Which of your products should be on Amazon next? A catalogue plan for manufacturers and wholesalers

Search for advice on adding products to Amazon and most of what you find is written for resellers. List 15 to 20 items, buy a few of each, add more every week. That works when you are buying stock to resell. It does not answer the question a manufacturer or wholesaler actually has: we make or stock thousands of products, a small share is on Amazon, so which ones go next?

The gap is often large. One prospect we reviewed this year had more than 25,000 products in its range and fewer than one in 25 of them listed on Amazon. In our experience that is common among manufacturers and wholesalers: the listed range gets all the attention, and the rest of the catalogue is never looked at with Amazon’s data.

This guide sets out how we choose. Score every unlisted product on three things, turn the scores into a profit estimate, launch a batch, and measure it. If you have not yet settled distributor pricing or who controls your listings, start with our guide to Amazon for manufacturers, which covers channel conflict, competition law and a first launch of 10 to 20 hero products. This guide is about everything after that.

Why “list more products” is the wrong starting point

Every product you add costs money before it earns any: photography, content, stock in Amazon’s warehouses, launch advertising and someone’s time. List a large range all at once and that money is spread so thinly that few products rank.

So the question is which products would make the most profit, and in what order. That needs three measurements per product:

• Proven demand. Are shoppers already searching for it on Amazon, and how often do those searches end in a purchase?
• Margin after Amazon’s fees. What is left from each sale after VAT, the referral fee and the FBA fee for its size, before advertising?
• How weak the rivals are. How much of that demand could a new listing realistically win?

Step 1: Cut the range to a shortlist you can research

You can’t research 20,000 products one at a time. A first pass with your own data cuts the range to a few hundred.

• Has a valid barcode and complete product data. No GTIN, no dimensions, no weight, no images: it can’t be listed well yet. Park it and fix the data separately (see the last section).
• Already sells somewhere. Trade orders, your own website, other marketplaces. A product nobody buys anywhere else is a weak first bet on Amazon.
• Isn’t already on Amazon through someone else. Search Amazon’s Add Products tool by each barcode⁹. If a reseller has already created a listing for your product, it is a different job: taking control of that listing, not creating a new one.
• Fits Amazon. Rule out products reserved for trade customers, and hazardous or very bulky items until you have checked the fees.

What is left is your shortlist: usually a few hundred products, which you can research properly.

Step 2: Measure demand with Amazon’s own data

There are three Amazon sources worth using, and it matters which one you use for products you have not listed yet.

Search Query Performance only covers searches associated with your branded products³, so for anything your range doesn’t already appear for, start with Product Opportunity Explorer.

Amazon describes Product Opportunity Explorer as a tool “to explore estimated customer demand for new product ideas"¹. A niche is a group of related searches, and for each search the Search terms tab shows “search volume, click share, search conversion and the top three clicked products"¹. The Insights tab covers today, 90 days and 360 days¹, which tells you whether demand is seasonal. Amazon says the tool is a guide only and “should not be a substitute for your own judgement"¹, so treat its numbers as estimates. Two practical points: the tool “will default to the US store on your first visit"¹, so switch it to the UK before you start, and check the period the search volume covers and convert it to a month.

For each shortlisted product, write down the one or two searches a shopper would most likely use to find it, their monthly search volume and their search conversion: the share of searches that end in a purchase. Our keyword research guide covers how to choose the right searches.

Step 3: Work out margin with the right size tier

Margin decides whether demand is worth having. Work it out per unit, before advertising, using Amazon’s UK rates:

• VAT. The standard rate is 20%¹¹, so one sixth of the price the shopper pays is VAT.
• Referral fee. A percentage set by category: Lawn and Garden is 15%, Tools and Home Improvement 13%, and Home Products 8% up to £20 and 15% above, each with a £0.25 minimum⁵. The fee is a percentage of the price the shopper pays, including VAT¹².
• FBA fulfilment fee. A flat fee per unit, set by the packed size tier and weight. Outside peak, UK fees run from £1.83 for a light envelope to £3.58 for a standard parcel up to 11.9 kg; oversize starts at £3.49⁶. From 15 October 2026 to 14 January 2027, parcels and large envelopes cost a little more⁶.
• Fuel and logistics surcharge. Amazon says it “will increase from 1.5% to 2%” of fulfilment fees from 2 November 2026⁷.
• Digital services fee. 2% of Amazon’s selling and FBA fees for UK sales⁵.
• Your landed cost. Product, packaging and delivery into Amazon.

The size tier is what people get wrong when they score a range from a price list. A standard parcel can be up to 45 × 34 × 26 cm; beyond that a product is oversize⁶. Use the packed size as it would ship, not the catalogue dimensions. Our FBA fees guide explains the tiers and dimensional weight.

The FBA calculator runs the sum for your own product.

Step 4: Judge how weak the rivals are

Demand you can’t win is worth nothing to you. Look at who takes the clicks and purchases today on each product’s main searches.

• Concentration. Both Product Opportunity Explorer and the Search Terms report show the top three clicked products for a search¹². Amazon defines click share as one product’s clicks after a search divided by all clicks after that search, and conversion share the same way for purchases². If three products take most of both, the search is held tightly and will be expensive to break into.
• Saturation. Product Opportunity Explorer’s Insights tab includes the number of products and top-clicked products in a niche, which Amazon says helps you “understand saturation in the niche"¹.
• Weak listings. Few reviews, poor images, missing details, slow delivery or a price well above yours are openings.
• Complaints. The Customer review insights tab groups review themes for a niche¹. A complaint your product solves is an opening.

