Repricing software with a margin floor you set yourself

Repricing software changes the prices in your store automatically, according to rules you write, using competitor prices, your cost and your margin floor as inputs. The rule is the product. A repricer with no margin floor is an efficient way to sell below cost across a whole catalogue before anyone opens a report. This page covers repricing your own storefront on Shopify, WooCommerce, BigCommerce or Adobe Commerce. It is not an Amazon buy-box repricer, and if the buy box is your problem, the next section says what to use instead.

What this repricing software does, and what it does not

There are two products in the market wearing the same word, and they solve different problems. Getting this wrong wastes an evaluation cycle, so here it is plainly before anything else.

So: if most of your revenue is Amazon FBA and the job is winning the buy box on ASINs you share with a dozen other sellers, use a dedicated Amazon repricer. The mechanics genuinely differ, the cadence is faster, and a storefront tool will not serve you well. We connect to Amazon Seller Central to monitor marketplace listings, track third-party sellers of your brand and detect below-MAP offers. We do not push buy-box price changes, and we would rather say so on this page than in month two of a contract.

Guardrails first: the floors every repricing rule needs

Most repricing failures are not strategy failures. They are missing-constraint failures. The rule did exactly what it was told, on an input that was wrong, with nothing underneath to stop it. Build the floor before you build the strategy.

• Cost floor. Never below landed cost, including freight and duty. This is the seatbelt, not the strategy. If your cost data is stale, everything above this line is decorative.
• Margin floor. The real stop. Set per category, not globally, because a 15 percent floor that suits appliances will quietly cap your accessories business.
• MAP floor. Where you are an authorized reseller, the brand's advertised price minimum overrides everything below it. See what MAP pricing is if you also need to enforce it on others.
• Ceiling. The guardrail people skip. Without one, a rule indexed to the market follows a competitor's pricing error upward, or spikes when every rival goes out of stock and the remaining sample is one overpriced listing.
• Max change per day. Bound the blast radius. A 5 percent daily cap means a bad input costs you 5 percent for one day, not 40 percent before Monday.
• Minimum change threshold. Do not move a price for $0.03. Small changes churn your feeds, your caches and your customers' trust for no measurable gain.

How to design a repricing rule, in seven decisions

A rule is not a strategy dropdown. It is seven decisions, and six of them are about restraint. Work through them in this order.

1. 1

   Scope: which products this rule owns

   Define by category, brand, margin band or tag. Rules must not overlap ambiguously; where two could claim a SKU, precedence is explicit. Start with one category you know well, because your first rule is a hypothesis.

2. 2

   Strategy: what the target price is a function of

   Match, undercut, index to the market average, hold a margin, or step down on a schedule. The next section compares the common ones and names how each fails.

3. 3

   Floor: the price the rule may never go below

   Cost floor, margin floor and MAP floor all evaluate, and the highest of them wins. When a floor blocks a proposed change, that is an exception worth reading, not a silent no-op. Repeated floor hits usually mean your cost moved or your matching is wrong.

4. 4

   Ceiling: the price the rule may never exceed

   Anchor it to something real, such as your own list price or a multiple of the market median. A ceiling protects you from a thin competitor sample as much as from a bad number.

5. 5

   Limits: how far and how often it may move

   Maximum change per day as a percentage or an absolute amount, plus a cap on how many times a single SKU can move in a week. Bounded movement is what makes automation recoverable.

6. 6

   Exclusions: what it must never touch

   New launches inside their first weeks, loss leaders, bundles and kits, anything with fewer than three in-stock matched competitors, and anything whose match confidence is below your threshold. Thin evidence is the most common cause of a confidently wrong price.

7. 7

   Output: where the proposed change goes

   Review queue, CSV export, or a direct push to the connected store. New rules propose, they do not push. Promotion to automatic is something a rule earns by being right for a few weeks.

Ecommerce repricing strategies, and how each one fails

Every strategy is right somewhere and dangerous somewhere else. The failure column is the one to read.

The arithmetic: what a 5 percent price cut has to earn back

Before automating anything, be certain the people writing the rules can do this on a whiteboard. A price cut is not a discount off revenue. It comes straight out of contribution margin, which is a much smaller number, so the volume it needs to break even is much larger than intuition suggests.

required unit lift = price cut / (current margin per unit - price cut)

  $100 price, $60 cost, 5% cut
  = 5.00 / (40.00 - 5.00)
  = 5.00 / 35.00
  = 0.143, so 14.3% more units just to stand still

Walk it through with real quantities. A product sells at $100.00 and costs you $60.00, so contribution margin is $40.00 a unit. Sell 100 units and you bank $4,000.00 of gross profit. Cut the price 5 percent to $95.00 and the margin per unit falls to $35.00. To hold that same $4,000.00 you now need 4,000 / 35 = 114.3 units. That is a 14.3 percent volume increase bought with a 5 percent discount. The question is not whether the cut wins units. It is whether it wins that many.

This arithmetic ignores returns, payment fees, fulfillment cost per order and any cannibalization of a higher-margin substitute, all of which make the real break-even worse. It also ignores whether demand responds at all, which is a measurable thing rather than a matter of opinion. Price elasticity explains how to test it on a subset before you write it into a rule.

Automated repricing without losing control

Automation is not a switch, it is a promotion path. The sequence below takes about a month and is the difference between a repricer you trust and one you turn off after an incident.

1. 1Weeks one and two: monitor only. No rule enabled. Establish your price index and position mix so you know where you actually sit. Teams routinely discover the SKUs they worried about were fine while a quiet corner of the catalogue sat thirty index points high.
2. 2Week three: one rule, one category, review queue only. Approve every proposal by hand and count how often you disagree.
3. 3Week four: read your disagreement rate. If you are rejecting more than a small share of proposals, the rule is wrong or the matching under it is wrong. Fix the input, not the threshold.
4. 4Then promote, narrowly. Move that one rule to automatic push inside a tight daily change cap, keeping the review queue for anything that hits a floor, a ceiling or a thin competitor sample.
5. 5Widen scope, not aggressiveness. The next step is a second category on the same conservative settings, never the same category on looser ones.

Every change is stored with the rule that proposed it, the competitor observations that triggered it, the old and new price, the timestamp and the approver. That history is what lets you answer why a price was $84.99 in March, six months later, without anybody reconstructing it from memory.

Where the new price actually lands

Approved changes push directly to Shopify, WooCommerce, BigCommerce and Adobe Commerce, or leave as scheduled CSV or SFTP exports for an ERP or PIM to consume, or fire as webhooks. The repricing module keeps all three routes open at once, because most catalogues have a subset of SKUs that should never be touched by an automated push.

The same logic applies to marketplaces, retail media and any print or email campaign already in flight. Repricing frequency should be set by the slowest system that shows your price to a customer, not by the fastest one you own. Refresh runs daily on Starter, twice daily on Growth, four times daily on Scale and hourly on Enterprise, and picking the fastest available tier is not automatically the right answer.