Dynamic Pricing

Dynamic pricing is the practice of algorithmically adjusting product prices in response to real-time demand signals, competitor pricing, inventory levels, and customer data — enabling retailers to optimise revenue and margin across their assortment without manual intervention. It exists on a spectrum from simple rule-based repricing to AI-driven elasticity models that process dozens of variables simultaneously.

How it works

Dynamic pricing systems draw on multiple input streams simultaneously: website traffic, wishlist adds, cart abandonments, inventory levels, competitor pricing, customer purchase history, time-of-day signals, and external factors such as weather or news cycles (Alhena AI, 2026-04-03). Modern AI systems evaluate up to 60 variables in milliseconds (Alhena AI, 2026-04-03).

Three main architectural approaches exist:

Rule-based: Explicit if-then conditions ("if competitor drops below £X, match within 2 hours"). Fast to implement, brittle under complex market conditions.

Elasticity-based optimisation: McKinsey's five-module framework covers long-tail pricing, elasticity modelling, Key-Value Items (KVI), competitive response, and omnichannel coordination — arriving at optimal prices "over weeks or months rather than minutes or hours" for most retailers, contrasting with airline/hotel real-time models (McKinsey, 2021-07-09).

AI/ML-driven: Machine learning models train on historical transaction data and feedback loops without ongoing human intervention; Algolia distinguishes these from rule-based systems as "training themselves over time" (Algolia, 2025-05-15). Effective models typically require 12–24 months of clean transactional data (multiple trade sources, as-of 2026).

McKinsey advises that the most important prerequisite is cost data quality: "the quality of your cost data is almost always terrible" and fixing item-level cost data is "by far the most important thing" before deploying pricing automation (McKinsey, 2021-07-09).

Adoption (as-of 2026)

Deloitte's 2026 Retail Industry Global Outlook (n=330 global retail executives, surveyed Oct–Nov 2025) found 48% of retailers currently using AI for pricing and promotions optimisation, with 38% planning deployment within 12 months — making it the second most adopted AI use case in retail after fraud detection (Deloitte, 2026-01).

A TCS 2026 report found 42% of brands prioritising profitable growth plan to implement dynamic pricing (cited by Shopify, 2026).

Amazon's repricing frequency is itself contested:

Revenue and margin benchmarks (as-of 2026)

McKinsey benchmarks dynamic pricing at sales growth of 2–5% and margin increases of 5–10% across retailer pilot programs (McKinsey, 2017-03-27). These figures are widely cited but originate from a 2017 article.

A named European fashion retailer implemented AI-driven dynamic pricing in 2025, reporting a 12% increase in sales while maintaining profitability (tgndata, 2025-10-03). No retail name, methodology, or baseline is disclosed.

56% of consumers may abandon purchases entirely when they encounter an unexpected price change (Morning Consult, cited by Alhena AI, 2026-04-03) (as-of 2026).

Consumer perception

Gartner (October 2024 survey, n=303 US consumers): 68% of consumers report feeling taken advantage of when brands use dynamic pricing; 80% say brands with consistent pricing are more trustworthy; 42% would spend more if consistent pricing was guaranteed; 79% experienced an unexpected pricing scenario (surge pricing, hidden fees, or unforeseen rate hikes) in the prior year. Gartner analyst Kate Muhl described "suspicion and frustration" as "fueling distrust and price paranoia" (Gartner, 2024-12-16).

A majority of US adults oppose both dynamic pricing and AI-informed dynamic pricing (Morning Consult, cited by Alhena AI, 2026-04-03).

Algolia positions "dynamic discovery" — using customer data to serve the most relevant products at a fixed universal price rather than varying prices per customer — as an ethical alternative, citing consumer distrust of per-customer price variation (Algolia, 2025-05-15).

Legal and regulatory landscape (as-of 2026-07)

EU — existing obligations:

• The EU Omnibus Directive requires retailers to display the lowest price from the past 30 days when running promotions, and to disclose when prices are personalised based on browsing history or purchase behaviour (DynamicPricing.AI, 2026-01-21).
• EU rules prohibit drip pricing — the total price including all taxes and mandatory fees must be shown from the first consumer touchpoint (DynamicPricing.AI, 2026-01-21).

EU — incoming regulation:

• The Digital Fairness Act (expected Q3 2026 draft): EU consumer groups Euroconsumers, Football Supporters Europe, and Live DMA are calling on EU legislators to introduce a targeted ban on dynamic pricing in live events under the Act; the Regulatory Scrutiny Board reviewed the Act on July 1, 2026 (Pollstar News, 2026-06-23). The Act will also ban fake urgency, pre-ticked boxes, and difficult cancellation flows — mechanics frequently deployed alongside dynamic pricing (DynamicPricing.AI, 2026-01-21).
• The EU AI Act and broader global AI guidelines are moving toward addressing fairness and transparency requirements for algorithmic pricing systems (Lexology, undated).

