Dynamic pricing in eCommerce: how to build a pricing strategy based on data, margin, and automation

Price is one of the most important business parameters in eCommerce. It affects not only conversion, but also margin, cash flow, inventory turnover, the effectiveness of advertising campaigns, position relative to competitors, and brand perception. Despite this, many online stores treat pricing in a highly operational way: as a value entered in the admin panel, a discount in a promotional campaign, or a reaction to a competitor's price cut.

In mature eCommerce, price should not be accidental. It should be a strategic decision.

Dynamic pricing, or dynamic price management, is one of the most advanced ways to optimize pricing policy. However, it is not about chaotically changing prices or automatically being the cheapest on the market. Its goal is to adjust the price to current business conditions: demand, supply, stock level, margin, seasonality, competition, advertising costs, and the goal assigned to a given product.

When implemented well, dynamic pricing can increase profit, improve inventory turnover, and reduce unnecessary markdowns. When implemented poorly, it can lead to a price war, margin loss, promotional chaos, and declining customer trust.

Price as a profit lever, not just a marketing element

Many online stores focus on increasing traffic: SEO, Google Ads, Meta Ads, marketplaces, affiliate marketing, newsletters, and marketing automation. These are obviously important areas, but even the best traffic will not solve the problem if the store has a poorly designed pricing policy.

Price directly affects financial results. What's more, its impact is often greater than that of a small increase in conversion or traffic.

Let's assume the store sells a product for $100. The purchase cost is $60, so the gross margin on the product is $40.

If the store lowers the price to $90, the gross margin falls to $30. That means that with the same number of units sold, gross profit drops by 25%. To earn as much as before, the store has to sell significantly more units.

This shows the basic problem with many promotions: higher sales do not always mean higher profit.

A store may sell more products, have higher revenue, better advertising campaign statistics, and at the same time earn less. That is why professional price management should be based not only on revenue, but above all on margin, gross profit, inventory turnover, and profitability after sales costs.

What is dynamic pricing?

Dynamic pricing is a price management model in which a product's price changes based on previously defined rules, market data, or predictive algorithms. The price change may result from demand, product availability, competitors' prices, sales history, seasonality, logistics costs, advertising costs, stock level, price elasticity, and the business goal assigned to the product.

In practice, dynamic pricing may mean, for example, that the store:

keeps a higher product price when competitors are out of stock,

lowers the price of a product that is sitting in the warehouse,

does not discount bestsellers with low stock levels,

adjusts the price to the season,

raises the price of a product that has high conversion despite no promotion,

changes prices only for selected variants, sizes, or colors,

sets different pricing rules for its own store, the marketplace, and advertising campaigns.

It is also important to distinguish dynamic pricing from personalized pricing. Dynamic pricing most often concerns changing a product's price depending on market or business conditions. Personalized pricing, by contrast, means varying the price for specific customers or customer groups based on data about them. The OECD points out that personalized pricing in the digital environment is a separate issue and is linked to consumer data analysis, automation, and potential consequences for competition and consumer protection.

This distinction is crucial. Changing the price of children's sandals from $59 to $49 after the season is dynamic pricing. Showing two different prices to two customers based on their browsing history, location, or shopping profile is already the domain of price personalization, which is far more sensitive legally and reputationally.

Dynamic pricing is not a strategy for being the cheapest

The biggest mistake in thinking about dynamic pricing is the assumption that the system should automatically track competitors and set a slightly lower price.

This is a very dangerous approach.

If a store always tries to be the cheapest, it quickly falls into a price war. A competitor lowers the price to $79, the store drops to $77, the competitor responds with $75, and after a few rounds nobody is making money. An increase in the number of orders may look good in sales reports, but the margin gets destroyed.

Professional dynamic pricing does not ask: “how to be the cheapest?”. It asks: “what price best achieves this product's business goal?”.

