Amazon tracks dozens of seller performance metrics, but only a handful directly determine whether you keep your selling privileges, win the Buy Box, or get buried in search results. This guide breaks down which numbers actually matter, how they interconnect, and what thresholds you need to maintain.

The three-tier hierarchy of Amazon seller metrics

Amazon's metrics system works like a pyramid. Foundation metrics at the bottom determine whether you can sell at all. Middle-tier metrics affect your account health rating and customer trust signals. Top-tier metrics influence Buy Box eligibility and search visibility.

Most sellers monitor everything equally, which leads to wasted effort on vanity metrics while missing early warning signs in critical areas. Understanding this hierarchy helps you prioritize where to focus your monitoring and improvement efforts.

Foundation tier: The metrics that get you suspended

Three metrics directly trigger account-level actions when they cross Amazon's thresholds. These are non-negotiable—fall below standard on any of them and you risk suspension regardless of how well you perform everywhere else.

Order Defect Rate (ODR) must stay below 1%. This metric combines three customer complaint types: negative feedback, A-to-Z Guarantee claims, and service chargebacks. Amazon calculates ODR over a 60-day rolling window.

What makes ODR particularly dangerous is that a single high-value order with a defect can spike your rate significantly if your order volume is low. A seller with 50 orders in 60 days needs zero defects to stay comfortably under 1%. One defect puts them at 2%, triggering immediate account review.

Cancellation rate threshold is 2.5%. Amazon counts any order you cancel after the customer places it, whether the cancellation was your decision or the customer's request. The exception: if the customer initiates cancellation through their account before you ship, it doesn't count against you.

The common trap here is inventory sync failures. Your listing shows in-stock, customer orders, then you discover you're out of stock and cancel. Amazon treats this as seller-caused cancellation regardless of the reason. This is why inventory management systems that sync in near-real-time matter for multi-channel sellers.

Late shipment rate must stay below 4%. Amazon starts the clock when the customer places the order and stops it when you mark the item shipped with a valid tracking number. For FBA sellers, this metric is largely automated—Amazon handles fulfillment and rarely misses shipment deadlines.

FBM sellers face more complexity. Amazon expects shipment within the handling time you set in your shipping settings. If you set 2-day handling time, you must ship within 2 days of order placement. Missing this window counts as late even if the package arrives on time to the customer.

How these foundation metrics interconnect

The three foundation metrics feed into your Account Health Rating, but they're not weighted equally in Amazon's evaluation system. ODR carries the most weight because it directly measures customer dissatisfaction.

Here's how the cascade works in practice: High ODR triggers account review. During review, Amazon examines your cancellation and late shipment rates as secondary factors. If all three metrics show problems, you move from warning to suspension much faster than if only one metric is elevated.

Amazon also looks at the trend direction. A seller at 0.9% ODR trending upward faces more scrutiny than a seller at 0.95% ODR trending downward. This is why weekly monitoring matters more than monthly reviews—you need to catch degradation early enough to investigate root causes before crossing thresholds.

The defect investigation window

When your ODR approaches 1%, you have roughly 2-3 weeks to identify and fix the root cause before it becomes a suspension risk. Most defects trace back to one of four issues:

  • Product condition mismatches (item condition doesn't match listing description)
  • Shipping damage (inadequate packaging for FBM sellers)
  • Late delivery creating customer frustration that manifests as negative feedback
  • Listing accuracy problems (images or descriptions don't match actual product)

The investigation priority is finding which products generate defects. If 80% of your defects come from one ASIN, that's a product problem not a process problem. If defects are evenly distributed, you have a systemic fulfillment or communication issue.

Middle tier: Account health and customer experience metrics

These metrics don't directly trigger suspension at specific thresholds, but they heavily influence your account health dashboard rating and how Amazon treats you in borderline situations.

Valid Tracking Rate (VTR) measures what percentage of your shipments have tracking information that confirms delivery. Amazon expects 95%+ VTR for most categories. Fall below this and you lose eligibility for certain shipping options and may see Buy Box suppression.

