A brand launch runs about seven months and breaks in the same five places every time. Plan it as gated phases: size the market, validate with 200 units, prove rating, conversion, and acquisition cost, then scale. Write the kill criteria before you spend, because you will not write them afterwards.
The short version
- Gates, not milestones. A milestone is something you celebrate. A gate is something that can stop you.
- Validation before inventory. 200 units and $5,000 to $10,000 buys the answer far more cheaply than a container does.
- Four products can be tested at once, which spreads the risk that any single one fails.
- The kill criteria are written on day one, in advance, while you are still capable of being objective.
- Expect four to six hours a week from yourself during the active phase, whoever else is working on it.
You are further along than the plan assumes
Most people asking for a launch plan already have something: a product idea, a supplier quote, sometimes stock in a garage. The plan below is written to be entered at any point, and the honest advice is to start at step one anyway. The steps you skipped are the ones that will surface later, at the worst possible moment, as a problem you cannot fix with advertising.
The plan
- Size the segment. Name the specific competitive set your product will appear against and establish whether the market is large enough to be worth entering after acquisition cost. Gate: a number you would defend in front of an investor.
- Find the gap. Read the negative reviews of the leading listings and identify the repeated, fixable complaint. Differentiation comes from what customers already dislike, never from invention. Gate: one specific defect you can fix in manufacturing.
- Select and sample the supplier. Quote several factories at once, write a specification that removes ambiguity, and expect more than one sample round. Gate: a sample you would sell under your own name.
- Run the legal track in parallel. Trademark filing, Brand Registry, barcodes, category approvals. Started at step one, not at step six.
- Build the listing and creative from the sample. Keyword architecture, main image, secondary images that answer the objections you found at step two, A+ content. Gate: a page a stranger understands in five seconds.
- Phase one, validation. 200 units, $5,000 to $10,000, up to four products in parallel if you have the budget. The purpose is information, not profit. Gate: rating holding, conversion rate acceptable, acquisition cost trending down.
- Phase two, scale. Only now do you commit real inventory and real budget. Reorder before you run out, because going out of stock at this point destroys ranking you paid for.
- Review against the criteria at each window. Scale what is working, fix what is fixable, kill what is neither.
Where it breaks, ranked by cost
| Failure mode | What it costs | The gate that prevents it |
|---|---|---|
| Entering a market too small to support acquisition cost | The entire launch budget, slowly | Step one, sizing |
| Skipping validation and ordering a full container | Capital locked in unsellable stock | Step six, phase one |
| No written kill criteria | Months of spend defending a decision | Step eight, decided in advance |
| Vague supplier specification | Extra sample rounds, delayed launch, quality variance | Step three, the spec |
| Listing built before the gap is understood | Traffic that does not convert, at full price | Step two, then step five |
| Trademark started late | Launch delayed while stock sits | Step four, in parallel |
| Running out of stock mid-launch | Lost ranking, paid for twice | Step seven, reorder timing |
The one that cost me the most
Early on I poured money into a failing product for three months, convinced that better advertising would turn it around. The rating was slipping, returns were higher than the category, and every week I had a reason why next week would be different. The ads did not turn it around. The product was wrong, and every dollar after the first month was paying to postpone an obvious conclusion.
That is where our scale, fix, or kill criteria come from. Before spend begins, we write down what would make us stop: rating trend, return rate, conversion rate, and acquisition cost trajectory, each measured over a defined window. The criteria exist because I have proven, personally, that they cannot be written honestly in the middle of a launch you are emotionally attached to.
When you evaluate any provider, this is the question I would ask: what would make you tell me to stop selling this product? Anyone who cannot answer specifically is either inexperienced or paid to keep going.
What most agencies will not tell you
A launch plan sold as a document is worth very little. The value is entirely in whether someone enforces the gates when the news is bad, and the gates are inconvenient by design. Every one of them is a place where the honest answer might be to stop, and stopping is the outcome nobody in the room is paid for.
The second thing is timing expectations. Seven months for a full brand is normal. Sellers who are told six weeks are being sold a listing, not a brand, and the difference shows up around month four when the product has a page, a small pile of reviews, and no ranking.
Related answers
- Step-by-step Amazon product launch checklist
- Amazon product launch mistakes to avoid
- How long does an Amazon launch phase take
- Checklist for Amazon Brand Registry and trademarks
- Done-for-you Amazon management: the complete guide
The gates we hold every brand to, including the stopping conditions, are set out at Flapen.