We turn this into one number: the share of the search’s purchases a good new listing could realistically win within a few months. It is a judgement, so keep it simple. In our example we use 2% where three established products hold most clicks and purchases, 5% where the top is contested, and 10% where the top products are weak. Write down why, so you can check it later.

Step 5: Turn the three scores into one ranking

Multiply the three together and you have an estimate of monthly profit before advertising for every shortlisted product.

Here is how three products from an invented garden range compare, all in Lawn and Garden. All the figures are illustrations. Fees: Amazon’s UK non-peak rates (in force again from 15 January 2027) plus the 2% fuel and logistics surcharge that starts on 2 November 2026.

The sums: 20,000 × 8% × 2% = 32 units × £14.98 = £479.36; 6,000 × 10% × 10% = 60 units × £13.11 = £786.60; 9,000 × 10% × 5% = 45 units × £6.33 = £284.85 (or 45 × £7.68 = £345.60 in the 44 cm box).

During peak (15 October 2026 to 14 January 2027) standard parcels cost slightly more⁶: margins with the 2% surcharge fall to £14.85, £12.97 and £6.33, and the order does not change.

The planter has the fewest searches of the three and still comes first, because its rivals are weak. The tool set has the biggest search and the best margin, but three strong brands already hold it. A plan built on search volume alone would have launched the tool set first and spent its launch budget fighting for 2% of the purchases.

These are estimates before advertising and storage, not forecasts to bank, and they are only as good as the share you chose in step 4.

Step 6: Launch a batch, and write the forecast down first

Take the top 20 to 30 products from the ranking as your first expansion batch. In our view that is enough to learn from and few enough to give each one proper images, content and launch advertising. Before anything goes live, record each product’s searches, expected share and monthly profit estimate, so you have something to judge the results against.

Check whether the batch qualifies for Amazon’s New Selection Programme (2026). For a branded parent ASIN that is new to FBA, Amazon reduces the referral fee to 10% (or your existing rate, if lower) on the first 100 units and 5% on the next 100, sold within 120 days of your first FBA inbound, and adds free storage, free returns processing and free liquidation on the first 200 units⁸. In the planter example, 10% instead of 15% on £59.99 is worth about £3.00 a unit on the first 100.

Check the conditions. “New-to-FBA” means the parent ASIN “has not had an FBA shipment in the past 12 months by any seller"⁸, so a product a reseller already sends into FBA won’t qualify. You need a “Maximum trailing 6-month IPI score of 300 or higher (if you have an assigned score)”, and Amazon excludes “Books, DVD, Music, Software and Computer/Video Games, Video Game Consoles and Video Game Accessories"⁸. Sellers moved over from the earlier version of the programme had to accept the updated terms by 31 October 2026 to keep getting benefits on new ASINs⁸. If that applies to you, check your status in the seller incentive dashboard.

Our product launch guide covers the launch itself: Vine, vouchers, the first advertising and a 90-day budget.

Step 7: Measure at 30, 60 and 90 days

Once the batch is live, Amazon’s own reports tell you whether each product is meeting its estimate.

• Search Catalog Performance. One row per ASIN: search impressions, clicks, basket adds and purchases across all searches⁴. Seen but not clicked is usually an image, price or review problem; clicked but not bought, a listing problem.
• Search Query Performance. Now that the products are listed, their searches start to appear in your brand’s view³. Compare your share of purchases on each chosen search with the share you expected.
• Profit per product. Units, price, fees, advertising and returns, product by product. Our Profit X-Ray does this, and our guide to diagnosing a sales drop explains how to split sales into traffic, conversion and price.

At 90 days, put each product in one of three groups. This is how we do it; set your own thresholds from your margins.

Then re-score the shortlist and pick the next batch. After two or three rounds, your share estimates rest on your own results.

Make the barcode the single source of truth

Every step above depends on joining your product records to Amazon’s catalogue and reports. In a large range, that join is where things go wrong.

Amazon matches products by barcode. Its UK help page says product IDs “help ensure product page accuracy by matching listings to products that already exist in the Amazon catalogue”, and that it will “confirm the product IDs against the GS1 database"⁹. Using false product IDs is prohibited and “can result in your ASIN creation privileges being removed"⁹.

So use the barcode (the GTIN: a 13-digit EAN or a 12-digit UPC⁹) as the one key that joins everything: your product system, your trade price list, your Amazon listing file and your reports. On one large range we worked on, matching products to Amazon listings by title and colour paired about one product in ten with the wrong listing, often a near twin in another size or variant. We moved to barcode-only matching and kept title and colour as a check, not the key.

Four Amazon rules to plan around:

• Packs need their own barcode. Amazon says that for a bundle or a pack you must “add the correct product ID for that pack and item quantity rather than the product ID for a single item contained in the pack"⁹. Give each Amazon-only multipack its own GTIN.
• No exemption if a GS1 barcode exists. You cannot apply for a product ID exemption if the product “has an existing GS1-approved barcode"¹⁰.
• Exemptions can be withdrawn. Amazon “may expire previous product ID exemptions for brands that initially didn’t have Global Trade Item Numbers, but have since acquired GS1 barcodes"¹⁰. Buy GS1 barcodes later and expect to move exempted products onto them.
• Exemptions are by category, not by product. Once approved, one exemption covers every product in that category, but you need a separate one for each store and category¹⁰.

Before your first expansion batch, run a barcode audit: every product with its GTIN, whether that GTIN already has an Amazon listing, who sells on it, and whether its packed dimensions and weight are recorded. It is dull work, and it makes steps 2 to 5 trustworthy. Our guide to Brand Analytics shows how to read the search reports once the data joins up.

We can do this with you; it starts with a Profit Audit. Our How we measure page explains how we work out every £.