For one product, the goal will be maximizing margin. For another, fast inventory turnover. For a third, increasing basket value. For a fourth, maintaining premium positioning. For a fifth, generating traffic and acquiring new customers.

The same pricing logic should not be applied to the entire assortment.

Pricing as a decision-making system

An online store should not treat price as a single value. Price should be part of a decision-making system.

A bestseller should be managed differently from a premium product, a seasonal product, or a product that has been sitting in the warehouse for 120 days.

Example:

Product A costs $49, sells every day, has a good margin, and a low return rate. The goal may be to maximize profit and maintain the price.

Product B costs $129, but has not sold for three months. The goal may be to free up cash and warehouse space, even at the expense of a lower margin.

Product C costs $19, has a low margin, but often ends up in the cart together with a product priced at $89. The goal does not have to be unit profit, but increasing the average order value.

Product D costs $99, has a strong brand, limited availability, and a full size range, while competitors have only individual variants. The goal may be to maintain the price rather than match the cheapest offer on the market.

This shows that dynamic pricing should operate at the level of product segments, not the entire store.

The most important pricing strategies in eCommerce

Dynamic pricing does not replace classic pricing strategies. Rather, it combines them into a more advanced model. To implement it well, you need to understand the basic approaches to pricing.

Cost-plus pricing, or pricing based on cost and margin

This is the simplest pricing model. The store takes the purchase cost of the product and adds a specific margin.

Example:

The product cost is $40. The store sells it for $80. The gross margin is $40.

The advantage of this approach is simplicity. The store knows how much it wants to earn on the product and sets the price according to a specific formula.

The problem is that cost-plus pricing ignores the market. It does not take into account how much the customer is actually willing to pay, what prices competitors have, whether the product is unique, whether it is sitting in the warehouse, or whether demand for it is growing.

Cost-plus pricing is a good starting point, but a weak end point. If a store uses only this model, it may sell too cheaply products for which customers would pay more, and too expensively products that the market values lower.

Competition-based pricing, or pricing based on competition

In this model, the store sets the price relative to competitors.

If competitors sell the product for $74, the store can set the price at $72, $74, $79, or $85. The decision depends on whether the store has better availability, faster delivery, better reviews, greater trust, easier returns, or a stronger brand.

Competitor price monitoring is very valuable, but it should not lead to automatically copying the lowest price.

A store does not have to be the cheapest if it offers greater value to the customer. In eCommerce, you compete on price only partly. The customer also considers product availability, delivery time, shipping cost, store credibility, reviews, photos, product description, return policy, and the overall shopping experience.

Value-based pricing, or pricing based on value

In this model, price results not only from the cost of the product, but from the value the customer sees in the offer.

Two products may have a similar purchase cost, for example $35, but one sells for $59 and the other for $89. The difference may come from the brand, design, photo quality, description, reviews, packaging, product story, uniqueness, or better presentation.

Value-based pricing is especially important in premium categories, fashion, home furnishings, children's products, cosmetics, lifestyle products, and everywhere the customer is buying not only function, but also emotion, aesthetics, convenience, and trust.

In this model, the store monetizes not only the product, but the entire value proposition.

Psychological pricing, or psychological price points

Psychological pricing is not only about endings such as $99 instead of $100. It is an entire architecture of purchase decisions.

Examples:

A price of $49 may be perceived as acceptable for an impulse purchase.

A price of $99 often acts as a threshold before entering the category of a 'more expensive' product.

A price of $149 may position the product as premium, while still remaining below the $150 barrier.

Free shipping from $75 may increase average order value if the typical basket is around $62.

A bundle for $119 may seem attractive if the products purchased separately cost $139.

Psychological pricing works best when it is consistent with the architecture of the offer. Simply ending a price with .99 will not build an advantage if the customer does not understand the product's value.

Inventory-based pricing, or pricing based on stock

This is one of the most practical strategies in eCommerce.

Price should depend not only on competition, but also on how many units of the product are in stock and how quickly the product is selling.

