Repricing strategy for e-commerce: segments, positions and guardrails

Repricing is the routine of adjusting your prices as the market moves. A repricing strategy is the set of decisions that tells you, in advance, how to adjust them: which products, against which competitors, towards which position, within which limits. Without one, repricing turns into a series of reactions. With one, most price changes become routine and the hard cases stand out.

The six parts of a repricing strategy

1. An objective per product group. Not every product should be priced for the same goal. Some exist to win customers, some to earn margin, some to clear.
2. A reference set. The competitors you actually compare against for each group — usually a short list of credible sellers, not everyone who lists the product.
3. A target position. Where you want to sit against that reference set: below, at or above the market.
4. Triggers. The events that prompt a review: an undercut, a competitor stock-out, slow sales, a cost change.
5. Guardrails. A floor, a ceiling and a maximum change per step, plus any supplier minimum advertised prices.
6. A cadence and an owner. When changes are reviewed and who approves them.

Step 1: split the catalogue

Treating every product the same is the most common repricing mistake. A simple split into three groups goes a long way:

• Traffic products. The items customers compare most and use to judge whether your store is expensive: well-known brands, popular models, products you advertise. Objective: stay competitive, because these shape your price image.
• Core range. Products that sell steadily and are compared less closely. Objective: earn margin while staying reasonable against the market.
• Long tail and own brand. Products with few or no direct competitors. Objective: margin. Review them less often and rarely against competitors.

A fourth group, clearance, applies temporarily to anything you need to sell through: end-of-line, seasonal, overstock.

Step 2: express position as a price index

"Be competitive" is not a rule anyone can apply. A price index makes position measurable: your price divided by the market median, multiplied by 100. An index of 100 means you are at the median, 95 means 5% below it, 108 means 8% above.

Once position is a number, you can set a target per group — for example 97–100 for traffic products, 100–105 for the core range — and review only the products that have drifted outside their band. That turns a catalogue of hundreds of prices into a short exceptions list. The idea is explained in more detail in our competitive pricing strategy guide.

Step 3: choose your repricing rules

Most workable repricing strategies are built from a handful of rule types:

• Position rule: keep the index within a band. The default for traffic and core products.
• Stock-aware rule: when the cheapest credible competitors are out of stock, the market price is effectively the next one up. Hold or raise by a small step.
• Velocity rule: if a product is selling well above its usual rate at its current price, test a small increase; if it has stalled, review it.
• Markdown rule: for clearance, reduce in planned steps over a set period, and stop when sell-through is on track.
• Cost rule: when a supplier raises prices, review every affected product against its floor before the new stock arrives.

Rules of the form "always beat the cheapest competitor by one cent" deserve a warning. Applied by two sellers at once, they drive both prices down to the floor. If you use them at all, limit them to a few products, add a floor well above cost and exclude competitors without stock. For the trade-offs between rule-based and machine-learning approaches, see rules vs ML repricing.

Step 4: set guardrails that cannot be overridden by accident

• Floor: cost plus fees plus your minimum margin. Recalculate it when costs change.
• Ceiling: usually the RRP, so a competitor stock-out cannot push you to a price that looks opportunistic.
• Maximum step: limits how far one change can move a price, which keeps changes reversible and measurable.
• Minimum advertised price: where a brand has a MAP policy, treat it as a second floor.
• Approval threshold: small changes can go through on routine; large ones get a second look.

A note on marketplaces

If you also sell on marketplaces, treat them as a separate channel with their own strategy. Marketplace repricing is often about winning the featured offer on a shared listing, which rewards different behaviour from pricing on your own store. Keep the two price lists separate in your rules, and make sure a marketplace price war does not leak into your own site.

Step 5: decide cadence and approval

Daily competitor data and a fixed review rhythm suit most online stores. Fast-moving categories at busy times of year may justify more frequent checks; slow categories rarely need more than weekly. Whatever you choose, batch the changes, record why each was made, and keep a human in the loop for anything outside the usual bands.

A worked example

Illustrative figures, invented to show how the parts fit.

An outdoor-gear store has 600 products. It classes 80 as traffic products with a target index of 97–100, 320 as core range at 100–106, and 200 as own-brand and long tail, reviewed quarterly. Its floor is cost plus fees plus 12%; its maximum step is 5%; changes above 5% need the owner's approval.

On a given morning, daily data shows 14 products outside their bands. Six traffic products sit at 102–104 after competitors cut prices; four of those competitors have stock, so those four are moved down one step each. Five core products sit below 100 because competitors have raised prices; they are moved up one step. Three products are flagged because the cheapest competitor reading is less than half the median — almost certainly an error — and are left alone pending a check.

How to tell whether the strategy works

Judge a repricing strategy on gross profit, not on how often you are cheapest. Track, per product group:

• Gross profit and units sold, month on month.
• The share of products inside their target index band.
• How many changes were made, and how many were reversed.

For individual changes, compare sales before and after — see how to measure whether a price change worked.

Where software fits

The strategy is yours; software handles the data and the arithmetic. Luska shows a price index per product (100 is the median of competitor prices read in the last 48 hours), the gap to the cheapest competitor and your margin. On the Professional plan it suggests prices based on the target margin, price position, rounding and approval threshold you set, never below an 8% minimum margin and never more than 15% in one change. Suggestions wait in an approval queue.

FAQ

What is the best repricing strategy?

There is no single best one. A strategy that segments the catalogue, sets a position per segment against the market median and enforces a floor is a sound starting point for most stores.

How often should I reprice?

Review daily data, but change prices only when a product leaves its band or a trigger fires. Most products should not change most days.

Should I reprice against marketplaces?

Include marketplace sellers only if your customers genuinely compare with them. Many marketplace listings come from sellers your customers would not buy from.