> For the complete documentation index, see [llms.txt](https://sourcesheets.gitbook.io/sourcesheets-new-seller-blueprint/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://sourcesheets.gitbook.io/sourcesheets-new-seller-blueprint/part-4-sourcing-and-buying-your-first-products/7.5.1-7.5.3-the-gate-based-evaluation-model.md).

# 7.5.1 The Gate-Based Evaluation Model

Before analysing individual metrics, it helps to understand how we evaluate products, not just what we look at. In this Blueprint we use a gate-based evaluation model, not a scoring system. That's intentional.

## Why We Use Gates Instead of Scoring

Many tools and guides encourage scoring products: add points for high ROI, subtract points for competition, weigh factors against each other. This sounds logical, but it creates a dangerous habit: trade-offs. Scoring encourages sellers to think "the profit makes up for the risk" or "the demand outweighs the competition." In reality, Amazon doesn't work on averages. A single hard problem can invalidate an otherwise strong product.

That's why we use gates. A gate is simple: the product either passes or it fails. There's no balancing act.

## Applying the Model to Product Analysis

Think of it like the game Guess Who. You start with many possibilities, and each question removes options. You're not trying to prove who the character is. You're trying to eliminate who it isn't.

Product analysis works the same way. Every gate asks a binary question: does the maths work, is demand stable, is competition manageable, are there hard risks present? If the answer is no, the product is removed. There's no reward for surviving more gates, only permission to move to the next one.

## Why Most Products Should Fail

This is the most important expectation to set early. If most products you analyse pass, something is wrong. Amazon is crowded, margins are thin, and risk is unevenly distributed. The purpose of structured evaluation isn't to find reasons to buy. It's to find reasons not to. What that looks like in practice, and how to act on a rejection, is covered in Phase 6: Final Decision Logic.

## What the Gate Model Protects You From

Using gates instead of scoring stops you justifying bad decisions with good numbers, ignoring hard risks because other metrics look strong, getting emotionally attached to a product, or chasing excitement instead of stability. It forces discipline: every product is treated the same, and every decision follows the same structure.

## How This Model Will Be Used Going Forward

In the sections that follow, this gate-based model is applied consistently. Each phase introduces new gates: financial reality, demand and stability, competition and Amazon behaviour, and risk flags and hard stops. A product only moves forward if it survives the current gate. There are no shortcuts. This structure is what lets sellers make calm, repeatable decisions even in a market that constantly changes.

## A Note on Experience, Judgement, and Reality

This framework is designed primarily for new and early-stage sellers. It provides structure at a point where judgement is still forming and confidence can run ahead of experience.

Product analysis isn't purely mechanical. Tools and gates create discipline, but they don't replace experience. Over time, sellers develop an instinct for products that goes beyond metrics, subtle signs of instability, patterns in seller behaviour, brand quirks, listings that technically pass checks but still feel wrong. These signals are hard to explain in advance because they're learned through repetition, exposure, and mistakes. There's no shortcut to this.

This framework isn't meant to turn you into an expert overnight. It exists to slow you down, remove emotion from early decisions, and protect you from the most common and costly beginner mistakes while that judgement develops.

As you analyse more products, you'll move through these gates faster. Certain issues will stand out immediately, and some checks will become instinctive rather than deliberate. That's not a failure of the framework. It's the outcome it's designed to support.

This model doesn't predict outcomes or guarantee profit. Amazon is dynamic, sellers behave unpredictably, and conditions change without warning. Even experienced sellers get caught out sometimes. Passing analysis simply means a product isn't obviously bad at the point you checked it, nothing more.

Used correctly, this structure gives you a safe baseline to build pattern recognition without paying unnecessary tuition in losses. Every product you analyse, including the ones you reject, contributes to that learning. The goal isn't perfection. It's consistency, survival, and progress.