If a product has 300 units in stock and has sold only 8 times over 60 days, the store can gradually lower the price from $59 to $49, and later to $39.

If another product has only 4 units in stock and sells every day, there is no logical reason to lower the price from $89 to $79 just because a competitor temporarily launched a promotion.

Excess inventory is frozen capital. Lack of stock is lost sales. Pricing should manage both.

Demand-based pricing, or pricing based on demand

In this model, the price changes depending on demand.

A seasonal product may cost $69 before the peak season. At the peak of the season, its price may rise to $79, and after the season fall to $59, and then to $49.

In children's fashion, examples include sandals, jackets, winter shoes, swimwear, or school products. Demand for them is strongly seasonal, so keeping the same price all year round often makes no business sense.

Demand-based pricing requires caution, however. A customer may accept a price increase resulting from seasonality or limited availability, but responds worse to changes that look random, unfair, or opportunistic.

What data is needed for dynamic pricing?

Dynamic pricing is only as good as the data it works on.

If the store has incorrect purchase costs, outdated stock levels, incorrectly calculated marketplace commissions, or does not include advertising costs, price automation will only speed up bad decisions.

A mature pricing system should use several layers of data.

Product data

This group includes: purchase cost, regular price, minimum price, maximum price, target margin, category, brand, seasonality, product status, size variants, color variants, and information on whether the product is new, a bestseller, a premium product, or a clearance product.

Without this data, the system does not know what role the product plays in the offer.

Sales data

What matters here are: number of units sold, conversion, average order value, return rate, time since last sale, sales by channel, promotion history, and the impact of price changes on sales.

Information that a product is selling is not enough on its own. You need to know whether it is selling with a good margin, in which channel, at what customer acquisition cost, and with what return rate.

Inventory data

Dynamic pricing should take into account stock level, turnover, time in stock, supplier availability, planned deliveries, and whether the product is at the end of the collection.

A product that has been sitting for 180 days requires different logic than a product that sells every day and has limited availability.

Marketing data

In eCommerce, price is strongly linked to advertising.

A product may look profitable at the gross margin level, but after adding the cost of the advertising campaign, it stops being worthwhile. That is why it is worth analyzing cost per click, ROAS, CAC, the product's share in campaigns, profitability after advertising cost, and the impact of price on conversion.

If a product priced at $99 has a good margin but requires very expensive paid traffic, its real profitability may be lower than that of a product priced at $69 that sells organically.

Market data

This group includes competitors' prices, competitors' availability, marketplace position, Google Shopping prices, market promotions, and offer visibility.

It is important, however, not to look only at price. A competitor's price without information about availability can lead to wrong decisions.

If a competitor has the product for $89 but only in one size, while the store has a full size range and fast shipping, automatically dropping to $87 may be an unnecessary loss of margin.

Example dynamic pricing rules

At the beginning, dynamic pricing does not have to rely on advanced artificial intelligence. In many stores, well-designed business rules are a better first step.

Minimum margin rule

The system should not lower the price below a specified margin level, unless the product has clearance status.

Example:

The product's regular price is $89. The minimum price that allows an acceptable margin is $69. The system may change the price in the range from $69 to $89, but it does not go lower without a manual decision or changing the product's status to outlet.

Low-stock rule

If stock is lower than 5 units, the system does not lower the price automatically.

Example:

The product costs $79. A competitor lowers the price to $72, but the store has only 3 units. The system keeps the price at $79, because the risk of losing margin is greater than the potential benefit of faster sales.

Stock aging rule

If a product has not sold for 90 days, the system lowers the price by a specified percentage or moves the product to the clearance segment.

Example:

The product costs $59 and has not sold for 90 days. The price falls to $53, and after another 30 days without a sale to $49.

Competition rule

If the competition is cheaper by less than $3, the system does not change the price.

