Most sellers can repeat what their agency promised. Far fewer could tell you what anyone did to the account last Tuesday.
That gap is the whole problem. "Full service" is a sales label, not a scope, and the difference between the two only surfaces around month six.
This covers how a full-service Amazon agency actually runs a store: what they take over, what happens in the first 90 days, what a normal month looks like, what it costs, and where the model stops working.
How this was put together:
- Statements of work compared across agency, hybrid, and managed-platform models
- Cadence and staffing claims checked against what sellers report publicly, not against agency sales pages
- Platform mechanics verified against Amazon's own Seller Central and Amazon Ads documentation
Xneeti is one of the models covered here. The questions below work whoever you end up hiring.
By the end you'll know what to expect in month one, what to demand in month three, and when to leave.
What "full service" actually covers on Amazon
Nothing stops a PPC shop from calling itself full service. The label is free. The statement of work is the only document that describes what you bought.
Nine areas make up a real management engagement.
The top row is the one every agency is good at. Amazon sponsored ads are the easiest part of an account to manage well, and the easiest part to mistake for the whole account.
Now look at the last three rows. Inventory coordination, account health, and finance recovery are where most retainers quietly end. They're also the three that cost you money silently instead of visibly, which is why nobody notices they're missing.
How an agency takes over your store: the first 90 days
Engagements rarely fail in month eight. They fail in week two, when access is half-granted and nobody wrote down what the numbers looked like before anyone touched anything.
Access is where control quietly changes hands. An agency should work under named user permissions inside your Seller Central and your ad account. You keep Brand Registry. You keep the DSP seat. If they ask for a shared login or hold the seat in their own name, you're renting your own store back from them, and you'll find out the day you try to leave.
The baseline matters more than it sounds. Without one both sides agreed to, month three becomes an argument about what performance looked like in month zero. Pull it from your Amazon ads dashboard and your Business Reports on the same day, and get it in writing.
The recurring cadence: what happens in a normal month
Steady state is a loop. How long that loop takes to close tells you more about an agency than any case study on their site.
Week 1: performance review and catalog sweep
Search term reports get mined, negatives added, bids moved. The Amazon PPC ads side of this is the part everyone does.
The part fewer do is checking the top ASINs for Buy Box loss, suppressions, and new competitor entries in the same pass. An ACoS drop means nothing if you lost the Buy Box on your best seller last Thursday.
Week 2: content and creative
Conversion by ASIN decides the creative queue. Listings that slipped get new imagery or A+ modules, and Brand Analytics shows which competitors moved into your priority terms this month.
This is also when Amazon video ads creative gets planned, which for most sellers is the thing that never happens. Production is slow, nobody owns it, and it drops off the list every month until someone makes it a standing item.
Week 3: upper funnel and brand protection
DSP audiences get rebuilt from sponsored search signals, and new-to-brand rate gets checked against what you spent to get it. The hijacker sweep runs whether or not anyone complained that month.
Week 4: reporting, inventory, planning
Profit by SKU gets compiled, rank movement reviewed, days of supply checked against next month's planned spend. Running aggressive campaigns on fifteen days of inventory is how sellers delete their own rank.
Here's the part that never appears in a proposal.
The left column moves hourly. The right column moves weekly. Everything that happens in between is invisible until it reaches revenue, and by then it's a quarter old.
Where full-service management usually breaks down
These aren't bad agencies. They're structural problems, and they show up in almost every engagement that stalls.
Account load is the number nobody asks for. A strategist carrying thirty accounts is running templates, whatever the retainer says, because there aren't enough hours in a month to think properly about thirty Amazon businesses.
Ask how many your named strategist carries. Then ask what they did on your store last week, specifically.
How to tell whether your store is genuinely being managed
The monthly report is the confession. If everything in it came out of the ad console, the ad console is the whole engagement.
Count the rows. If fewer than four appear in your monthly report, you're paying full-service rates for PPC management. That isn't always the wrong trade, but you should know it's the trade you made.
What full-service Amazon store management costs
The range is wide because the label means nothing. Compare the model and what sits inside it, not the headline number.
Figures are observed market ranges, not fixed rates.
