Amazon Marketing Services is one of those phrases that refuses to leave agency conversations.
Technically, AMS is old language. Amazon unified Amazon Marketing Services, Amazon Media Group and Amazon Advertising Platform under Amazon Advertising in 2018, and the current brand is Amazon Ads. Sponsored Products, Sponsored Brands, Sponsored Display, Amazon DSP, Amazon Marketing Stream and Amazon Marketing Cloud now sit in a much broader retail media system than the original AMS console ever did.
Commercially, though, agencies still hear “AMS” every week. It appears in old retainers, client briefs, vendor-side contracts, job descriptions, dashboards, training videos and folder names. A client asks for “AMS management” when they actually mean Amazon Ads strategy. A vendor-side stakeholder says “AMS spend” when the operator is looking at Sponsored Brands Video, DSP audiences and hourly budget pacing. A new account manager joins the agency and inherits three years of decisions labelled with a name Amazon no longer uses.
The named mistake I see is treating AMS as a campaign console instead of a change-control problem. The agency optimizes bids, adds keywords, pauses targets and reports ACOS, but nobody can quickly answer who approved the spend change, which margin file was current, whether stock cover allowed the move, or whether the client had agreed that this SKU was still in growth mode.
My stance: marketplace agencies with five or more people need an Amazon Marketing Services change ledger. Not because AMS is the modern product name. Because the old phrase often hides old operating habits: revenue-led reporting, platform-only KPIs, unclear approval rights and campaign changes that are logical inside Amazon Ads but weak inside the client’s profit model.
This guide is for marketplace agencies in Germany, the United States and cross-border teams managing Amazon work for multiple clients. The goal is simple: turn “AMS management” from a vague service line into a controlled operating layer where every meaningful PPC change has evidence, margin context, stock permission and client traceability.
What the public AMS content explains well
The existing content on Amazon Marketing Services is useful, especially for teams that need the vocabulary cleaned up. Productsup explains AMS as the former name for Amazon’s advertising tools and walks through visibility, targeting, budget control, analytics and ad formats. Its older rebrand article also captures the 2018 shift from separate names like AMS, AMG and AAP into one Amazon Advertising brand.
Adference is especially clear for German teams: AMS is historical language, Amazon Ads is the current term, and the old phrase still appears in older articles, tool names and contracts. That matters when an agency works with DE vendor accounts where stakeholders still use the legacy wording.
Amazowl’s guide is older but useful for context. It explains how early AMS was associated with vendors, Sponsored Products, headline ads, product display ads, second-price auctions, minimum CPCs and the way paid placements consume Amazon search real estate. That history helps explain why senior clients may still associate AMS with “serious brand advertising” rather than the broader Amazon Ads stack.
Pacvue, Channable and other retail media software players focus more on the current operating layer: Amazon Marketing Stream, DSP, campaign automation, bulk control, multi-marketplace advertising and cross-retailer reporting. Their angle is scale and activation, which agencies absolutely need.
What most of this content does not solve is the agency governance layer. It explains what AMS was, what Amazon Ads is, which ad formats exist and why automation helps. It rarely answers the question an agency director actually faces on Monday morning: “Which of last week’s 148 campaign changes should the client trust, which ones require approval, and which ones moved spend without enough profit evidence?”
The agency problem is not terminology. It is decision memory.
AMS terminology creates confusion, but confusion is not the expensive part. The expensive part is lost decision memory.
An Amazon specialist raises bids because conversion rate improved. A catalog manager later discovers the promoted variation was merged under a weaker parent. A finance analyst updates landed cost from €8.40 to €9.70, but the ad rule still uses the old break-even ACOS. The client approves “more aggressive AMS” in a meeting, but nobody records whether that means Sponsored Products only, Sponsored Brands too, or DSP retargeting with a separate budget. Three weeks later the QBR becomes a debate about interpretation instead of a discussion about profitable growth.
That is why agencies need a change ledger. The ledger does not replace Amazon Ads. It sits above the platform and records the commercial permission behind the change.
For every meaningful Amazon Ads action, the ledger should answer seven questions:
- What changed? Campaign, ad group, keyword, product target, budget, bid, placement multiplier, rule or negative.