Example:

The store sells the product for $79, the competition for $77. The difference is too small to automatically destroy the margin. The system keeps the price.

Competitor availability rule

If the competition does not have the product in stock, the store does not lower the price, and in selected cases may raise it.

Example:

The product costs $99. The competition previously sold it for $94, but is currently out of stock. The store has availability and fast shipping. The system keeps the price at $99 or tests a price of $104.

Seasonal rule

30 days before the end of the season, the system starts gradual markdowns.

Example:

A seasonal product costs $79. In the first stage, the price falls to $69, later to $59, and at the final stage of the clearance sale to $49.

Gradual markdowns are often better than a one-time aggressive price cut because they allow control of margin and turnover pace.

Advertising campaign rule

If a product has a low ROAS, the system should not automatically lower the price. First, you need to check whether the problem is the price, description, photos, availability, delivery cost, traffic quality, or the campaign itself.

Example:

The product costs $119. It gets a large number of visits from advertising, but has low conversion. A reduction to $99 may help, but it may also fail to solve the problem if the product page has weak photos, lacks a size chart, or delivery is too expensive.

Dynamic pricing cannot replace merchandising and marketing analysis.

Dynamic pricing and price elasticity

One of the most advanced elements of pricing is price elasticity, that is, demand's response to a price change.

Not every product reacts to price in the same way.

A product with high price elasticity has many competitors, is easy to compare, has few differentiating features, and the customer can quickly find an alternative. Examples include popular accessories, electronics, standard cosmetics, or products available in many stores.

A product with low price elasticity has a strong brand, uniqueness, limited availability, emotional value, or an urgent purchasing need. The customer does not give it up so easily just because the price increased by a few dollars.

Example:

The product costs $39 and sells 100 times per month. After a reduction to $35, sales rise to 108 units. At first glance, the reduction works because sales increased. But if unit margin fell more than volume rose, the store may be earning less.

Another product costs $99 and sells 20 times per month. After a reduction to $89, sales increase to 40 units. In such a case, the reduction may be justified if the additional volume compensates for the lower margin.

Dynamic pricing without elasticity analysis often leads to a false conclusion: “the discount works because sales increased.” The real question is: did profit increase?

Dynamic pricing and product variants

In many stores, a product's price should not be analyzed only at the model level, but also at the variant level.

This applies especially to fashion, footwear, children's products, home furnishings, and categories in which products have sizes, colors, or versions.

Example:

Shoes cost $89. Sizes 28, 29, and 30 sell very well, but size 35 is sitting in the warehouse. Lowering the price of the entire product to $79 may be a mistake. A better solution may be to discount only the slow-moving sizes.

The same applies to colors. If the black color sells well at $99, and the yellow color has been sitting for 100 days, there is no reason to discount all variants. Dynamic pricing should take variant turnover into account, not only the main product.

This is especially important in stores that sell seasonal collections. Sometimes the problem is not the price of the whole product, but individual variants that block warehouse space.

Dynamic pricing and marketplaces

Marketplaces introduce additional complexity into pricing policy.

A price that is profitable in the store's own shop may be unprofitable on a marketplace after adding commission, advertising costs, fees for highlighting, service costs, and potential returns.

Example:

The store sells a product for $79 in its own shop and has a good margin. On the marketplace, the same product also costs $79, but after commission, campaign cost, and additional fees, actual profitability is much lower.

That is why dynamic pricing should take the sales channel into account. The same gross price does not mean the same profitability.

In practice, a store may have different rules for:

the online store,

Allegro or another marketplace,

Amazon,

Google Shopping,

B2B sales,

remarketing campaigns,

outlet offers.

The biggest mistake is managing price only at the product level, without considering the sales channel.

Dynamic pricing and promotions

Dynamic pricing is not the same thing as a promotion.

A promotion is a tactical activity. It may relate to Black Friday, a seasonal sale, a coupon, a discount code, a “second product cheaper” offer, or free shipping.

