Dynamic pricing for small online stores: a practical starting point

Dynamic pricing has a reputation built by airlines and ride-hailing apps: prices that move every few minutes, driven by models nobody can explain. For a small online store it means something much plainer — prices that respond to what competitors charge, what you have in stock and what is actually selling, reviewed on a schedule and kept inside limits you set in advance.

What dynamic pricing means at small scale

Strip away the jargon and dynamic pricing is a loop: collect a few signals, decide whether a price should change, change it, and check what happened. For a store with a few hundred products, three signals cover most of what matters:

• Competitor prices for the same or directly comparable products, including whether those competitors actually have stock.
• Your own stock — how much you have, how fast it is moving and whether anything is ageing.
• Your own sales — units and revenue per product, before and after each change.

It is worth being clear about what dynamic pricing is not. It is not showing different customers different prices for the same product at the same moment; that is personalised pricing, which carries different legal and reputational risks. We cover the line between the two in dynamic vs personalised pricing.

When it is worth the effort — and when it is not

Dynamic pricing pays off when most of the following are true:

• You sell products that customers can find elsewhere: branded goods, standard parts, anything with an EAN or a model number.
• You have more products than you can sensibly review by hand each week.
• Your margins leave room to move. If a 3% price cut would wipe out your profit on an item, there is little to optimise.
• Competitors change their prices often enough that a price set in January is wrong by March.

It is rarely worth it for unique products (handmade goods, your own brand sold nowhere else), for very small catalogues where a monthly review is enough, or where supplier agreements fix the price you can advertise. In those cases, good initial pricing matters far more — see how to price products for an online store.

Set the guardrails before anything else

Every problem story about automated pricing — the book listed at absurd prices, the store that matched a competitor's typo — is a story about missing guardrails. Decide these before you change a single price:

1. A floor. The lowest price you will accept for each product: landed cost plus fees plus a minimum margin. Nothing goes below it, whatever a competitor does.
2. A ceiling. Usually the RRP, or a fixed percentage above your normal price. It stops a competitor stock-out from pushing you to a price that damages trust.
3. A maximum step. How far one change may move a price. Small steps are easier to reverse and easier to measure.
4. Supplier rules. If a brand sets a minimum advertised price, that is a second floor. See MAP monitoring.
5. Rounding. Decide whether prices end in .99, .95 or whole numbers, so changed prices still look deliberate.

Pick a cadence you can keep

Small stores rarely need to reprice more than once a day, and many do well with daily data and weekly decisions. The point of a cadence is to make pricing a routine rather than a reaction:

• Daily: collect competitor prices and stock status; look only at the exceptions — new undercuts, competitor stock-outs, your own slow sellers.
• Weekly: decide and apply changes in one batch, so you know what changed when.
• Monthly: review whether last month's changes did what you intended, and adjust the rules rather than individual prices.

Five rules that cover most situations

You do not need a model to start. These five rules, applied inside your guardrails, handle most of what a small store sees:

1. Main competitor out of stock: hold your price, or raise it by one small step if you have stock and the product is selling. Do not raise it to the ceiling; the competitor will be back.
2. A credible competitor undercuts you on a key product: review it rather than matching automatically. Check whether they have stock, whether shipping changes the picture, and whether your sales have actually dropped. More on this in how to respond when a competitor lowers prices.
3. You are well below the market and selling steadily: test a small increase and watch sales for two weeks.
4. Stock is ageing: mark down in steps, not in one large cut, and stop when it starts moving.
5. Nothing has changed: leave the price alone. Changing prices for the sake of it adds noise and makes measurement harder.

A worked example

This example is illustrative; the store and figures are invented to show the mechanics.

A store sells 250 coffee-equipment products, about 90 of which are branded items also sold by four competitors. The owner sets a floor of cost plus fees plus 10%, a ceiling at RRP, a maximum step of 5% and .99 rounding. Each morning she looks at a list of exceptions only. In a typical week that list might show three products where a competitor has dropped below her, two where the cheapest competitor is out of stock, and a handful of slow movers. She reviews ten or so prices, not 250, and changes perhaps six. The other 160 own-brand and niche products she reviews quarterly.

The important part is what she does not do: she does not chase every competitor move, and she does not touch prices that are selling well at their current level.

The legal and trust points to know

• Price-reduction announcements in the EU. If you advertise a discount, the reference price must be the lowest price you applied in a period of at least 30 days before the reduction, under Article 6a of the Price Indication Directive. Frequent small up-and-down moves make "was" prices harder to justify, so keep discount messaging separate from routine repricing.
• Volatility is visible. Customers who see a price change twice in a day lose confidence. A maximum step and a fixed cadence prevent that.

Spreadsheet, rules or software

With a few dozen products and two or three competitors, a spreadsheet and a weekly half-hour will do. Beyond that, the time goes on collecting prices rather than deciding them, and that is where software earns its keep. Luska, for example, checks the competitor product pages you add once a day, flags undercuts and stock-outs, and on the Professional plan suggests prices with a reason and a confidence score. Its suggestions never go below an 8% minimum margin or move more than 15% at a time, and nothing changes until you approve it.

Measure, or you are just moving numbers

A price change you never evaluate is a guess. At minimum, compare units and revenue for the week before and the week after each change, and note anything else that changed at the same time — a promotion, a stock-out, a season starting. The method is set out in how to measure whether a price change worked.

FAQ

Is dynamic pricing legal for a small online store?

Changing prices over time based on market conditions is generally lawful. The rules to watch are how you advertise discounts, any supplier pricing agreements, and personalising prices to individuals, which has its own disclosure duty in the EU.

How often should prices change?

As often as the market gives you a reason, and no more. For most small stores that means a few products a week, not the whole catalogue every day.

Do I need sales data to start?

No, but you need it to know whether your changes worked. Start with competitor prices and your costs; add sales data as soon as you can.