- Why now? Search term evidence, stock arrival, margin update, retail readiness fix, competitor move, seasonal plan or client request.
- Which SKU economics applied? Contribution margin, fees, fulfilment cost, expected return rate and break-even ACOS at the time of the change.
- What was the stock permission? Days of cover, inbound timing, FBA/WFS/3PL split and whether the SKU could safely absorb more demand.
- Who approved it? Operator, account lead, client, finance, catalog owner or pre-approved automation rule.
- What is the rollback trigger? Spend ceiling, ACOS ceiling, TACOS movement, conversion drop, stock floor or Buy Box issue.
- When does the decision expire? Date, spend threshold or evidence threshold after which the change must be reviewed again.
This sounds administrative until you manage ten clients, 2,400 advertised SKUs and three specialists making daily changes. Then it becomes the difference between a scalable agency product and a collection of talented operators who keep context in their heads.
Scenario 1: Keller Home Goods and the €6,800 “good ACOS” mistake
Imagine Keller Home Goods, a fictional German kitchenware brand spending €18,000 per month on Amazon Ads across DE, FR and IT. The agency sees a ceramic pan set improving from 31% ACOS to 22% ACOS over ten days. The specialist increases the Sponsored Products exact bid from €0.82 to €1.05 and raises the daily budget from €120 to €190.
Inside Amazon Ads, the move looks sensible. Sales are rising, ACOS is below the client’s 28% target, and the search term report shows repeat conversions on “keramik pfanne set”.
Inside the business, the move is weak. The landed cost file was updated three days earlier after a freight surcharge. Contribution margin per unit dropped from €11.40 to €8.10. The true break-even ACOS moved from 32% to 23%. The SKU also had only 16 days of FBA cover, with the next inbound shipment not due for 24 days.
Over the next two weeks, the campaign spends an extra €6,800. Reported ACOS stays near 24%, so the dashboard does not scream. But the SKU sells through too quickly, loses stock for six days, and the agency has to explain why a “profitable” AMS decision created a stockout and thinner profit.
A change ledger would have blocked or capped the move. The bid increase would require current margin confirmation and a stock-permission flag. The rule might read: if contribution margin is below €9 or stock cover is below 21 days, bid increases above 15% need account-lead approval and a spend cap. The specialist could still act, but not in a way that silently spends beyond the SKU’s permission.
Scenario 2: Northstar Pet and the $11,000 Sponsored Brands ambiguity
Now take Northstar Pet, a fictional US pet accessories brand with $42,000 monthly Amazon Ads spend. In a planning call, the client says, “Let’s push AMS harder for the new travel carrier range.” The agency hears growth permission. The specialist launches Sponsored Brands Video, Sponsored Products manual campaigns and a small Sponsored Display retargeting test.
After 30 days, the range has $63,000 in attributed ad sales from $15,400 spend. The Sponsored Products campaigns are acceptable at 24% ACOS. Sponsored Brands Video spends $11,000 at 41% ACOS. The client is frustrated because they expected a conservative launch, not full-funnel expansion.
The problem is not that Sponsored Brands Video is bad. It may be exactly the right format for a new carrier range that needs education. The problem is that “push AMS harder” was never translated into decision rights.
A proper change ledger would split the approval into lanes:
- Sponsored Products launch budget: pre-approved up to $4,000 spend or 32% ACOS.
- Sponsored Brands Video: test-only, capped at $2,500 until review-rate and conversion benchmarks are met.
- Sponsored Display: manual approval required because attribution and incrementality are less direct.
- Portfolio-level guardrail: total launch spend cannot exceed $9,000 before the next client checkpoint.
The same Amazon Ads work can still happen. The difference is that the agency can prove the client authorized the lane, the spend ceiling and the learning objective. That protects trust when the early numbers are messy.
Scenario 3: Alpine Tools and the hourly data trap
Amazon Marketing Stream makes agency work faster. Hourly campaign metrics, budget consumption and near-real-time signals can help teams react before yesterday’s report arrives. For agencies, that is powerful. It is also dangerous when speed outruns permission.
