Many manufacturers make products that reach buyers under other companies’ names. They know the product better than anyone, yet the brand on the shelf sets the price and keeps much of the margin.
An own brand on Amazon changes that. The hard part is choosing where to enter, because a category that looks full from the top can still hold an open lane.
This is how we plan that entry. We clean the market data, find the lane, check the margin and test a first version before anything scales.
Where the manufacturer started
The manufacturer in this build had a long production history and real product competence, but little visibility with end customers. An own brand on Amazon was on the table as its next growth move.
Amazon’s Brand Registry accepts a pending trademark application only in some cases. Brands without a registered mark are pointed to its IP Accelerator programme. So the trademark office is checked first, before any brand work runs ahead of the registration.
We mapped the category and looked for the place where a new brand could win. Then we wrote a staged plan with a stop after the first two days. The ranking and review charts redraw the shape we found.
Clean the market before sizing it
Marketplace research tools make a category look bigger than it is. Summed row by row, the listings in the brand exports put this category at a multiple of its real size. Each row is one listing, yet the revenue column repeats the parent product’s total on every variant.
We collapse the listings into their parent products before any total is calculated. The ranking of brands holds either way. The market size does not, and a business case built on the raw total would carry the error.
One price per unit of content
Listings in this category are sold as single packs, multipacks and bundles of different sizes. Compared by list price, a large pack looks expensive and a small one looks cheap, and neither comparison means much.
We read the pack content out of every title and turned each listing into a price per unit of content, the way a shelf label shows it. Only then could we sort the market into price bands, from a crowded cheap band to a thin premium band.
Reviews guarded this market more than prices did
We then plotted every relevant listing by price per unit of content against its number of reviews. Most listings had few reviews. The ones with the longest review histories sat in the cheap band, with more than a thousand reviews each.
In our reading, a newcomer cannot buy that history quickly, and a lower price would not beat it. Programs like Amazon Vine help at the start. Vine lets a brand-registered seller using fulfillment by Amazon collect up to 30 reviews per parent product. Vine also runs in Germany, France, Italy and Spain, and it belongs in the launch budget, but thirty reviews do not close a gap of more than a thousand. Higher up the price scale the picture changed. One premium listing also had a long review history, but some listings priced several times higher per unit had far fewer reviews. In the middle band, between the cheap band and those premium listings, a new brand competes mainly on the product, where a manufacturer has real advantages.
Aim at the gap below the leaders
Ranked by monthly revenue, a few leaders stood out, and the brands behind them formed a plateau. Between the sixth and the seventh brand, revenue dropped sharply, and below that the long tail began.
The target was the gap, not the leader. Above it, a new brand would have to take share from established names. Below it, a brand tends to stay too small for its logistics to scale. The gap gives a focused range a size to aim for, and it set the revenue goal of the plan.
The lane came from combining attributes. Each single attribute in the category, such as the base material or the processing method, was already crowded. In our reading of the market, the price premium sat where two of them met, and that combination was still rare.
Margin decides before the products do
A manufacturer selling direct can keep part of the margin that intermediaries would take. The plan earmarks that share for customer acquisition on the marketplace, where a new brand has to buy its first visibility. That is the thesis of the plan, and only a first version can test it.
Every candidate product then runs through three filters, in this order. It has to fit the brand. Its unit economics have to work after production, fulfillment fees and preparation for the fulfillment center. And it has to clear a minimum margin, set before the search begins.
A first version in one small batch
The first version in the plan is deliberately small. A couple of listings and one small batch, packed by hand and sent to the fulfillment center as parcels. Pallets, a third-party preparation service and new ERP processes come only after the market has answered.
The first two days of the work check the account and validate the data, and a stop-or-go decision follows. In the plan, a go means the checked data still shows candidate products that pass the three filters. Without a clear go at that point, the work stops and the spending ends there.
The first version has its own gate. If buyers do not respond, the plan adjusts it and tests again, and a second no stops the work there too. Each later stage opens only after its own go: a preparation service, then processes in the ERP, then pallet freight, then more variants.
The work after launch
A plan without the operating load understates the cost of the move. We wrote that load into the plan from the start: batch tracking, packaging rules for the fulfillment center, advertising, regular updates to the offers and stock planning.
Marketplace rules also change often, partly because of new EU requirements. For an own brand sold in the EU, some duties are concrete from the first sale. The General Product Safety Regulation has applied since December 2024 and brings safety and traceability duties for each product. In Germany, the packaging has to be registered in the LUCID packaging register before it is first sold. Storing stock in other EU countries through Amazon’s pan-European fulfillment brings VAT registration in each of them.
Keeping up with these rules is part of running an own brand and does not end at launch.
What the analysis rests on
Revenue figures in marketplace research tools are estimates, not actual sales. The price bands rest on one snapshot of list prices, without promotion history. Both serve to rank brands and find a gap. Sales data from your own first version then takes over.
The price split in the margin chart is a template, not a costed model, and it starts from the price after VAT. Break-even volume, cash tied up in the first batch, returns, storage fees, unsold stock and the length of the expensive advertising phase need real product costs.
If you are weighing your own brand
- Clean the data before you size the market. One repeated revenue column can inflate a total several times over.
- Compare products on the unit buyers pay for, the price per unit of content a shelf label shows.
- Plot price against reviews. Where reviews are thin and prices are high, a newcomer often has room.
- Aim at the revenue gap below the leaders, and set the margin filter before choosing products.
- Put a stop-or-go decision after two days, before any stock is packed, and keep the first version small.
If your factory already produces for other brands, we can run the same analysis on your category with you. You get the gap, the margin filter and the first version, each with its stop-or-go written down before any stock moves.
Credits
Market analysis, business logic and staged plan by Capcelerate.