Every August we get the same call. A brand doing $150K–$400K a month on Amazon has just finished their mid-year review, looked at twelve months of agency invoices, and done the arithmetic: $6,000 a month is $72,000 a year, and you can hire an experienced Amazon brand manager for $85,000. Add a little for benefits and it’s roughly a wash — except now the person is yours, full-time, in your Slack, not splitting attention across other brands.
It’s a completely reasonable calculation. It’s also the wrong one, and not for the reason you’d expect an agency to give you. We’re not going to argue that in-housing is a mistake — for a meaningful number of brands it’s exactly right, and we’ll tell you which ones. The mistake is that brands compare a salary to a retainer, when the thing they’re actually replacing is a coverage surface.
Here’s the honest version of that math, the five ways this goes wrong, and the specific reason August is the worst month of the year to make this call.
The comparison brands run vs. the one they should
The standard comparison is one line against one line. Agency retainer versus salary. On that basis in-housing usually looks like a small win.
The real numbers, as of 2026: an Amazon brand manager runs $82,000–$95,000 in base salary, and fully loaded — payroll taxes, benefits, equipment, software seats, recruiting cost, and the management time of whoever they report to — you’re realistically at $130,000–$150,000 for that one person. That alone reframes a $72,000 retainer.
But the loaded-cost correction isn’t the interesting part. The interesting part is what one person can actually cover.
A functioning Amazon program is five distinct disciplines:
- Advertising — campaign architecture, bid management, search-term hygiene, budget pacing, and the diagnostic skill to tell a listing problem from a campaign problem
- Creative — hero images, image stacks, A+ content, video, and the merchandising judgment behind them
- Catalog and content — listing structure, attributes, variations, flat files, and staying ahead of Amazon’s constant field-level changes
- Operations — inventory planning, restock limits, shipment problems, FBA fee classification, aged-inventory triage
- Account health and brand protection — policy compliance, suppressions, Brand Registry, reseller enforcement, case management
These are genuinely different skill sets. The person who is excellent at bid strategy is very rarely the person with the eye to tell you your slot-two image is the reason your conversion rate is soft, and neither of them is the person who knows what the aged-inventory surcharge does to your Q4 margin.
Your $85,000 hire is strong at one or two of these and passable at the rest. That’s not a knock on them — that’s what $85,000 buys, and it’s what almost every job description in this market is quietly asking for. To genuinely replicate the surface, you’re looking at a minimum of three people: an advertising specialist, a content and catalog manager, and someone owning operations and account health. That’s $200,000–$350,000 in salaries before tools, benefits, or the person managing them. All-in, a real in-house team runs $12,000–$20,000 a month.
So the honest comparison isn’t $6,000 versus $7,000. It’s $6,000 for five disciplines at partial allocation versus $11,000 for one discipline at full allocation. Both are defensible purchases. They’re just not the same purchase, and brands routinely sign up for the second thinking they bought the first.
The five mistakes we see
1. Hiring for the job you have, not the job you’ll have in six months.
Brands write the job description around their current pain — usually ACOS. They hire a PPC specialist. Six months later the thing hurting them is a catalog problem, a suppression, or a creative refresh that never happened, and the PPC specialist is doing it badly because they’re the only person there. Hire for the discipline that will still be your constraint in a year, not the one that’s loudest this week.
2. Not budgeting for the tool stack.
Agencies amortize software across their whole book. You won’t. Advertising software, listing and catalog tools, keyword and market research, analytics, creative production — depending on your stack this runs $500–$2,500 a month on top of salary, and it’s almost never in the spreadsheet that justified the hire. It shows up in month two as an unbudgeted surprise.
3. Ignoring single-point-of-failure risk.
This is the one that actually costs money. When your Amazon manager takes two weeks off in December, who runs the account? When they quit in March — and Amazon talent turns over fast, because everyone in this market is being recruited — what leaves with them? The bid logic they never documented. The negative keyword rationale. The relationships in Seller Support. Which flat file does what.
We take over accounts after this exact event regularly, and the recovery is not fast. The institutional knowledge walks out and the account coasts on autopilot for the eight to twelve weeks it takes to hire and ramp a replacement. On a $300K/month account, a quarter of coasting is a far larger number than the retainer you were trying to save. If you in-house, document as though the person is leaving, from week one.
4. Underestimating the ramp.
Even an experienced Amazon hire needs 60–90 days to learn your catalog, your margins, your supplier constraints, your seasonality, and where the bodies are buried. That’s not a reflection on them; it’s the same silent quarter any transition carries. Brands budget for the salary starting in month one and the output starting in month one. Only one of those is true.
