view in Seller Central looks alarming if you are used to fixed daily budgets. If your tool does budget reallocation, make sure you are tracking total account spend rather than individual campaign budgets. Set a hard ceiling on total daily spend in the tool's settings if one is available.

When a rising ACOS is actually a problem

Not every ACOS increase is a healthy learning curve. Here are the signs that something is genuinely wrong.

• The learning period has passed (the vendor's stated window, or at least four weeks) and ACOS has not started trending downward.
• Total ad spend has increased but total sales (organic plus advertising) have not. This means the tool is not generating incremental revenue, just shifting organic sales into paid.
• TACOS is rising, not just ACOS. A higher ACOS with flat or declining TACOS is defensible. A higher ACOS with a higher TACOS is a net loss.
• The tool cannot show you which specific changes it made and why. If the only explanation is "the algorithm is learning," with no data to back it up, you are flying blind.
• Your conversion rate on the listing itself has dropped. If clicks are up but conversions are down, the problem may be on the listing, not the campaign. Check your A+ Content and product images before blaming the tool.

A checklist before you blame the software

1. Confirm the learning window. Ask your tool vendor how long it takes and whether your catalogue size extends it.
2. Compare ACOS and TACOS together. If TACOS is flat or falling while ACOS rises, the tool may be doing its job.
3. Check the attribution lag. Look at ACOS for completed attribution windows (at least 14 days old), not real-time or same-day data.
4. Review search term reports. Are you paying for irrelevant queries? If so, add negatives immediately.
5. Check your listing conversion rate. More traffic to a weak listing just wastes ad spend. Run your listing through a listing quality scanner to rule out content issues.
6. Ask for a change log. A credible tool should show you exactly which bids it changed, which keywords it added or negated, and which budgets it moved.
7. Compare to the same period last year. Seasonal shifts in CPC and conversion rates can account for part of the change.

The TACOS perspective

ACOS in isolation is a misleading metric. It only measures the efficiency of your paid clicks, not the total return on your advertising investment. If a PPC tool raises your ACOS from 25 percent to 35 percent but your organic ranking improves and total sales grow by 20 percent, your TACOS may actually drop.

This happens because Amazon's ranking algorithm factors in sales velocity. More paid sales can push your product higher in organic results, generating free clicks that offset the higher ad cost. The effect is not guaranteed and not instant, but it is real enough that experienced sellers track TACOS as the primary metric and treat ACOS as a secondary indicator.

If your PPC tool does not report TACOS alongside ACOS, calculate it yourself: total ad spend divided by total revenue (from Business Reports, not from the ad console). That single number tells you more about your advertising health than any campaign-level ACOS figure.

What to do if the tool genuinely is not working

If you have given the tool a fair learning period, checked the attribution lag, reviewed search terms, and TACOS is still rising, it is reasonable to pause the tool and revert to manual management. Before you do, export everything: your campaign structure, keyword lists, negative keyword lists, bid history, and any audience segments the tool created. Some tools make this easy. Others do not, and that is worth knowing before you sign up. Check the tool comparison guide for what to look for in a vendor.

When you revert, do not delete the campaigns the tool created. Pause them and review the data. Some of those keywords and audience segments may be worth keeping even if the automated bidding did not work out.