What is dynamic pricing in retail? A 2026 decisioning view

Dynamic pricing in retail means changing a product's price over time, in response to demand, inventory, competition and channel signals — to optimise margin, sell-through, or both. That definition fits on a slide. The hard part is everything underneath.

Most explainers stop at "prices that move" and move on. They miss the architectural choice every retailer makes whether they notice it or not: do prices move because a rule fired, or because a decision was weighed?

Those are two different disciplines sharing one label. The confusion costs margin. This article defines the term cleanly. It names where it came from, separates the two flavours that get conflated, and lays out the five questions a head of pricing should ask before signing anything in 2026.

A one-sentence definition that holds up

Dynamic pricing is the practice of varying a retail product's price over time. The price reacts to signals about demand, inventory, competition, channel and customer context — with an explicit objective function (margin, sell-through, share, or a weighted blend).

Three words in that sentence carry the load. Signals: the inputs the price reacts to, which extend well beyond competitor scraping. Time: cadence matters as much as direction — hourly, daily, weekly, or seasonal moves are different products. Objective: a price change without a stated goal isn't optimisation, it's noise.

A common shorthand is "AI-powered pricing". That phrase tells you nothing useful. Both rule-based and optimisation-based systems get labelled AI in pitches. The honest version names the objective, the cadence, and the logic shape — not the buzzword.

Where it came from — and why retail isn't airlines

Dynamic pricing originated in airlines and hotels, where it's called yield management. A seat on tomorrow's flight is perishable in a way a t-shirt isn't: at takeoff, an unsold seat is worth zero forever.

That single-asset, single-window, single-channel shape is what made yield management tractable. American Airlines built the first serious system in the 1980s. Hotels followed in the 1990s. Both industries had one SKU per pricing decision (a specific seat, a specific room-night), one competitor set per route, and a clean clock — the takeoff or check-in time.

Retail breaks every one of those assumptions. A fashion chain prices multi-SKU × multi-store × multi-channel: tens of thousands of SKUs across hundreds of stores. Demand profiles differ store by store, and the online and offline channels either diverge or must stay aligned by law.

The competitor set is fragmented and shifts by category. The clock isn't a flight time — it's a season curve, a markdown calendar, a vendor MAP window. Each pricing decision is also entangled with markdown, replenishment, and allocation decisions on the same SKU. That coupling has no analogue in a hotel room.

Importing yield-management thinking into retail without acknowledging that coupling is what produces the "sophisticated tool, disappointing margin" pattern. The technique is sound. The shape of the problem is not the same.

The two flavours nobody distinguishes

Walk into any vendor pitch and you'll hear "dynamic pricing" used for two completely different architectures. Naming them is the first useful thing this article can do.

Rules-based dynamic pricing

Rules-based dynamic pricing is the discipline of writing pricing logic as if-then chains, then executing them fast and at scale. If competitor X drops below threshold Y, match within Z% — unless margin falls below floor F. The platform's contribution is speed and consistency: the same logic applied to 50,000 SKUs across 200 stores in minutes.

This flavour has three honest virtues. It is auditable — every price change traces back to a named rule. It is predictable — the pricing team can replay any decision in plain English. And it is governance-friendly — overrides, freezes and scope changes are immediate.

Rules-based pricing reliably captures most of the easy margin lift. It plateaus, in our experience, around 60–70% of what optimisation-based pricing achieves on the same assortment — and that range varies widely by category. For regulated products, contractual pricing and long-tail commodities, it's often the right answer.

Decision-driven dynamic pricing

Decision-driven dynamic pricing models the demand curve directly. It optimises the price–elasticity–frequency loop, coordinates with markdown and inventory decisions on the same SKU, and recomputes the trade-off as the data moves.

The output isn't a rule firing. It's an arbitration: given today's stock cover, today's competitor stance, today's brand-tier constraints, this is the price that best serves the multi-objective target on this SKU in this store. The system can explain the weights it used. It cannot always express its choice as a one-line rule.

This flavour captures more margin. It is also harder to govern and requires a decision layer rather than just a price engine. The operating model shifts too: the team sets policy, the system handles the routine arbitration.

For the full buyer's-side mental model on this split, the long-form treatment is in rules dressed up, or actual decisions?. Most "what is dynamic pricing" pages collapse these two into one, which is exactly why so many evaluations end up comparing incomparable products.

Who actually does dynamic pricing well in retail

The vendor landscape splits into three archetypes. Naming them descriptively, not head-to-head, is the most useful frame.

