Repricing without burning your margin

Automated repricing makes a simple promise: if the competitor drops, you drop, and you do not lose the sale. It works, and it is also the fastest way to make your margin evaporate without noticing, because every individual cut looks reasonable while you make it.

The lowest price is not a strategy

Chasing the lowest price makes sense in one case: when you are the one with the lowest costs. If you are not, you are playing a game somebody else wins structurally, and every cut is a step towards their win.

The practical point is that elasticity is not uniform across a catalogue. On some products buyers compare to the cent and price decides everything. On others the decision is availability, delivery, trust in the seller, support. Applying the same rule to both groups is a reliable way to give away margin where it bought you nothing.

Before touching a price you need the margin

It sounds obvious and almost nobody has it per product. Without per-item margin, a repricing rule is working blind: you can know whether you are dearer than a competitor, not whether you can afford not to be.

The margin to use is not just price minus purchase cost. Shipping belongs in it, so does that category's return rate, so do the fees of the channel you sell on. On a product with 20% returns the real margin is a different number from the one on paper, and pricing rules should know that.

Guardrails, which is the part that matters

A repricing rule without limits is an automated way of being wrong faster. You need at least three brakes.

A floor per product, calculated from margin rather than guessed. You do not go below it, even when the competitor does.

A movement limit, so you do not chase a mistake. If a competitor mistypes a price or is clearing stock, a rule with no brake follows them into the ditch.

An exclusion list. New arrivals, loss leaders, items with brand constraints, anything that must not move automatically for reasons the system cannot know.

When not to reprice

Worth writing down, because it is the part people skip.

Do not reprice when the competitor you are following is not comparable: a seller with no support, no warranty or far longer delivery is not selling your product, even if the barcode matches.

Do not reprice on cuts that last a few hours, because you are reacting to noise.

And do not reprice when price is not the problem. If the page does not say what is in the box, the buyer is not choosing the cheapest: they are choosing the only one they understood. Cutting the price of a thin page is paying for a problem that writing would have solved.

How often

It depends how fast your market moves, but the real criterion is different: every price change carries a cost that appears nowhere. It confuses anyone comparing, it puts the feed out of step with the page until it refreshes, and it makes your own data unreadable, because you will no longer know whether a change in sales came from price or from something else.

Moving less, with clearer rules, almost always pays better than moving constantly.

Frequently asked questions

How do I get per-product margin if I do not have it?

Start from what you have: purchase cost and selling price give a gross margin, already enough to set a prudent floor. Then refine it with shipping, returns and fees. An approximate floor that exists protects far more than a perfect margin that does not yet.

Do I risk a price war with a competitor doing the same?

Yes, which is why the floor is not optional. Two systems chasing each other without a limit converge on cost. A fixed floor turns the chase into a decision: below that threshold you let the sale go, knowingly.

Does repricing hurt search visibility?

Not the price itself. The mismatch does: if the feed says one price and the page another because one updates late, products can be suspended. The more often you change, the more that risk matters.

One rule for the whole catalogue, or many rules?

A few rules for groups of products that behave alike. One rule for everything ignores differences in elasticity; one rule per product is unmaintainable. The right level is usually the category, with exceptions stated.

Is it useful if I sell own-brand products with no direct competitors?

Competitor repricing buys you little, but the guardrail part stays useful: knowing per-product margin and having a floor changes how promotions get built, and promotions are where own brands leave the most margin on the table.