Dynamic pricing is a continuous process. It includes decisions about whether a product should cost $49, $59 or $69 depending on the data and the business goal.

This is an important distinction, because many stores try to fix pricing mistakes with promotions. If the regular price is set incorrectly, the store starts rescuing sales with discounts. Over time, the customer learns that it is not worth buying at the regular price, because a promotion will appear soon.

In this way, the store loses control over price perception.

A mature pricing policy should clearly separate:

regular price,

promotional price,

clearance price,

minimum price,

marketplace price,

campaign price,

bundle price,

price resulting from dynamic pricing.

Dynamic pricing, law, and transparency

In Europe, dynamic price management must be designed with consumer regulations in mind, especially when communicating promotions. In the context of the Omnibus Directive, when announcing a price reduction, sellers must inform customers of the lowest price from a specified period before the reduction; in practice in Poland and in many industry materials, this is discussed as the obligation to show the lowest price from the 30 days before the promotion.

This means that a store should not artificially raise the price just to show a bigger discount later. Practices of this kind are legally risky and harmful to reputation.

Even greater caution is needed with personalized prices. The Federal Trade Commission in the U.S. has published materials concerning so-called surveillance pricing, meaning practices in which data such as location, browsing history, or other consumer information may be used for individual price setting.

From the perspective of an online store, the safest approach is simple: dynamic pricing should be based primarily on product, inventory, market, and sales data, not on hidden price differentiation for specific people.

The more the price depends on data about a specific customer, the more important transparency, regulatory compliance, personal data protection, and the risk of losing trust become.

The biggest mistakes when implementing dynamic pricing

Automatically being the cheapest

This is the most common and most expensive mistake. The store implements competitor monitoring, sets the rule “be $1 cheaper,” and after a few weeks notices that sales increased, but margin fell.

Dynamic pricing should not be a machine for destroying margin.

No minimum price

Every product should have a minimum profitability threshold. Without it, the system can fall to a level at which the sale generates a loss.

The minimum price should take into account not only purchase cost, but also commissions, payments, delivery, packing, returns, discounts, and advertising costs.

Ignoring marketplace costs

Selling through a marketplace often has different economics than selling in your own store. If the pricing system does not take commissions and channel costs into account, it may set prices that look good only at the revenue level.

No product segmentation

You cannot price bestsellers, new arrivals, premium products, outlet items, seasonal products, and end-of-stock items in the same way.

Each segment should have its own rules.

Too frequent price changes

Frequent and visible price changes may weaken customer trust. If a customer sees that a product cost $89 in the morning, $99 in the afternoon, and $79 the next day, they may feel that the price is random or manipulative.

Dynamic pricing should have limits on the frequency of changes.

Treating promotions as a pricing strategy

Promotions are a tactical tool. They will not replace well-designed pricing.

If a store sells through discounts all the time, the regular price stops being credible.

No control after implementation

Dynamic pricing requires dashboards, alerts, and regular audits.

The store should know:

which prices were changed,

why they were changed,

what the impact on sales was,

what the impact on margin was,

whether the system is violating minimum profitability rules,

whether the changes are too frequent,

whether premium products are being discounted unnecessarily.

Automation without control is risky.

How to implement dynamic pricing step by step?

Step 1: calculate the true margin

The first step is not choosing a tool, but calculating the true profitability of products.

You need to include:

purchase cost,

marketplace commissions,

payment costs,

delivery cost,

packing cost,

return cost,

advertising cost,

discounts,

service costs,

differences between sales channels.

Only then can you determine what price is truly profitable.

Step 2: divide products into segments

Example segments:

Core,

Premium,

Bestseller,

Seasonal,

Outlet,

High-stock,

Low-stock,

Ad-driven,

Marketplace-sensitive,

New collection,

End of collection.

Segmentation is necessary because different products have different goals.

Step 3: set minimum and maximum prices

For each product or segment, it is worth defining a price range.