Four line items get quoted separately after signing more often than any others: design, A+ production, new listing creation, and DSP. Ask about all four before you compare quotes. A $4,000 retainer with three of them billed hourly is not cheaper than a $7,000 retainer that includes them.
Scope also varies more than rate cards suggest across Amazon ads management services, so two agencies quoting the same number can be selling very different amounts of work. Your category changes the math too. What Amazon ads cost per click decides how much of that retainer the account can realistically pay back.
What changed: managing a store now means managing AI discovery
Amazon's COSMO layer powers Rufus, and Rufus reads your listing to answer a shopper's question rather than match a string. Most catalogs were written for the opposite job.
Ask any agency how they handle Rufus. If the answer is keyword density, they're managing a 2019 store.
Questions to ask before you hand over your store
Every bad engagement was avoidable on the sales call. Eight questions surface the gaps while you can still walk away.
- Ask for the statement of work line by line, and which of the nine areas above sit outside it.
- Ask how many accounts your named strategist carries at the same time.
- Ask who makes catalog edits, and whether that person has Seller Central access to do it themselves.
- Ask how often bids get reviewed, and whether that's a person or software.
- Ask to see a real client report with the numbers redacted, not a sample template.
- Ask what happens to campaign structures, creative files, and rank history if you leave.
- Ask specifically how they handle Rufus and AI-assisted discovery.
- Ask for one account where performance dropped, and what they changed in response.
Most Amazon product ads management companies will answer the first four without blinking. Questions five through eight are where the differences appear.
Four red flags, any one of which is enough to walk: guaranteed rankings, a shared login, a DSP seat in the agency's name, and a monthly report containing only ACoS.
When you don't need full-service management
Full service is the wrong call for plenty of sellers, and an agency that won't say so is selling rather than advising.
- Under roughly 25 SKUs, with a founder still close enough to the catalog to spot problems personally
- Ad spend below $10,000 a month, where bid work alone covers most of the available upside
- One marketplace, one category, stable seasonality, no unauthorized resellers
- Listings that were never built properly, where a rewrite beats a retainer by a wide margin
- A strong internal team with one specific gap that a specialist closes more cheaply
If you're keeping it in-house, Amazon ads software covers most of the day-to-day work. The real question is whether anyone opens it on a schedule.
The economics flip past 50 SKUs, past $15,000 in monthly spend, or the first time a reseller undercuts your price. Whichever comes first.
Where the agency model hits its ceiling
None of this is a competence problem. A weekly review cycle can't cover a platform whose auction resets every hour, and hiring a better strategist doesn't change that.
Three limits, all of them staffing rather than skill. Cadence, because people review on a schedule and Amazon doesn't. Account load, because attention divides. And the split between ads, listings, and inventory, because those three usually sit with different people who meet on Thursdays.
Which is why the question has moved. It's less about who manages your store, and more about what runs between the human reviews.
How Xneeti manages your store
Continuous execution and strategic judgment are two different jobs. Most agencies ask one overloaded strategist to do both, and execution is the half that slips.
Xneeti splits them. The AI layer adjusts bids by hour, day, and placement against your account's own conversion patterns rather than category averages. It tracks competitor keywords into your campaigns and runs n-gram analysis across your search terms, adding negatives before the waste reaches your ACoS report. Sponsored Brands Video gets generated through an in-house module, which removes the production bottleneck that stops most sellers running video at all.
It also covers the three rows most retainers drop. The SEO module rewrites listings for both A10 and Rufus. The inventory predictor watches sales velocity, ad spend rate, and supplier lead time together, then flags a reorder before the danger window instead of after the stockout. Payout intelligence reconciles every Amazon financial event and tells you in plain English what landed and what's still owed.
On top of that sits a dedicated strategist who reviews every action the AI takes, carrying roughly half the accounts of an industry-average manager. This article told you to ask about account load, so that's the number worth stating.
The fit is scaling brands whose catalogs outgrew manual review. Under 20 SKUs, running it in-house is still cheaper, and we'll say so on the call.
If you want to see what a continuous layer would find in your account, book a demo and get the gaps mapped before you commit to anyone.
