A checklist is not useful because it has many boxes.
It is useful because each box answers a decision question, records the evidence and makes the unknowns visible.
That distinction matters when you are choosing products for Amazon. “Demand checked” might mean you looked at a Best Sellers page for thirty seconds. It might also mean you recorded a signed-in marketplace signal, compared a trend over time, reviewed the competitive offer and stated exactly what still needs proving.
Those are not the same check.
Use this page as a working record before supplier outreach, sampling or a first order. It is designed to sit underneath the broader guide to finding products to sell on Amazon UK and Europe. It does not replace the detailed product opportunity scorecard.
The examples are illustrative. No supplier, product test, sales result or Amazon dashboard outcome has been invented.
1. The three status labels
Use one of these labels for every important row:
- Verified: supported by a source or document you can inspect, with a store, scope and date.
- Estimated: a clearly labelled model input or third-party estimate that still needs testing.
- Unknown or blocked: missing, contradictory, inaccessible or too important to assume.
Do not use “yes” as a status when the evidence is only a guess. A good checklist makes it easy to write “not yet proven”.
2. How to use the checklist
- Create one record per exact product configuration.
- Set the marketplace, country, currency and date before collecting evidence.
- Save the source URL or document beside every material input.
- Mark whether a signal is actual, relative, estimated or assumed.
- Write the stop condition before a supplier or product list persuades you to buy.
- Finish with one next action and one owner.
If the candidate changes size, material, pack count, brand, supplier or fulfilment route, open a new record or version the old one. Do not quietly replace the assumptions in an old record.
3. 1. Candidate identity
Start with the thing you are actually researching.
| Check | Evidence or input | Status | What it does not prove |
|---|---|---|---|
| Product name is specific | Plain-language description and use case | That the listing or supplier uses the same name | |
| Exact configuration is recorded | Dimensions, material, colour, size, pack count and condition | That the supplier can produce or supply it | |
| Target marketplace is stated | Amazon UK, a named EU store or a defined expansion set | That one store's data transfers to every country | |
| Customer problem is written in one sentence | Problem, occasion or job to be done | That customers will pay your target price | |
| Product identifiers are mapped | EAN, UPC, GTIN, MPN or ASIN where relevant | That an identifier proves brand authority or listing approval | |
| Research date is recorded | Date and time zone | That a dynamic result will remain unchanged |
If you cannot describe the product precisely, pause the rest of the checklist. A vague product creates false matches in demand, supplier and fee research.
4. 2. Demand evidence
Use more than one signal, but keep the signals separate.
Amazon Product Opportunity Explorer describes a workflow around customer needs, niches, search behaviour, competition, reviews and returns. The exact signed-in fields and marketplace access must be checked before relying on them.
The Amazon UK Best Sellers pages provide a public snapshot, not a unit-sales report. Google Trends provides relative, normalised interest rather than absolute Amazon demand.
| Check | Evidence or input | Status | What it does not prove |
|---|---|---|---|
| Customer need is named | Niche, use case or problem statement | That the need is underserved for your exact product | |
| Amazon signal is recorded | Tool name, store, view, date and relevant metric | That the tool guarantees demand or success | |
| Public marketplace signal is dated | Category, product, store and observation date | Units sold or future demand | |
| Trend comparison is reproducible | Term or topic, geography, range and search type | Amazon conversion or sales volume | |
| Seasonality is checked | Peaks, troughs, events and production lead time | That a peak will repeat | |
| Evidence type is labelled | Actual, relative, estimated or assumed | That different evidence types can be added as if equal | |
| Demand question is written | What would change your decision? | That collecting another number will answer it |
Demand notes to record
Write a short paragraph below the table:
The strongest signal is [source]. It supports [narrow claim]. It does not establish [limitation]. The next evidence action is [action].
This small discipline stops a research note becoming a page of disconnected numbers.
