The AI Guide analyzes any product you're researching and gives you a clear verdict on whether it's worth selling: Recommended, Consider, or Not Recommended. Each analysis includes a confidence level, plain-language reasoning, key pros and cons, and a suggested test quantity for your first order.
The AI Guide is built for sellers in their early stage. It weighs the factors that matter most when you're placing small test orders: real profitability at current market prices, demand strength, buy box competition, price stability, and brand or IP restrictions.
Note: AI analysis is for guidance only and may not always be accurate. Always apply your own judgment before purchasing inventory.
Prerequisites
- An active ProductChimp account
- A product open in the analysis view
- Recommended: your cost price for the product. The AI Guide can run without one, but profitability results are far more reliable when you provide your actual cost. Your saved Settings (sale price, discount, fulfillment costs, desired profit and ROI targets) are also factored in automatically.
How to Run an Analysis
- Open the product you want to evaluate.
- Click the AI Guide button.
A panel opens confirming the ASIN it's about to analyze: - The AI Guide shows the cost price it will use — pulled from what you've entered for this product. Check that it's correct before continuing.
- Click Analyze Now. The analysis usually completes within a few seconds.
- Review your results:
- Verdict — Recommended, Consider, or Not Recommended, with a one-line summary.
- Confidence level — Low, Medium, or High, indicating how clear-cut the data is.
- Reasoning — a short explanation of the deciding factors, with specific numbers.
- Pros and cons — the strengths and watch-outs the AI found in the data.
-
Suggested test units — a conservative first-order quantity (typically 5–10 units) when the verdict is Recommended or Consider.
- To re-run the analysis — for example, after updating your cost price or Settings — click Analyze Again:
Understanding the Verdicts
Recommended — The core sourcing case holds: you can buy in, sell through, and make an acceptable profit. A Recommended product will still usually have cons listed; these are things to watch, not reasons the AI held back.
Consider — The product could work, but something genuinely threatens the profit case — most often a thin or borderline margin. The decision hinges on your actual cost and your own judgment.
Not Recommended — Either the numbers don't work (for example, the market price floor is at or below your break-even cost), or a hard blocker exists — such as an Amazon brand restriction, Amazon itself dominating the buy box, or demand too low to sell through even a small test order.
What the AI Guide Analyzes
- Profitability — Your cost price against the maximum viable cost, ROI at the current sale price, and how live offers are priced relative to your cost. If you've applied a discount, the discounted cost is used as the basis for all margin calculations.
- Demand and sales velocity — Monthly sales volume and BSR, including whether demand is growing or declining over 30, 90, and 180 days.
- Competition and buy box rotation — How the buy box is shared among sellers, the FBA/FBM split, and whether Amazon itself is on the listing. Healthy rotation across multiple sellers is a positive; one seller dominating is a red flag.
- Price stability — Price trends and churn, including price erosion and buy box suppression events.
- IP and brand risk — Amazon brand restrictions and internally flagged brands, plus cases where brand approval is required.
- Your criteria — Where available, results are compared against your desired profit, desired ROI, and BSR range from Settings.
Why the Same Product Can Get Different Verdicts
The verdict depends heavily on your numbers — especially your cost price. The same ASIN can be Not Recommended at a $5.00 cost (if the break-even cost is, say, $3.48) and Recommended at a $3.00 cost with a healthy ROI. Market data is only half the picture; the AI Guide is evaluating whether this deal at your cost makes sense.
If your verdict looks off, check:
- The cost price shown before you clicked Analyze Now
- Any discount you've applied to the cost
- Your sale price, fulfillment, and storage assumptions in Settings
Then click Analyze Again.
Troubleshooting
The verdict is Not Recommended, but the pros look great. Strong demand and clean competition can't overcome a deal that loses money. If the reasoning points to profitability, the market price is too low relative to your cost. Try to source cheaper, then re-run the analysis.
My profit or ROI numbers look wrong. Confirm your cost price was entered for this product, and review your Settings — sale price, discount, FBM cost, and storage assumptions all feed the calculation. Without a cost price, the AI Guide focuses on the maximum viable cost instead of exact profit figures.
A con was listed, but the product was still Recommended. That's by design. Cons are informational — they tell you what to monitor (a brand approval step, mild price erosion, a concentrated buy box). The verdict reflects whether the overall deal works, not a tally of cons.
The analysis mentions "buy box suppression." This means Amazon temporarily hid the buy box because it judged the listing price too high. It slows sell-through even if sales are still possible, which is why it's flagged as a notable risk.
FAQs
Does the AI Guide use my Settings presets? Yes. Custom sale price, discount, fulfillment costs, storage assumptions, and your desired profit/ROI targets are all reflected in the analysis when configured.
What does the confidence level mean? It reflects how clear-cut the data is. High confidence means the signals point strongly in one direction; Low confidence means the data is mixed or limited, and your own judgment matters more.
How many units should I buy if it's Recommended? The AI Guide suggests a conservative test quantity, typically 5–10 units, adjusted down when competition is heavy or prices are volatile. It's a starting point for a first test order, not a scaling plan.
Can I rely on the verdict alone? No. The AI Guide is a decision aid — it surfaces the numbers and risks quickly, but it can miss context and may not always be accurate. Treat it as one input alongside your own research.
Is there a limit on how many analyses I can run? Analysis availability depends on your plan. See your plan details on the Billing page.