The recurring ones are structural. Building a brand around a product you personally liked, differentiating by invention instead of by documented complaints, funding a failure because stopping feels like defeat, and never writing down the conditions under which you would stop. Score yourself on all four before the next order.
The short version
- Product selection is the mistake that cannot be fixed downstream. Everything after it inherits the error.
- Differentiation is research, not creativity. The complaints in competitor reviews are your specification.
- The absence of a stop rule is itself the mistake, and it is the one that empties bank accounts.
- Brand building and product building are different jobs. Most sellers do the second and call it the first.
- Score the build, do not describe it. A number forces a decision that a paragraph lets you avoid.
The three months I spent proving this myself
Early on I had a product that was not working. The rating was drifting down, conversion was flat, and the cost of acquiring a customer kept climbing. I put more money into advertising for three months, because the alternative meant admitting the product was wrong, and I told myself the ads would turn it around. They did not. The money went, the stock stayed, and the decision I should have made in week three got made in month four anyway.
Everything below comes out of that, particularly the scoring row nobody enjoys filling in.
Score your brand build
Rate each row from 0 to 3, where 0 means not done, 1 means loosely done, 2 means done properly, and 3 means done with evidence you could show a skeptic. Multiply by the weight.
| Row | Weight | What full marks looks like |
|---|---|---|
| Market sized before product chosen | 5 | Written analysis of size, growth, returns and rating gap, dated before the product decision |
| Differentiation traced to complaints | 5 | Specific one and two star review themes mapped to specific product changes |
| Unit economics after all costs | 4 | Margin modeled with fees, freight, returns and acquisition cost at three price points |
| Validation order sized to learn | 4 | Small first order, defined questions it will answer |
| Stop rule written before launch | 4 | Triggers, thresholds and a review date agreed in writing |
| Brand assets beyond one product | 3 | A registered mark, a positioning line and a second product that makes sense beside the first |
| Conversion path tested | 3 | Primary image compared on data, listing answering the top three objections |
| Traffic plan named | 2 | Named demand sources with owners, not just an advertising budget |
Maximum is 90. Above 70, keep going. Between 50 and 70, fix the lowest weighted rows before the next purchase order. Below 50, you have a product purchase, not a brand build, and the next order is the most dangerous decision on your calendar.
The stop rule, in practice
The row people skip is the last one they need. Ours is a scale, fix or stop decision reviewed against four signals over a defined window, agreed before launch rather than during the argument.
| Signal | What you watch | What it usually means |
|---|---|---|
| Rating trend | Direction over the last fifty reviews, not the average | Product or expectation problem. Fixing ads does nothing |
| Return rate | Against your model and the category norm | Quality, sizing or listing accuracy |
| Conversion rate | After the listing has been properly rebuilt once | Price, offer or product fit |
| Acquisition cost trajectory | Direction over weeks, not a single reading | Competitive pressure or weak conversion |
The point of writing it down early is that the version you write while optimistic is the fair one. Every version written mid crisis is generous to the decision you already want to make.
What most agencies will not tell you
Nobody is paid to tell you to stop. A monthly fee continues while a product limps, and the conversation that ends it is the conversation that ends the invoice. That is not a character flaw in the industry, it is arithmetic, and you should assume it applies to us as much as to anyone else. The defense is a stop rule agreed in advance, in writing, so the decision is a calculation rather than a negotiation.
The second thing: most private label brands are not brands. They are a collection of unrelated products behind one logo, which is fine for cash flow and terrible for valuation. If you intend to sell the business one day, the second product has to make sense next to the first, and that constraint should shape your research from the beginning.
Related answers
- Amazon product launch mistakes to avoid
- Top categories for private label beginners on Amazon
- How to find profitable niches on Amazon without PPC burnout
- Questions to ask before hiring an Amazon agency
- Done-for-you Amazon management: the complete guide
Score your build, then send us the two lowest rows for a free written second opinion at Flapen.