5. 3. Competition and offer structure
Competition research should describe the shelf, not just count reviews.
| Check | Evidence or input | Status | What it does not prove |
|---|---|---|---|
| Leading brands are recorded | Brand, product and store | That a new offer cannot compete | |
| Visible offer structure is recorded | Seller count, variations, condition and fulfilment | That the visible count is complete or stable | |
| Price range is recorded | Low, typical and high price with pack-size context | That the lowest price is sustainable | |
| Price changes are noted | Observation dates and promotion context | That a single price is the market norm | |
| Review themes are grouped | Repeated complaints and desired features | That one complaint represents all customers | |
| Listing quality is assessed | Images, measurements, instructions and comparison clarity | That improving content guarantees conversion | |
| Brand concentration is considered | How many brands own the visible demand? | That a generic product has no opportunity | |
| Product and offer questions are separated | Niche competition versus seller-level competition | That a product-level gap guarantees an offer-level gap |
Do not invent a review threshold. A mature listing with many reviews may be hard to displace, but a low-review listing may simply indicate low demand or poor product quality. Record the observation and explain the interpretation.
6. 4. Supply path and supplier evidence
The goal is not to find a supplier name. It is to identify a credible route to the exact product.
| Check | Evidence or input | Status | What it does not prove |
|---|---|---|---|
| Supplier type is named | Manufacturer, brand distributor, wholesaler, agent or other | That the business actually plays that role | |
| Discovery channel is recorded | Brand site, trade fair, association, directory or referral | That the channel vets every supplier | |
| Legal entity is recorded | Legal name, registration location and trading name | Reliability, product quality or brand authorisation | |
| Contact authority is considered | Independently sourced contact and explanation of role | That the contact can change the product or payment terms | |
| Exact product is traceable | Catalogue, model, identifiers, specification and pack | That the quote, goods and intended listing will agree | |
| Sample question is written | What will be measured or compared? | That a sample proves production consistency | |
| Commercial evidence is specified | Quote, terms, invoice path, payment beneficiary and lead time | That a supplier-issued document is independently true | |
| Authorisation or compliance questions are listed | Product-specific evidence needed before purchase | That a general certificate covers every market or version |
Use the European supplier discovery guide to build a longlist and the supplier verification guide to investigate a serious candidate. Use the invoice and brand-authorisation guide when the product is branded.
7. 5. Economics before enthusiasm
Do not ask whether a product has a good margin until you have defined the cost basis.
| Check | Evidence or input | Status | What it does not prove |
|---|---|---|---|
| Sale-price basis is clear | Customer price, discount and VAT treatment | That the price will remain available | |
| Supplier cost basis is clear | Unit, pack, case, currency and validity date | That the quote covers every landed cost | |
| Marketplace fees are checked | Store, category, product inputs, route and date | That a remembered percentage is current | |
| Fulfilment route is named | FBA, merchant fulfilled or another route | That the route is eligible for this product | |
| Landed cost is built | Freight, duty, import charges, prep and inbound inputs | That tax treatment is correct without professional review | |
| Returns are allowed for | Rate, average cost and unsellable-stock treatment | That a borrowed rate matches this product | |
| Advertising is explicit | Launch allowance, percentage, cost per order or separate test budget | That the campaign will deliver the assumed sales | |
| Downside case is shown | Price, cost, returns or demand change | That the base case is a forecast |
Use the Amazon UK fee and revenue estimator with the actual product dimensions, weight, category, price, store and fulfilment method. Keep the input record. Link the detailed calculation to the landed-cost guide, Amazon contribution-margin guide and wholesale profitability guide.
8. 6. Cash and operational exposure
Per-unit contribution does not tell you how much cash the first order consumes.
| Check | Evidence or input | Status | What it does not prove |
|---|---|---|---|
| MOQ is recorded | Units or cases and any tiered pricing | That the MOQ is commercially sensible | |
| Case pack is recorded | Units per carton and order multiples | That the supplier will split cases | |
| Payment timing is recorded | Deposit, balance, credit and payment date | That a supplier will offer better terms | |
| Full lead time is recorded | Production, inspection, freight, customs, receiving and availability | That the date is guaranteed | |
| Stock cover is modelled | Weeks of stock at the working demand assumption | That the demand assumption is right | |
| Reorder exposure is visible | Reorder date, cash and lead-time sensitivity | That the next order will be available | |
| Downside cash case is run | Slower sales or higher costs | That stock can be exited at the modelled price | |
| Exit route is considered | Return, liquidation, other channel or controlled write-off | That a secondary channel will absorb stock |
If a candidate requires more cash than the business can safely commit, mark it blocked even if the per-unit model looks positive.