For FBM sellers, VTR is often the forgotten metric. Using shipping methods without tracking—or using tracking numbers that don't update properly in Amazon's system—degrades this score. The fix is using Amazon Buy Shipping or integrated carriers that automatically feed tracking data into Seller Central.

On-Time Delivery Rate tracks whether packages arrive by the delivery date you promised at checkout. This differs from late shipment rate—you can ship on time but still have late delivery if you chose an inadequate shipping speed.

Amazon calculates promised delivery date based on your handling time plus the carrier's estimated transit time. If you set 2-day handling and choose standard shipping with 5-7 day transit, your promised delivery is 7-9 days out. The package must arrive within that window.

Customer response time measures how quickly you respond to buyer messages. Amazon expects responses within 24 hours. This metric doesn't have a published threshold, but patterns of slow response correlate with higher ODR—customers who feel ignored are more likely to open claims.

The practical challenge is weekend coverage. A message received Friday evening that you don't see until Monday morning is already outside the 24-hour window. Sellers with consistent weekend sales volume typically need monitoring tools that alert them to new messages or vacation mode during uncovered periods.

Top tier: Buy Box and search visibility metrics

These metrics determine your competitiveness against other sellers, but won't get you suspended. They directly affect how much you sell.

Buy Box percentage shows how often your offer wins the Buy Box when customers view a product where you're competing with other sellers. Amazon doesn't publish the algorithm, but account health, price competitiveness, fulfillment method, and shipping speed all factor in.

The hierarchy matters here: sellers with poor foundation metrics rarely win the Buy Box even with the lowest price. Amazon prioritizes account health over price optimization. A seller at 0.8% ODR with FBA will usually beat a seller at 0.5% ODR using FBM, but a seller at 1.2% ODR gets suppressed regardless of price or fulfillment method.

Customer feedback score is your overall rating percentage. Amazon calculates this from your last 12 months of feedback. The practical threshold is 95%+. Sellers below 90% face significant Buy Box disadvantages.

What many sellers miss: feedback percentage matters more than raw feedback count. A seller with 47 positive reviews and 3 negative reviews (94%) performs worse in the Buy Box than a seller with 8 positive reviews and 0 negative reviews (100%). This is why feedback removal requests matter—removing even one negative piece of feedback can shift your percentage meaningfully when your total count is low.

Inventory Performance Index (IPI) affects FBA sellers specifically. Scores above 450 maintain standard storage limits. Fall below 400 and you face inventory restrictions during high-volume periods.

IPI combines four factors: excess inventory percentage, FBA sell-through rate, stranded inventory percentage, and in-stock rate for popular items. Amazon weighs excess inventory most heavily—aged inventory that hasn't sold in 90+ days drags down your score faster than other factors.

The IPI improvement priority stack

When your IPI drops below 450, fix issues in this order:

  1. Stranded inventory (inventory Amazon can't sell due to listing problems) — creates immediate drag with zero upside
  2. Aged inventory over 365 days — incurs long-term storage fees and Amazon counts it most heavily against excess inventory calculations
  3. Low sell-through on inventory under 90 days — indicates pricing or demand estimation problems you need to address before sending more units
  4. Out-of-stock popular ASINs — affects in-stock rate but only if you've sold the item consistently in the past 30 days

The reason for this sequence: stranded and aged inventory are pure waste with no recovery value. Low sell-through and stockouts at least represent sellable inventory where you can improve performance through pricing or restock decisions.

Metrics you can safely ignore

Amazon displays numerous metrics that have minimal impact on your account standing or sales performance. These are informational but don't require active optimization:

Return rate appears in your dashboard but Amazon doesn't penalize sellers for high return rates unless the returns are defect-related (damaged, not-as-described, etc.). Categories like apparel naturally have higher return rates. What matters is whether returns generate negative feedback or A-to-Z claims, which roll into ODR.

Refund rate measures refunds as a percentage of order value. Like return rate, this is informational. Amazon cares about why you issued the refund—if it's due to defects, it shows up in ODR. If it's customer preference, it doesn't affect account health.

