The Complete Amazon Repricing Guide

What Amazon repricing actually is

An Amazon repricer is software that automatically adjusts your product prices in response to competitor moves, Buy Box ownership shifts, and rules you configure. The motivation is simple: at any meaningful catalog size, manual repricing is impossible. Competitors change prices hundreds of times per day on contested ASINs. The Buy Box rotates every few minutes. Any seller still typing prices into Seller Central is leaving money on the table.

Repricers come in three rough generations:

Generation 1: Match-the-lowest scripts. These are worse than nothing. They collapse the Buy Box for everyone, race margin to the floor, and are responsible for the well-documented pricing storm incidents that periodically wreck contested categories. Avoid.

Generation 2: Rule-based with hard margin floors. Acceptable for stable catalogs. You set a min and max price per SKU and the repricer plays inside the box. Better than nothing, but lacks the strategic awareness that wins the Buy Box at HIGHER prices than competitors.

Generation 3: Strategic, Buy-Box-aware, TACoS-aware repricers with safety guardrails. This is the only generation worth using on a serious business. The rest of this guide explains what makes a Generation 3 repricer different.

Hard margin floors at the database level

The most important feature in any serious repricer is the hard margin floor. The repricer must not be able to submit a price below the floor — not because of a config bug, not because of a malformed cost record, not because of an upstream API hiccup. The constraint should be enforced at the database level so that even a runaway loop can not push a price below the floor.

Floor modes that should be supported:

• Fixed dollar. "Never go below $14.99." Simple, useful for legacy SKUs with stable economics.
• ROI %. "Never go below the price that produces 30% ROI." Ties floor to landed cost, recalculates automatically when COGS changes.
• Margin %. "Never go below the price that produces 25% margin." Ties to revenue not cost.
• Breakeven. "Never go below total fees + COGS." Useful for clearance modes where you want the repricer to drop hard but not produce a loss.
• % above COGS. "Never go below COGS + 15%." Hybrid between ROI and breakeven.

The breakeven mode is where the integration gets subtle. Which fees count? Referral fee yes, FBA fulfillment fee yes, but what about variable storage? PPC allocation? Inbound costs? The right repricer lets you toggle each fee component on or off so the floor reflects YOUR true breakeven, not an industry average.

Buy Box probability modeling

Generation 2 repricers reprice toward a price target. Generation 3 repricers reprice toward a Buy Box probability target. The distinction matters enormously.

When you reprice toward "match the Buy Box price minus 1 cent", you are assuming you will win the Buy Box if you have the lowest price. That is wrong. Amazon's Buy Box algorithm weights fulfillment channel, seller feedback, in-stock status, shipping speed, and historical performance — price is one factor among several. A Buy Box-aware repricer models your probability of winning the Buy Box at each price point, including those non-price factors, and chooses the price that maximizes expected revenue per Buy Box win.

A practical example: you and a competitor are both FBA on the same ASIN. Competitor has 4.3 stars, you have 4.8. The Buy Box-aware repricer recognizes that you will hold the Buy Box ~60% of the time at price parity, vs ~40% for the competitor. So instead of dropping to undercut, the repricer holds at parity (or slightly above) and lets the algorithm rotate to your favor. This wins more revenue at higher margin than dumb undercutting would.

TACoS-aware pricing (the part nobody else does)

Most repricers treat advertising and pricing as separate problems. They are not. A SKU with high TACoS is one where every dollar of revenue is being subsidized by ad spend. Dropping the price on a high-TACoS SKU to win more Buy Box is a double loss: less margin per unit AND continuing high ad spend per dollar of revenue.

A TACoS-aware repricer pulls live advertising data from your Amazon Ads account and factors it into pricing decisions. Specifically: when a SKU is carrying high TACoS, the repricer protects margin harder — it raises the floor, restricts how aggressively it will undercut, and prefers losing some Buy Box share over losing margin. When a SKU has low TACoS and healthy organic, the repricer can be more aggressive on price because the underlying business is stronger.

