Amazon is the whole business
If a single marketplace is most of your revenue and you intend to keep it that way, the capability is strategic and belongs inside. Buy expertise to start; own it once it is the company.
Amazon service · Build or buy
Hire the team, or hand over the work. Both are a real answer.
Every manufacturer that decides to take Amazon seriously arrives at the same fork: build the function in-house, or give it to an operator. This is the honest version of that comparison - what the in-house route actually requires, what it buys you, and where handing it over stops making sense.
Sell your product with us → Seller Central vs Vendor Central →
What the job actually is
It is not one hire. It is a department.
The reason build-versus-buy is a hard decision is that the “build” side is almost always underestimated at the start. An Amazon operation is not an Amazon manager. It is a standing set of functions, each of which has to be covered on the days the other ones are busy.
This is the list from our own org chart, because it is the list we staff ourselves:
Marketplace engineering - the channel connectors, the middleware that moves orders and stock between the marketplace and your ERP, and the secure access pattern the account logs in through. This is the function manufacturers almost never plan for and the one that breaks first.
Content and listings - titles, backend keywords, A+ content and imagery, built to a marketplace’s standard and rebuilt per language for every market you list in.
Pricing - not a spreadsheet. Someone reading the competition per listing, every day, and deciding where margin gets defended and where the price has to move.
Advertising - campaign structure, bids and budget, in every marketplace, against a return target somebody has to own.
Account health - the daily metrics that decide whether an account keeps selling. A suspension costs a season, and it is preventable only by watching it before it trips.
Customer care - buyer messages answered inside the marketplace’s own response windows, in the buyer’s language, including the weekend.
Logistics and returns - carrier selection per parcel, the paperwork at each border, and a returns path that runs the whole chain backwards.
Compliance and finance - registration and filing in every jurisdiction you sell into, and reconciliation against marketplace settlements that do not look like invoices.
The whole picture, drawn: what one Amazon sale actually rests on →
When in-house is the right answer.
We are not going to pretend this decision only goes one way. There are manufacturers who should build, and the pattern is consistent enough to describe.
One market, one language
Most of the cost of this function is multiplication - per marketplace, per language, per jurisdiction. Selling in one country makes the in-house version a genuinely small job.
You already have the systems
If the ERP is already integrated, the content already exists in every language you need and someone already owns pricing, you have paid the hard part. Do not hand it over to save the easy part.
What handing it over looks like
Passadena kept the account. Two of our people run it.
In October 2025 Passadena, a Greek fashion retailer carrying more than a hundred international labels, handed us its Amazon and Skroutz operations. Nothing changed hands: Passadena stayed the seller of record, the storefront and the ratings stayed in their name, and the buyer stayed theirs. Only the work moved.
We took over four jobs in one go - account health checked daily, marketplace access on fixed IPs and dedicated desks, their existing Entersoft ERP wired to the marketplaces through Linnworks, and buyer messages answered in our CRM. The flat catalogue-wide discount they had been running was replaced with pricing that reads the competition per listing.
We deliberately skipped Black Friday that year. Launching clean mattered more than launching fast. By the end of the quarter it was fully operational, and it runs today on two people on our side: a project manager and an operations lead.
Questions
Build-or-buy questions we get asked.
Why does an operator need fewer people than we would?
Because almost none of the functions above are full-time for one brand. A pricing analyst reading one catalogue is idle by lunchtime; the same person reading forty is not. We built the platform underneath this to run our own brands first, so adding an account does not mean adding a proportional number of people. That is the entire economic argument, and it is why two of our people can run what needed a team inside the brand.
Do we lose control if we hand this over?
That depends which model you pick, and it is worth being precise about. Under Operations as a Service you keep the accounts, the brand and the seller-of-record status, and we run the work underneath - that is what Passadena does. Under D2C or Distributor we buy the stock outright and sell as the merchant, which is genuinely handing over the commercial relationship. Five models, with the trade-offs printed, are on the models page.
Can we start with an operator and bring it in-house later?
Yes, and under OaaS it is the cleanest version of that plan, because the accounts were never ours to give back. You keep the account, the listing history and the ratings throughout. What you would be taking on is the work and the people to do it, at a point where you know from your own numbers what that work is worth.
What if we only need part of it?
That is common, and usually the honest starting point. Most manufacturers who call us already do their own content and want pricing, advertising and care covered, or they have the commercial side handled and need the integration and the compliance. Tell us which parts are already working and we will scope to the gap rather than to the whole.
Amazon is a trademark of Amazon.com, Inc. TPL S.A. operates on this channel as an independent seller and is not endorsed by, sponsored by, or affiliated with Amazon.com, Inc.