US — state regulation:

• New York Algorithmic Pricing Disclosure Act (effective 2025-11-10): retailers using personal data in pricing algorithms must display the notice "THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA," with civil fines up to $1,000 per violation (DynamicPricing.AI, 2026-01-21; Shopify, 2026).
• California AB 325 (effective 2026-01-01): prohibits common pricing algorithms used in anticompetitive agreements (Shopify, 2026).

US — federal antitrust:

• On May 14, 2026, DOJ Antitrust Division Acting Deputy AG for Criminal Enforcement Daniel Glad announced that companies using shared algorithmic pricing tools risk criminal prosecution — including potential prison sentences for executives and substantial fines for corporations (A&O Shearman legal alert, 2026-05-27).
• The DOJ criminal theory rests on a hub-and-spoke conspiracy framework: competitors (spokes) each feed competitively sensitive data into a common pricing algorithm (hub), and the resulting coordination forms the "rim" of a horizontal conspiracy. Four factors increase criminal risk: (1) competitors share non-public pricing, cost, or supply data with a common platform; (2) participants know their data will inform competitors' pricing; (3) the shared information would be unlawful to exchange directly; (4) compliance measures are absent or ceremonial (A&O Shearman, 2026-05-27).

Vendor landscape (as-of 2026)

Vendor	Positioning	Notes
Competera	Enterprise, elasticity + XAI	Explainable AI layer added 2026
Pricefx	Enterprise, configurable	Broad industry coverage
Intelligence Node	Competitor intelligence + pricing	Real-time competitor monitoring
PROS	B2B, sub-300ms engine	IDC MarketScape Leader 2025/2026
Aptos Revionics	Retail-specific AI pricing	Gartner MQ "Unified Price, Promotion and Markdown Optimization" category
Omnia Retail	Mid-market, omnichannel	Visual pricing logic editor
Prisync	SMB, competitor-led repricing	Simple workflow focus
DynamicPricing.AI	Shopify-native, SMB	Rule-based + AI models

Gartner's market category for this software is "Unified Price, Promotion and Markdown Optimization Applications" — covering dynamic pricing, Markdown Optimisation, and promotions in an integrated platform (Gartner Peer Insights, 2026).

The global dynamic retail pricing software market is projected to reach $36.9 billion by 2032 (Grand View Research, cited by Alhena AI, 2026-04-03; methodology unverified) (as-of 2026).

Fashion and apparel considerations

Fashion's price elasticity of demand is approximately -0.89 (relatively inelastic) versus -1.72 for electronics (americanimpactreview.com, undated; confidence: low, no primary source cited). The practical implication — if accurate — is that dynamic pricing in fashion yields smaller conversion lift but lower abandonment risk than in electronics.

Fashion faces a distinct inventory perishability problem: a product that sells well in September may be a markdown candidate by December, making AI-driven Markdown Optimisation the most common application of dynamic pricing logic in the sector rather than real-time consumer-facing price changes (Fashion AI School, 2025-10-01).

Boohoo and PrettyLittleThing tested real-time dynamic pricing in 2025; shoppers noticed the same items fluctuating in price within hours, raising trust and reputational concerns (Fashion AI School, 2025-10-01).

Luxury fashion avoids consumer-facing dynamic pricing — price stability signals exclusivity. AI may be applied to regional offers or personalised bundles behind the scenes rather than to listed prices (Fashion AI School, 2025-10-01).

Key fashion-specific risks: customer distrust from frequent price fluctuations; algorithms misreading demand signals (e.g., bots inflating apparent demand); geographic price discrimination ethics; regulatory intervention (Fashion AI School, 2025-10-01).

[!unverified] 90% of AI pricing initiatives reportedly never scale past the pilot stage in fashion retail, usually because underlying data and operating models are too weak (masterofcode.com, undated; no primary source cited; confidence: low).

Intersection with agentic commerce

As Agentic Commerce matures, AI agents shopping on behalf of consumers will compare prices algorithmically and in real time. Deloitte (2026-01) notes that 81% of retail executives believe generative AI will weaken brand loyalty by 2027 by focusing consumers on value over brand recognition — a structural pressure that may accelerate dynamic pricing adoption while increasing consumer sensitivity to price changes.

Key terms

Term	Meaning
Price elasticity	Percentage change in quantity demanded per 1% change in price; negative for normal goods
KVI (Key-Value Item)	SKU where consumers have strong price memory; small changes trigger disproportionate trust damage
Markdown optimisation	End-of-season or clearance price reduction managed algorithmically
Hub-and-spoke conspiracy	Antitrust theory: a shared platform (hub) coordinates price information among competitors (spokes)
Dynamic discovery	Algolia's alternative model: personalised product relevance at a fixed universal price