5. Confusing “in-house” with “in control.”
The pitch brands tell themselves is control. But control comes from knowing which numbers matter and asking for them — and a brand that never developed that muscle with an agency doesn’t suddenly acquire it by moving the work inside. We’ve seen in-house programs that are less transparent than the agency they replaced, because at least the agency sent a monthly report. If you can’t currently tell whether your ad spend is buying growth or buying efficiency, hiring someone doesn’t answer that question. It just moves who you’re not asking.
The August timing trap
Now the part that’s specific to right now.
If you’re reading this on August 1 and considering a hire to get ready for Q4, run the calendar forward honestly. A search started this week realistically closes in four to six weeks. Good Amazon operators are employed and have notice periods, so add two to four more. That’s a start date in late September or October — landing your new hire in the account precisely as peak begins, with zero context on your catalog, during the highest-stakes ten weeks of your year.
Meanwhile the decisions that determine your Q4 are being made now: inventory is already committed against the BFCM cutoffs, deal submissions are closing, creative needs to be validated by late September, and Q4 ad budgets have to be built off peak CPCs rather than Q3 baselines.
A hire made in August does not help your Q4. It complicates it. The honest sequencing is to make the transition decision in January, when a new person has three quiet quarters to learn the account before it matters — or to make it now and accept that the ramp is a Q1 investment, not a Q4 fix. What you should not do is treat the hire as your peak-season plan.
When in-housing is genuinely the right call
We’d rather tell you this than have you find out at month nine.
In-house wins when:
- You’re past roughly $5M a year on Amazon, where the work is genuinely full-time and the fully-loaded cost of specialists is justified by the revenue they’re managing
- Your operation is complex in ways that reward embedded knowledge — heavy variation catalogs, regulated categories, complicated bundling, or manufacturing constraints that require someone in your supply chain meetings
- Amazon is your primary channel and your primary strategic surface, not one of five places you sell
- You can hire more than one person. A team of three covering the surface beats an agency for many brands. One person covering the surface beats nothing.
Agency wins when:
- You’re doing under about $2M a year on Amazon, where a fully-loaded specialist is a large fraction of your contribution margin
- Your needs are spiky rather than constant — a creative rebuild, a launch, an international expansion — which is a bench problem, not a headcount problem
- You need five disciplines at partial allocation more than you need one at full allocation
- Continuity risk matters more to you than control does
And the honest middle, which is where most brands between $2M and $10M land: keep the specialist function in-house where embedded knowledge compounds — usually operations, inventory and account health, because those are yours and nobody outside your business can forecast your demand — and buy advertising and creative from a bench. That model puts your headcount where institutional knowledge is worth most and buys the disciplines where breadth and pattern density beat proximity.
FAQ
How much should I budget for a real in-house Amazon team?
One fully-loaded specialist is $130,000–$150,000. A three-person team covering the full surface is $12,000–$20,000 a month all-in including tools and benefits. If your plan is one hire at $85,000, budget honestly for what that person won’t cover.
Is one experienced generalist enough?
Below roughly $2M a year, often yes — the volume of work in each discipline is small enough that a strong generalist keeps the plates spinning. Above that, the generalist becomes the bottleneck, and the first thing to slip is always creative, because it’s the one with no daily alarm attached to it.
What if my agency isn’t performing? Should I in-house instead of switching?
Diagnose first. If the problem is that your agency has your account under-resourced, that’s a specific and answerable question — ask how many accounts your account manager runs. A healthy load is roughly 8–15; some shops run 30 or more, which no amount of automation covers. A bad agency is an argument for a better agency, not automatically an argument for headcount.
How do I protect against my Amazon manager leaving?
Documented SOPs from week one, admin-level access held by the business and not by an individual, a second person with Brand Registry access, and a standing export of your search-term and negative-keyword history. Everything you’d secure before switching agencies, secure before your employee’s last day — and you won’t know when that is.
Can I in-house PPC and keep an agency for creative?
Yes, and it’s one of the better-performing splits we see. Advertising rewards daily proximity to your margins; creative rewards seeing hundreds of listings across categories. Just make sure someone owns the handoff, because the most common failure in a split model is a listing problem being diagnosed as a campaign problem by the only person looking.
—
The brands that get this decision right aren’t the ones who ran the cheapest comparison. They’re the ones who wrote down all five disciplines, marked honestly which ones they’d be buying at full allocation and which they’d be leaving uncovered, and priced the gap. Whatever you don’t staff, you’re still going to pay for — just in aged inventory, a suppression nobody caught, or a creative refresh that slipped another quarter.
If you’re looking for a team that manages every lever — creative, advertising, and operations — Velocity Sellers works with brands doing $100K+/month on Amazon. Contact us for a free account audit.
