How to read Keepa charts like an algorithm

If you’ve ever stared at a Keepa chart, scratching your head and wondering if an ASIN is a goldmine or a money pit, you’re not alone.

For many Amazon sellers — especially beginners in online arbitrage (OA) or wholesale — Keepa feels like a foreign language:

• Too many lines.
• Too much noise.
• And no clear answer on whether to buy or walk away.

This doesn’t have to be confusing

But here’s the thing: algorithms don’t guess. They see patterns, probabilities, and context — and they use that information to make consistent, profitable decisions.

This step-by-step Keepa tutorial will teach you how to read charts like an algorithm so you can:

• Spot consistent winners.
• Avoid profit-killing traps.
• Build a more scalable and predictable sourcing process.

And once you have strong fundamental analysis, can you use an AI product sourcing tools to analyze thousands of ASINs for you in minutes.

Why Keepa is non-negotiable for Amazon sellers#

Keepa is more than a pretty graph — it’s a record of an ASIN’s life. It tells you:

• When prices spiked or dropped.
• How stable demand has been.
• When Amazon jumped in or dropped out.
• Whether there’s a recurring seasonal cycle.

Learning to interpret these signals is the foundation of profitable sourcing. Any substantial form of analysis requires historical data.

The core signals algorithms watch#

Before you can think like an algorithm, you need to understand what the data points actually mean.

Sales rank trends#

Sales rank is a proxy for demand — but context is everything.

• Consistently low rank: Strong, predictable demand.
• Sharp dips and spikes: Seasonal or unpredictable sales.
• Flatlining rank: A dead listing, suppressed, or a niche product with very low turnover.

Algorithm mindset: Don’t just look at the lowest rank. Evaluate the average and volatility over time to gauge stability.

Buy box behavior#

The buy box is where profits are made — or lost.

• Stable buy box price: Predictable profit margins.
• Frequent price suppression: Either Amazon interference or heavy competition eroding margins.
• Frequent price handoffs: Lots of sellers competing, increasing the chance of race-to-the-bottom pricing.

Algorithm mindset: Compare historical buy box averages to current pricing. If the current price is significantly below average, the listing may not recover soon.

Stock and seller history#

Stock levels and competition patterns reveal critical opportunities (and risks).

• Amazon competition: If Amazon is present and rarely goes out of stock, it’s almost impossible to compete profitably.
• Frequent OOS (out-of-stock) events: Opportunity to step in and grab the buy box when others run out.
• Seller churn: Indicates a high-velocity listing but also a potential price war.

Historical pricing#

Past behavior predicts future performance.

• Upward trends: Demand outpacing supply — great if you can secure inventory.
• Downward slides: Category saturation or declining relevance.
• Seasonal spikes: Toys before Christmas, backpacks before school, etc.

Secondary data layers#

Advanced algorithms also analyze:

• Review velocity: Are reviews coming in steadily or slowing down?
• Variations: Are parent-child relationships (e.g. multi-packs, size variations, etc) skewing perceived rank or sales?
• Category volatility: Some niches naturally have higher churn and volatility.

Thinking in patterns, not snapshots#

One of the most common mistakes FBA sellers make is zooming in too much.

A 30-day or 90-day chart might show a price drop that looks scary — but zooming out to 180 days or a full year often reveals that it’s just a seasonal pattern.

Example:

• A toy’s price drops every January as post-holiday clearance hits.
• By June, inventory levels tighten and prices normalize.

Algorithm mindset:

• Use wider windows to spot patterns.
• Look for cycles instead of reacting to temporary noise.

A step-by-step framework for reading any Keepa chart#

Here’s how to break down a chart like an algorithm would:

Step 1: Check demand stability#

• Zoom out to at least 180 days.
• Look for consistent dips in rank that indicate recurring sales.
• Identify whether the product moves steady volume or spikes unpredictably.

Step 2: Analyze buy box behavior#

• Look at historical pricing bands.
• Compare the current buy box to the 90-day and 180-day averages.
• Flag any price suppression or pricing instability.

Step 3: Evaluate competition#

• Is Amazon a competitor?
• How many FBA sellers are competing for the buy box?
• Is there frequent seller churn (high competition turnover)?

Step 4: Spot red flags#

• Price suppression.
• Inventory flooding after arbitrage videos or group leads.
• Rank instability following category changes.

Step 5: Run the math#

• Plug in fees, shipping, and prep costs.
• Base your profit projections on average prices, not peaks.
• Include a buffer for downward swings.

Step 6: Contextualize seasonality#

• Use Keepa’s yearly chart to identify predictable peaks.
• Match sourcing to cycles (e.g., buying seasonal items months ahead at low competition).

Common mistakes sellers make (and how to avoid them)#

Algorithms don’t get emotional — but sellers often do. Here are pitfalls to avoid:

Mistake	Algorithm approach
Judging by rank alone	Combine rank with price and competition data
Buying during spikes	Check historical averages for realistic pricing
Ignoring Amazon presence	Avoid competing unless the listing shows frequent OOS events
Overlooking seasonality	Use yearly charts to identify cycles
Focusing only on “fast flips”	Build a mix of steady, reliable sellers

Case studies: Human guesswork vs algorithm logic#

Example 1: The “hot toy”#

• Human: Sees low rank and high price in November, buys deep.
• Algorithm: Notes seasonal spike, predicts post-Christmas crash, avoids overbuying.

Example 2: Grocery listing#

• Human: Skips because price dipped last month.
• Algorithm: Notes the dip was from a temporary inventory flood; buy box recovered. Buys confidently.

Example 3: Mid-tier electronics#

• Human: Excited about low rank but ignores Amazon’s constant presence.
• Algorithm: Identifies low probability of winning the buy box and passes.

Automating the hard work with Axon#

Here’s the reality:

• You don’t have time to manually analyze hundreds of ASINs every week.
• Even if you did, human error creeps in — emotion, fatigue, or bias.

That’s why I’m building Axon, an AI-powered sourcing assistant that:

• Reads Keepa charts following the rules above 👆.
• Analyzes thousands of ASINs in minutes.
• Scores leads based on your unique risk profile and real-time market data.
• Flags profit-killers like price suppression or Amazon dominance.
• Highlights opportunities you might otherwise miss.

Imagine turning an hour of manual analysis into five minutes — and doing it with objective, consistent logic every time.

Pro tips for becoming “algorithm-like” even without automation#

While Axon can handle the heavy lifting, here’s how to start thinking algorithmically on your own:

• Create a checklist for every ASIN you analyze.
• Track your decisions and revisit them in 30/60 days to see where your judgment was off.
• Focus on averages, not outliers.
• Practice pattern recognition by looking at dozens of charts daily.
• Keep emotion out of the equation — data first, gut second.

Key takeaways#

• Zoom out: Wider time windows reveal patterns, not noise.
• Combine signals: Rank, price, competition, and history must be analyzed together.
• Think probabilities: Treat every buy as a data-backed bet.

Reading Keepa charts like an algorithm isn’t just about getting good at graphs — it’s about building a repeatable, scalable decision-making process.

Start practicing these steps today, and when you’re ready to take it to the next level, let Axon handle the heavy lifting so you can focus on growing your Amazon business.

👉 Join the waitlist for early access

Want more tips like this? Follow along as I build Axon in public, right here on the blog.