Run the arithmetic before the org chart. A single product launch needs $8,000 to $15,000 in capital and about seven months either way. Internal wins when you already own advertising, creative, and sourcing depth. External wins when buying those three costs less than hiring one person who covers two.
The short version
- The capital requirement is identical. $8,000 to $15,000 for one product, $25,000 to $50,000 for a five-product brand, whoever executes it.
- Only the labor line moves. That is the whole financial difference between the two playbooks.
- Seven months is the honest duration for a full brand launch. Compressing it is where money disappears.
- Validate before you scale. 200 units and $5,000 to $10,000 buys the data that decides everything after it.
- Ask how much attention you are buying, not how much capability exists somewhere in the building.
The number that does not change either way
Whichever playbook you pick, the launch capital is the same. Inventory, freight, trademark filing, samples, photography, and the ad budget that carries the first ninety days come to $8,000 to $15,000 for one product. A five-product brand lands between $25,000 and $50,000. Both paths buy from the same factories, ship on the same lanes, and pay Amazon the same fees.
That narrows the decision considerably. You are not choosing between two business models. You are choosing how to pay for one layer of the stack, the labor layer, while every other line on the sheet stays where it is.
Price the labor layer properly
| Cost line | Internal build | External partner |
|---|---|---|
| Advertising execution | A salaried specialist, fixed from day one | Inside the monthly fee |
| Creative production | Freelancers per asset, or a studio retainer | Inside the fee only if the firm owns a studio |
| Sourcing and quality control | Travel, inspection fees, or an agent commission | Inside the fee only if the firm has people in China |
| Software and data | Your licenses, annual commitments | Usually the firm's, already paid for |
| Ramp before productive | One to three months of salary with no output | Starts at the audit, first fixes inside 30 days |
| Cost of being wrong | A redundancy process | 30 days' notice |
Two rows in that table decide most comparisons. Ramp is a real cost that never appears in a salary budget, and the cost of being wrong is what you pay when the category turns out smaller than you hoped. A month-to-month arrangement prices that risk at 30 days. An employment contract does not.
Our published fee for the managed path runs from $800 a month for one product to $2,400 for five, all services included at every tier, with no commission and no onboarding charge. Put your own numbers on the internal column and compare the annual totals, not the headline.
The seven-month shape of a launch
- Research and market sizing. 90 or more data points before a product is chosen: growth trajectory, return rate, segment dynamics, and the rating gap that shows where the incumbent is weak.
- Sourcing and samples. Supplier shortlist, samples, a specification built from competitor negative reviews rather than from invention.
- Listing and creative build. Copy, images, A-plus content, and video, produced against the differentiation you found in step one.
- Phase one validation. 200 units, $5,000 to $10,000, up to four products tested at once. You are buying evidence on rating, conversion rate, and cost of acquisition.
- Phase two scale. Only once those three are proven. Inventory depth, channel expansion, and budget follow the evidence, never precede it.
Internal teams tend to compress steps one and four because they feel like delay. They are the two steps that protect the capital in the other three.
The ratio behind quality of attention
Capability on a website is not attention on your account. The question that separates the two is how many brands one person carries.
I run Flapen out of Abu Dhabi with a team of 50 operators looking after about 70 brands, which works out near 1.4 brands per operator. That number is the reason we can hold a launch cadence. Ask any external candidate for their equivalent figure, and ask the same of your internal plan, because a single hire covering your whole catalog plus a launch is carrying more than 1.4 of anything.
What a proposal will not tell you
The fee is the small number. On a $12,000 launch, three months of management is a fraction of the capital at risk, and choosing a partner on price alone optimizes the least important line. The expensive mistakes are a category that cannot carry the acquisition cost, a supplier who ships inconsistent units, and a launch that runs six weeks before anyone checks the conversion rate.
The second thing: neither playbook rescues a product that should not have been chosen. Internal teams protect their own product decisions because those decisions are theirs. External teams sometimes protect them because a killed product shrinks the account. Whichever you pick, write the kill criteria down before launch and give someone permission to invoke them.
Related answers
- Compare agency retainers vs in-house salaries for Amazon
- When to hire an Amazon agency
- Hybrid model for Amazon ops: what works
- How to launch your first product on Amazon
- Amazon agency pricing and economics: the complete guide
Bring us your launch numbers and we will size the market before quoting anything, at Flapen.