This is genuinely a feature SellerVault has and most other repricers do not. Bqool, Informed.co, Repricer.com, and Seller Snap are all repricer-only tools without integrated PPC data. The result is that they cannot make TACoS-aware decisions at all.

Anti-oscillation guards and circuit breakers

One of the failure modes of naive repricers is oscillation: prices zig-zag in narrow bands as multiple repricers chase each other. A typical oscillation cycle: you drop $0.10 to undercut, a competitor drops $0.10, you drop another $0.10, they drop again. Within an hour the price has fallen $5 with no actual buyer activity driving it. Margin has been silently destroyed across the catalog.

Anti-oscillation guards detect and stop these loops. The standard implementations:

• Narrow-band detector. If prices oscillate within a small band (say, 5%) more than N times per hour, lock the SKU at the current price for a cooling-off period.
• Velocity circuit breaker. If submission rate spikes above 10x the SKU's baseline, halt submissions and audit the cause before resuming.
• Floor enforcement. Hard margin floor (covered above) acts as an absolute brake.
• Storm detection audit. When multiple SKUs enter narrow-band oscillation simultaneously, fire an alert so a human can investigate.

These guardrails were added to SellerVault's repricer after surviving a real pricing storm in April 2026. The storm started with a competitor running a buggy algorithmic repricer on inactive listings; within 90 minutes the resulting cascade had moved hundreds of competing SKUs into low-margin territory. Most repricers had no defense against this. The post-incident review produced the four guardrails listed above and they have been in production since.

Audit trail with reason codes

Every price change should be logged with a reason code, timestamp, trigger event, and the rule that fired. When something looks wrong — and at scale, something always looks wrong — you should be able to trace the decision path in seconds.

Specific data the audit trail should capture:

• The SKU, the old price, the new price, the change delta
• The triggering event (competitor moved, schedule fired, rule re-evaluation, manual override)
• The rule that produced the decision (named, versioned)
• Buy Box probability before and after
• The repricer's confidence score for the decision
• Amazon's submission response (accepted, rejected, queued)
• TACoS context at the time of the decision

Without this, debugging "why did this SKU get priced at $9.99 yesterday afternoon" turns into a multi-hour archaeology project. With it, you click the SKU, click the timeline, and see the answer in 5 seconds.

Configuring your first rules

When setting up a repricer for the first time, the temptation is to enable it on the full catalog and let it figure things out. Resist. The right starting playbook:

Step 1: Enable on a subset. Pick 20-50 SKUs that span your catalog's range — fast movers, slow movers, contested ASINs, uncontested, premium and discount. Enable repricing on just this subset for the first 1-2 weeks.

Step 2: Set conservative floors. Use ROI % or margin % floors set 5-10 percentage points HIGHER than your eventual target. This gives you safety margin to discover whether the repricer's decisions match your expectations.

Step 3: Watch the audit trail. Every day for the first week, scroll through the audit trail and verify the decisions make sense. Look for surprises — anything that pricier or cheaper than you expected, any oscillation, any rejected submissions.

Step 4: Loosen floors. Once you trust the repricer's decisions on the subset, lower the floors to your true target floors and expand the SKU set in waves of 100-200 at a time.

Step 5: Enable strategic modes. Once base repricing is solid, enable strategic modes (TACoS-aware, schedule-based, velocity-aware) one at a time and verify each one's behavior on a subset before going catalog-wide.

This staged rollout is the same approach you would take to deploying any other automation that touches money. Skipping it produces bad surprises.

Metrics to watch after deployment

Once the repricer is running, the four metrics to watch closely:

1. Buy Box win rate. Should rise within 2 weeks of deployment. If it does not, the repricer is undercutting unnecessarily or your floors are too high.
2. Average margin per unit. Should be stable or rising, NEVER falling. If margin is falling, either floors are too low or competitor pressure has shifted.
3. Submission rate. Should be stable. Spikes indicate oscillation or unexpected catalog churn.
4. Rejection rate. Should be near zero. High rejections indicate listing or pricing rule misconfiguration.

Set up alerts on all four. SellerVault's repricing dashboard surfaces them by default; if your repricer does not, build your own dashboard to track them — do not run a repricer blind.