Example:

Regular price: $89

Minimum price: $69

Maximum price: $99

Clearance price: $59, but only after changing the product's status.

Such limits protect the store against automation errors.

Step 4: build business rules

At the beginning, it is better to implement simple and understandable rules than a complicated algorithm that nobody controls.

Examples:

Do not lower the price if stock is lower than 5 units.

Do not go below the minimum margin.

Do not react to a competitor difference smaller than $3.

Lower the price of a product that has not sold for 90 days.

Do not discount premium products without manual approval.

Change prices at most once a day or once every few days.

Step 5: test on a selected category

Dynamic pricing should not be implemented across the entire store right away.

A better approach is to test it on one category, one segment, or a selected group of products. This makes it possible to check the impact of changes on sales, margin, turnover, and customer behavior.

Step 6: measure profit, not just sales

The most important metrics are:

gross margin,

gross profit,

conversion,

average order value,

inventory turnover,

sell-through rate,

share of discounted products,

ROAS after margin,

number of price changes,

impact of price on returns,

profitability by channel.

If a store measures only revenue, it may consider a bad decision a good one.

Step 7: establish governance

Dynamic pricing requires accountability.

It is worth defining:

who approves the rules,

who can change price thresholds,

who is responsible for incorrect prices,

how often results are analyzed,

when the system requires manual approval,

which products are excluded from automation.

In a large store, pricing should not be solely the task of marketing. It is an area at the intersection of eCommerce, finance, purchasing, warehouse operations, performance marketing, and category management.

Example of an advanced scenario

The store sells a product for $120. The purchase cost is $65. The competition has a price of $109, but availability only in one size. The store has a full size range, fast delivery, and good reviews.

A simple pricing system would lower the price to $107 to be cheaper than the competition.

A mature pricing system may make a different decision.

It may keep the price at $120 if conversion is stable.

It may lower only selected slow-moving variants to $115.

It may leave popular sizes at the regular price.

It may discount only leftover stock.

It may add a bundle promotion instead of lowering the price of the main product.

It may raise the free-shipping threshold to improve average order value.

It may improve the product's exposure in the campaign instead of changing the price.

This shows that dynamic pricing does not always mean a simple change to the main price. Sometimes a better decision is to change the price of variants, add a bundle, change communication, run a limited promotion for a specific channel, or keep the price despite cheaper competition.

Should every store implement dynamic pricing?

Not every store needs an advanced algorithmic system right away. But every store should have a deliberate pricing policy.

A small store can start with simple rules:

minimum margin,

product segmentation,

control of slow-moving products,

competitor monitoring for selected bestsellers,

separate rules for seasonal products,

analysis of the impact of discounts on profit.

Medium and large stores should think about dynamic pricing more broadly: as a system connecting data from ERP, OMS, the eCommerce platform, marketplaces, advertising tools, the warehouse system, and competitor monitoring.

The greatest advantage is achieved not by the stores that lower prices the fastest, but by those that best understand when the price should be lowered, when it should be maintained, and when it should be raised.

Summary

Dynamic pricing in eCommerce is not a magical algorithm or a simple machine for tracking competitors. It is an advanced profitability management process that requires data, rules, control, and an understanding of customer behavior.

Well-designed dynamic pricing combines analysis of margin, demand, stock, seasonality, competition, advertising costs, and the product's business goal. It allows better pricing decisions: not only lowering prices, but also maintaining them when a reduction makes no sense, and increasing them where the market accepts higher value.

The biggest mistake is equating dynamic pricing with being the cheapest. In mature eCommerce, price is not a tool of a desperate fight for sales. It is an element of strategy.

A store that manages prices well does not have to win every price war. It has to know on which products to make money, which products to turn over faster, which to protect from discounting, and which to treat as a tool for building the basket or acquiring customers.

That is exactly why dynamic pricing should be treated not as a feature in a system, but as part of the eCommerce business model.