Contact response time distribution shows the time breakdown of your responses (under 12 hours, 12-24 hours, etc.). The 24-hour threshold is what matters for account health. Whether you respond in 2 hours versus 18 hours makes no measured difference in Amazon's system, though faster responses likely improve customer satisfaction.

Pre-fulfillment cancel rate exists as a separate metric from overall cancellation rate but doesn't have a published threshold. As long as your total cancellation rate stays below 2.5%, Amazon doesn't take action on pre-fulfillment cancels specifically.

Monitoring strategy: Daily, weekly, and monthly priorities

Effective performance tracking requires different monitoring frequencies for different metric tiers.

Daily monitoring should cover foundation metrics only: ODR, cancellation rate, late shipment rate. These are the metrics that can trigger account-level actions. Check them every morning. If any metric crosses 75% of Amazon's threshold (0.75% ODR, 1.9% cancellation, 3% late shipment), investigate immediately.

Weekly review should include middle-tier metrics: VTR, on-time delivery, response time, and any recent customer feedback. Weekly cadence catches degradation early enough to implement fixes before metrics compound into foundation-tier problems.

Monthly analysis for top-tier metrics: Buy Box percentage, IPI score, and feedback score trends. These move slowly and require longer timeframes to identify meaningful patterns. Monthly review is sufficient unless you're actively working on improvement initiatives.

Setting up effective alerts

Manual checking misses problems that develop between review sessions. Automated alerts catch issues in real-time:

  • ODR alert at 0.75% — gives you roughly 10 days to investigate before approaching the 1% suspension threshold
  • Late shipment alert at 3% — provides 25% buffer before hitting the 4% threshold
  • Cancellation alert at 2% — flags the issue before you reach 2.5% limit
  • New negative feedback alert — immediate notification allows quick response or removal request submission

Most seller analytics tools, including SageSeller's account health monitoring, support configurable threshold alerts. Set them conservatively—it's better to receive an alert when nothing is seriously wrong than to miss early warning signs of metric degradation.

What to do when metrics start degrading

Metric degradation follows predictable patterns. Early intervention prevents foundation metrics from reaching suspension levels.

When ODR rises above 0.75%, immediately audit your last 60 days of orders for any common factors among defect orders. Sort by product, shipping method, customer location, and order value. Most sellers find one of two patterns: a specific product generating repeat problems, or a process issue affecting multiple products.

If product-specific, evaluate whether the issue is fixable through better listing content, improved packaging, or product quality improvements. If not fixable, consider removing the listing—one problematic ASIN can destroy your account health.

If process-related, the most common culprits are packaging inadequacy for FBM sellers, inventory sync delays causing stockout cancellations, or listing content that sets incorrect customer expectations.

When cancellation rate climbs, audit which orders you're canceling and why. Inventory-related cancellations indicate sync problems between your inventory management system and Amazon. Customer-requested cancellations that count against you mean customers aren't using the proper cancellation flow—you may need to guide them toward self-service cancellation to avoid the metric hit.

Late shipment issues for FBM sellers almost always trace to either unrealistic handling time settings or carrier performance problems. If you're consistently missing your handling time, extend it in your shipping settings. Better to promise 3-day handling and ship in 2 days than promise 1-day handling and ship in 2 days.

The metrics that matter most by business model

FBA and FBM sellers face different metric priorities because Amazon handles different parts of the fulfillment process.

FBA sellers should focus on IPI score and stranded inventory above all else. Amazon controls late shipment and delivery metrics for FBA orders, so those foundation metrics rarely become problems. Your main risks are inventory management issues that degrade IPI and buy Box suppression from poor feedback management.

FBM sellers must prioritize late shipment rate, VTR, and on-time delivery rate. You control the entire fulfillment process, which means more can go wrong. Foundation metrics are your primary risk. Monitor them daily and maintain buffer room well above Amazon's thresholds.

Hybrid sellers using both FBA and FBM need to segment their metrics by fulfillment method. A problem in your FBM operation can drag down account-level metrics even if your FBA performance is perfect. Amazon's dashboard shows combined metrics, but you need to track each channel separately to identify where issues originate.