Adobe Commerce development
Storefront design
Performance & SEO
AI & agentic commerce
Release tracker
Ecommerce dynamic pricing means changing prices based on conditions such as demand, stock, competitor prices, time or customer group, instead of setting them once. There are two broad approaches: rule-based pricing, where you write the rules ("10% off when stock is over 500 units"), and AI or algorithmic pricing, where a model recommends or sets prices toward a goal such as margin or sell-through. Either way, guardrails decide whether it helps or hurts: price floors, MAP policies, limits on how often prices change, and the disclosure rules that now apply in the EU and New York. Adobe Commerce handles rule-based pricing well out of the box; algorithmic pricing needs an external engine feeding prices in.
Rule-based vs AI pricing
| Aspect | Rule-based | AI / algorithmic |
|---|---|---|
| How prices are set | If-then rules written by your team | A model recommends or sets prices from data (sales, stock, competitors, seasonality) |
| Best for | Promotions, clearance, customer-group and B2B pricing, simple stock-based markdowns | Large catalogs with frequent competitor movement, perishable or seasonal stock, marketplaces |
| Data needed | Little; your own catalog and cost data | Clean sales history, costs, competitor feeds, stock and returns data |
| Explainability | Fully explainable | Varies; insist on reason codes for each price change |
| Main risk | Rules multiply and conflict | Optimizing the wrong goal, price wars, customer distrust, compliance gaps |
| Where it runs | Inside Adobe Commerce | External pricing tool or custom service, pushing prices into the store |
Most merchants should start with rules and add algorithmic recommendations for a subset of the catalog, reviewed by a person, before letting any system publish prices on its own. Our overview of AI in ecommerce covers where machine learning tends to pay off first.
Guardrails every dynamic pricing setup needs
- Price floor: landed cost plus a minimum margin per category. No rule or model should go below it.
- Price ceiling: for example, no more than a set percentage above your reference or list price, so a demand spike never produces an outrageous price.
- MAP compliance: where suppliers set minimum advertised price policies, the advertised price must respect them; keep MAP values as product attributes the engine reads.
- Change frequency and size: limit how often and how much a price can move in a day. It protects customer trust and your caches.
- Competitor matching rules: define which competitors count, ignore out-of-stock offers, and never follow a competitor below your floor.
- Fairness: do not let prices vary by signals that could act as proxies for protected characteristics. Prefer group-level prices (trade, member, region) over individual ones.
- Audit trail: log every price change with the rule or model reason, the old and new price, and who or what approved it.
- Kill switch: a single setting that freezes automated changes and reverts to list prices.
Legal checks before you launch
Pricing law differs by country; these are the rules most often missed. This is general information, not legal advice; confirm with counsel for your markets.
- EU personalized prices: the Omnibus Directive (EU) 2019/2161 amended the Consumer Rights Directive so that traders must tell consumers when a price has been personalized on the basis of automated decision-making. National laws carry this into force; Germany, for example, does so in Art. 246a § 1(1) No. 6 EGBGB. The European Commission states that the directive's amendments apply from May 28, 2022.
- EU price reductions: the same directive added a rule to the Price Indication Directive. When you announce a price reduction, the prior price shown must be the lowest price you applied in a period of at least 30 days before the reduction; the EU Court of Justice confirmed in September 2024 (Case C-330/23) that the announced reduction itself must be calculated from that price. Frequent dynamic changes make that history harder to track, so store it.
- New York: the Algorithmic Pricing Disclosure Act (General Business Law § 349-a) took effect on November 10, 2025. An entity that sets a price using personalized algorithmic pricing must show, clearly and at or near the price, the disclosure "THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA." The statute lists exceptions, for example for insurers and certain financial institutions.
- Competition law: sharing pricing algorithms or data with competitors, or using a common tool that coordinates prices, can raise antitrust issues.
What Adobe Commerce price rules can and cannot do
Adobe Commerce and Magento Open Source include a strong rule-based toolkit:
| Tool | What it does | Edition |
|---|---|---|
| Catalog price rules | Discounts on products before they are added to the cart, by product attributes, website, customer group and date range | Both |
| Cart price rules | Discounts in the cart by cart contents, totals, coupons, customer group; free shipping and buy X get Y | Both |
| Special prices and tier prices | Dated special prices; quantity breaks per customer group | Both |
| Customer segments in rules | Target cart rules and content at dynamic segments based on behavior and attributes | Adobe Commerce |
| Scheduled updates (staging) | Schedule price and rule changes as campaigns and preview them | Adobe Commerce |
| Shared catalogs (B2B) | Company-specific catalogs and custom prices | Adobe Commerce B2B |
What they cannot do natively:
- Watch competitor prices or demand and adjust automatically.
- Optimize prices toward a goal such as margin or sell-through.
- Enforce floors and MAP across all rules at once; overlapping rules can stack unless you set priorities and "discard subsequent rules" carefully.
- Price per individual shopper. Prices are set per website, customer group or segment, which is also the safer design for fairness and disclosure.
How to run algorithmic pricing with Adobe Commerce
- Keep the engine outside. A pricing tool or custom service calculates prices, applies the guardrails above and keeps the audit log.
- Push prices in bulk. Use the REST price endpoints, such as
POST /V1/products/base-prices,/V1/products/special-priceand/V1/products/tier-prices, ideally through the asynchronous bulk API for large batches. - Batch changes on a schedule. Every price change affects indexers and invalidates cached pages. A few scheduled batches a day are far kinder to performance than constant single updates.
- Keep promotions as rules. Let the engine set base or special prices and keep marketing promotions in Commerce rules, with clear stacking priorities.
- Monitor outcomes. Track margin, conversion and price-change counts per category, and review exceptions weekly.
Pricing works best alongside other levers. Member or loyalty pricing is often a better route than individual prices; see ecommerce loyalty programs. For tailoring content and recommendations instead of prices, see ecommerce personalization.
FAQ
Is dynamic pricing legal?
Generally yes, with conditions: honest reference prices, required disclosures for personalized or algorithmic pricing in some places, no collusion, and no discrimination. Check the rules in each market you sell to.
Will dynamic pricing hurt SEO or shopping ads?
Not by itself, but your product feed, structured data and landing page price must match. Push price changes to feeds at the same time as the store.
Can I show different prices to logged-in customers?
Yes, through customer groups, segments or B2B shared catalogs. Make sure full-page cache variations are set up correctly so the right price is shown.
Next step
Start with the guardrails and an audit log, then automate. Mistri's AI for ecommerce work covers pricing recommendations and the data they need, and our Magento extension development team builds the price import, guardrail and rule logic inside Adobe Commerce.
Sources
Directive (EU) 2019/2161 (EUR-Lex); Attorney General James warns New Yorkers about algorithmic pricing as new law takes effect (New York Attorney General, November 2025); Catalog price rules (Adobe Experience League); Cart price rules (Adobe Experience League); New York General Business Law § 349-a (New York State Senate); Court of Justice press release No 152/24, Case C-330/23 Aldi Süd (CURIA, September 2024); Price Indication Directive (European Commission); Art. 246a EGBGB (Gesetze im Internet, German Federal Ministry of Justice); Content staging (Adobe Experience League); Create and delete customer segments (Adobe Experience League).