Alpine Tools, a fictional cross-border hardware brand, spends €27,000 per month on Amazon Ads. The agency uses hourly data and sees that conversion between 18:00 and 22:00 is 36% stronger than the daily average. The team raises evening bids by 20% across five campaigns. Revenue improves for four days. Then the warehouse flags that the promoted bundle is using a component shared with a higher-margin B2B kit, and the evening push is stealing stock from a better channel.
The hourly signal was real. The permission was incomplete.
A change ledger keeps fast data useful by assigning speed limits. Hourly signals can trigger a temporary bid window, but only inside SKU-level stock and margin boundaries. For example: evening bid uplifts can run for 72 hours, capped at €1,500 extra spend, only when stock cover is above 30 days and contribution margin is above 18%. If any condition fails, the action becomes a recommendation instead of an automated change.
What belongs in an AMS change ledger
The ledger should be simple enough for operators to use daily and structured enough for directors to audit. I would include six layers.
1. Campaign identity and role
Every campaign needs a role: launch, harvest, defend, scale, clear stock, protect brand, test creative, conquest competitor, or maintain presence. Without a role, every KPI debate becomes fuzzy. A 40% ACOS might be unacceptable for a mature exact campaign and perfectly acceptable for a capped launch-learning campaign.
2. SKU margin and break-even ACOS
Amazon Ads cannot decide profitability from ad-attributed revenue alone. The ledger should hold the current contribution margin, marketplace fees, fulfilment cost, return assumption and break-even ACOS per SKU or product group. FiveX helps here by connecting advertising performance to SKU profitability, so bid and budget decisions can be checked against real margin rather than a static target.
3. Stock and retail readiness permission
Spend should not scale into weak availability. The ledger needs days of cover, inbound dates, Buy Box or offer status, review rating, content readiness and suppressed-listing warnings. FiveX inventory insights and marketplace monitoring can turn those signals into a simple permission state: scale, cap, wait or review.
4. Approval rights
Not every action needs a meeting. Agencies should define thresholds. A junior specialist may lower a bid by 10% inside an agreed rule. A 35% budget increase on a hero SKU may need the account lead. A DSP or Sponsored Brands Video test may need client approval. The point is not bureaucracy. The point is to make approval predictable before the account is under pressure.
5. Automation rule history
Rules need memory too. If automation pauses targets above 45% ACOS, increases bids when share of voice drops, or caps campaigns when stock falls, the agency must be able to show which rule acted and why. FiveX ad automation logs and AI recommendations are useful because the action is not just “the tool changed it”; it becomes a reviewed decision with traceable conditions.
6. Expiry and rollback
Every temporary decision should expire. Launch budgets, seasonal bid uplifts, Prime Day defence, clearance pushes and competitor conquest tests all need review dates. Otherwise yesterday’s exception becomes tomorrow’s baseline.
The operating cadence I would use
For a five-person marketplace agency, I would keep the cadence practical.
Daily: review exceptions only. Stock below threshold, spend pacing above permission, margin file changes, campaigns out of budget, suppressed listings, sudden conversion drops and automation actions above a defined value.
Weekly: run the change review. Pick the top 20 changes by spend impact, not by how dramatic the percentage looks. Ask: did the evidence match the action, did the action stay inside permission, and does it need to continue?
Monthly: reconcile the client story. Show which AMS/Amazon Ads decisions created profitable growth, which were tests, which were rolled back, and which client-side blockers limited performance. This is where FiveX reporting helps agencies move from “here is your ACOS” to “here are the decisions we made, the profit guardrails we used and the next approvals we need.”
Where agencies should draw the line
My strong opinion: agencies should stop selling “AMS optimization” as if the work is only inside the Amazon Ads console. That phrase undersells the risk and overpromises the simplicity.
The better service promise is: Amazon Ads decisions with profit control. That includes campaign structure, bids, budgets and search terms, yes. But it also includes SKU economics, inventory, catalog readiness, automation governance, client approvals and decision memory.
If a client only wants cheaper clicks, they can find a cheaper operator. If they want a marketplace agency that protects margin while scaling Amazon, the change ledger becomes part of the product.
That is the opportunity. AMS may be an old name, but it points to a modern agency problem: Amazon advertising has become too connected, too fast and too financially important to manage as a set of isolated campaign tweaks. The agencies that win will not be the ones with the prettiest AMS report. They will be the ones that can prove why each meaningful change was allowed to happen.