9. 7. Product risk and stop conditions
The spreadsheet should not hide the risk screen.
| Check | Evidence or input | Status | What it does not prove |
|---|---|---|---|
| Category and product approval | Current Amazon path for the target store | Approval is guaranteed or permanent | |
| Product compliance scope | Market, product type, labels, records and responsible parties | A general certificate covers every version | |
| Brand and IP position | Brand, design, packaging and claims questions | A supplier's statement settles the legal question | |
| Hazmat or restricted attributes | Batteries, chemicals, cosmetics, food or other relevant features | A normal fulfilment route is available | |
| Size and handling exposure | Dimensions, weight, breakage, leakage and storage | Every unit will arrive sellable | |
| Seasonal and date risk | Demand cycle, expiry, event and production timing | A seasonal peak will recur | |
| Returns and support burden | Likely failure modes and customer-service steps | A low return assumption is safe |
Use the product risk screen as the early stop gate. The Amazon restricted-products guide and UK and EU compliance guide own their detailed topics.
Write one sentence under the table:
Stop before [action] if [evidence or condition] is still unresolved.
For example: “Stop before paying a deposit if the supplier cannot reconcile the exact model, pack count and invoice entity.”
10. 8. The decision record
Finish the checklist with a decision. Choose one:
- research another source;
- request a sample;
- request a comparable quotation;
- revise the specification;
- run a small, controlled test;
- pause until a dependency is resolved; or
- reject the candidate.
Record:
| Decision field | Entry |
|---|---|
| Candidate and version | |
| Decision owner | |
| Decision date | |
| Strongest evidence | |
| Most important unknown | |
| Hard stop | |
| Next action | |
| Evidence required to change the decision | |
| Next review date |
“Promising” is not a decision. “Request a sample after the supplier confirms the material and case pack” is a decision.
11. Copyable on-page checklist
Use this table as the complete HTML version. A later spreadsheet or PDF can mirror it, but the page must remain useful without a download.
| Check | Evidence or input | Source or file | Status | Checked on | What it does not prove | Next action |
|---|---|---|---|---|---|---|
| Exact candidate identity | ||||||
| Customer need | ||||||
| Amazon demand signal | ||||||
| Public marketplace observation | ||||||
| Google Trends comparison | ||||||
| Seasonality | ||||||
| Competitive brands and offers | ||||||
| Price spread and changes | ||||||
| Review themes | ||||||
| Supplier type and route | ||||||
| Legal entity and contact | ||||||
| Exact product and identifier | ||||||
| Sample or specification | ||||||
| Quote and payment terms | ||||||
| Amazon fee input | ||||||
| Landed cost | ||||||
| Returns and advertising | ||||||
| MOQ and case pack | ||||||
| Lead time and cash cycle | ||||||
| Restrictions and compliance | ||||||
| Fulfilment and damage risk | ||||||
| Stop condition | ||||||
| Decision and next action |
12. What this checklist cannot do
It cannot guarantee demand, Amazon approval, a Featured Offer, supplier quality, compliance, a sale price or profit.
It can make the decision traceable. That is the point.
When you have completed the record, use the opportunity scorecard to compare it with other candidates. Then move only the candidates with a clear next action into supplier research, sampling and detailed cost work.
13. Sources and review note
This checklist uses Amazon's public product-research and fee guidance, Amazon UK public marketplace pages and Google's official Trends methodology. Amazon tools, fees and programmes can change by store, account and product. Recheck volatile sources before acting on a live candidate.
This is an educational buying control, not a promise of sales, approval, ranking or profit.