Price-optimisation pure-plays. Companies whose core IP is pricing — competitor scraping, elasticity modelling, markdown optimisation. They are deepest on the pricing surface itself.

The trade-off is breadth: pricing decisions stay isolated from the markdown, allocation and replenishment decisions on the same SKU, unless the retailer wires the coordination layer themselves.

ERP and merchandising modules with a pricing tab. The big retail-suite vendors offer pricing as one module among many. The strength is integration: the price change reaches the POS, the e-commerce front and the receipt format without re-entry.

The trade-off is depth — the optimisation logic is usually generations behind the pure-plays, and the rule engine often dominates over genuine decisioning.

Decision-layer-first platforms. Newer entrants build the coordination layer first — the intelligence layer sits across pricing, markdown, replenishment and allocation, with execution handled by the orchestration layer. Pricing is one of several decisions the platform makes.

The trade-off is maturity on the pricing surface specifically — they trade pure-play depth for cross-decision coherence.

None of these is universally "better." The fit depends on what the retailer already has, where the margin is leaking, and how much coordination across decisions matters for the assortment. A retailer with a clean ERP and isolated pricing pain sits in a different position from one whose markdown decisions routinely contradict replenishment on the same SKU.

The pattern we see most often in 2026: a retailer bought a pure-play five years ago and plugged it into the ERP. They are now discovering the pricing surface is only a third of the margin lift available. The other two-thirds live in coordination — across pricing, markdown, allocation, and replen on the same SKU. Adding a second pure-play for each decision multiplies vendors. It does not multiply the value.

That is the pattern the third archetype is built to address. It is not always the right answer — but in 2026, it is the right question.

Five questions a head of pricing should ask in 2026

These cut through demo theatre faster than any RFP scorecard.

1. Is the goal margin, sell-through, or competitive position? A pricing system optimising for margin behaves differently from one optimising for sell-through. "All three" is not an answer — it's the absence of a stated objective. Force the trade-off explicit before evaluating any tool.

2. At what cadence do prices actually need to change? Hourly pricing makes sense for electronics where competitor cycles compress. It is overkill for slow-turn home furniture where weekly is plenty.

The honest cadence answer narrows the vendor list immediately. Buying hourly capability you don't need pays for engineering you won't use.

3. How does pricing coordinate with markdown and allocation decisions? A markdown on a SKU that another team is reallocating in parallel is a self-inflicted loss. If the pricing tool sits in a silo, the coordination cost lives in the team's heads.

This is the question that separates decision-fragmentation pain from genuine cross-decision platforms.

4. Who owns the override governance? Every dynamic pricing system needs a clear answer to "who can freeze, scope-narrow, or override a decision, on what timescale, with what audit trail?". Vendors that don't answer this fluently haven't shipped to a mature retailer recently.

5. What's the bridge from forecast to executable price? A demand forecast is an input to a pricing decision, not the decision itself. The bridge — applying the constraints, the brand rules, the channel-specific rounding, the execution propagation — is where most pricing projects stall. We have argued this at length in from forecasting to decision: why ML isn't enough.

These five questions sort the market faster than any feature checklist. Vendors who handle them fluently are in a different category from vendors who change the subject.

The Solya angle

This article is a definitional anchor, not a Solya pitch. The mental model — yield-management heritage, rules vs decisions, the cadence and coordination questions — works whether you eventually buy from us or not.

Where Solya sits in this landscape: pricing is one of several decisions handled by the same decision layer. That keeps the markdown, replenishment and allocation arbitrations coherent on the same SKU. The architectural choice has its own trade-offs, argued in full in our companion piece on rules vs decisions in dynamic pricing. If you're scoping a 2026 pricing evaluation, that's the deeper read.

The question to take into the next vendor meeting

Before the next dynamic pricing demo, write down — in two sentences — the objective, the cadence, and the coordination cost of your current pricing logic. If you can write it cleanly, you have already done more diligence than most evaluations ever do.

If you can't, the gap isn't the vendor's. It's the absence of a shared definition of what "dynamic pricing" even means for your assortment. That definition is the work this article tried to start.

---

Want a buyer's-side read on your dynamic pricing options?

At Solya, we offer retail leadership teams a 30-minute diagnostic focused on the rules-vs-decisions line in pricing. We look at where your current pricing logic sits today. We name the coordination cost with markdown and allocation, and what changes if you cross from one flavour to the other.

You'll walk away with:

• A clear map of where your current pricing logic is rules-based vs decision-driven
• The categories in your assortment where the rules-vs-decision distinction is most expensive
• The evaluation criteria that separate the two architectures, beyond the marketing labels