Many retailers are not creating demand.
They are renting it with margin.
That is the promotion trap.
In my last article, Stop Pricing Products. Start Allocating Margin., I argued that retailers need to stop treating margin as a static output and start treating it as an enterprise resource.
That argument only becomes real if the operating model changes.
Because here is the uncomfortable truth: many retailers are not using margin deliberately. They are spending it accidentally.
They are giving it away through promotional calendars, reactive competitive matching, disconnected pricing decisions, and organizational routines that were built for a slower market.
Then they call the result growth.
But if sales only move when margin is sacrificed, the business is not growing as much as it thinks.
It is buying demand with its own economics.
The danger is not that promotions exist. Promotions can be useful. The danger is building a commercial model that needs promotions to breathe.
For retailers that want to use margin as a resource, the next shift is clear: stop running pricing and promotions as separate programs. Start orchestrating price as one dynamic enterprise capability.
Dynamic pricing is not chaos. It is discipline.
The False Comfort of Regular Price Plus Promotion
Most retailers still operate with a familiar structure.
A product has a regular price — the everyday or base price customers see outside promotional windows. That price is usually shaped by margin goals, competitive position, category role, vendor dynamics, historical architecture, and sometimes an optimization model.
That word choice matters. Here, “regular price” means the retailer’s own everyday or base price outside promotional windows — not the manufacturer’s list price or MSRP. It contrasts directly with the sale price or promotional price, which is exactly the old model this article challenges.
Then, at specific points in the year, the product enters a promotional calendar. The savings story gets attached: 20% off, save $10, buy one get one, member-only deal, limited-time event.
On the surface, the system feels logical. It gives the business structure. It gives merchants and marketers a calendar. It gives customers a familiar signal of value.
But the model has a hidden cost.
It teaches the organization to think of the regular price as the default and promotion as the lever.
That is where the problem starts.
Because customers do not experience regular price and promotional price as two separate corporate concepts. They experience one thing: the price in front of them when they are deciding whether to buy.
The retailer sees a pricing program and a promotion program. The customer sees a price.
That gap matters.
Promotion Reliance Is Not a Marketing Strategy
The first major problem is promotion reliance.
When a retailer becomes promotion-reliant, revenue may still look healthy. Sales may spike during events. Traffic may increase. Units may move. Vendors may participate. The weekly readout may look good.
But too often, the business has not created demand. It has rented it.
It has borrowed volume from the future. It has trained the customer to wait. It has used margin to manufacture urgency.
Promotions should be a choice. They should not be life support.
A promotion-reliant retailer can find itself keeping stores, systems, inventory, teams, digital platforms, and marketing channels running for 52 weeks while conditioning customers to respond during a handful of promotional moments.
That is an expensive way to operate.
The issue is not simply that discounts reduce margin. Everyone knows that.
The deeper issue is that the organization starts designing around the discount. Everyday pricing becomes less important. Promotional events become the commercial heartbeat. The business gets better at planning deals than understanding value.
Over time, the retailer is not just discounting products. It is discounting its own pricing capability.
Static Prices Create Weak Learning
The second problem is analytical.
If regular prices — the everyday or base prices outside promotional windows — do not move often enough, retailers do not generate enough clean learning.
Elasticity models need variation. They need meaningful price movement and observable customer response across different demand conditions. They need to see what happens when price changes in normal periods, seasonal periods, high-traffic periods, low-traffic periods, competitive periods, and inventory-constrained periods.
If regular prices stay flat for long stretches, the model is forced to infer too much from too little.
That is when retailers get into trouble.
A model may overestimate elasticity. It may underestimate elasticity. It may suggest that a price increase can be taken without meaningful unit risk. It may miss the point where the customer starts to behave differently.
Then the recommendation goes into market, performance disappoints, and people say, “The model was wrong.”
Sometimes the model was not the root problem. The operating model failed to create the data the model needed.
That is a critical distinction. Better algorithms can help, but no algorithm can fully rescue a weak learning system.
Promotional Data Is Not as Clean as It Looks
A common response is: “We do have price variation. We run promotions.”
But promotional prices are rarely pure price signals.
Promotional periods often come with email campaigns, app placement, loyalty targeting, circular visibility, endcaps, vendor funding, seasonal traffic, urgency messaging, associate focus, and competitive response.
So when a retailer blends regular-price data and promotional-price data and calls the result elasticity, it may not be measuring true price response.
It may be measuring price plus media. Price plus placement. Price plus seasonality. Price plus urgency. Price plus customer conditioning.
That is not a clean elasticity signal.
It is a confounded commercial event.
And when that signal becomes the foundation for future pricing decisions, the retailer can end up optimizing against noise.
This is one of the quietest but most damaging consequences of the traditional promotion model: it gives the illusion of learning while muddying the signal.
The business thinks it is getting smarter.
In reality, it may simply be getting more confident in a distorted read of customer behavior.
Competitor Matching Is Not a Strategy Either
Many retailers try to compensate by anchoring prices to competitors.
Competitive intelligence matters. It should absolutely be part of the pricing system.
But competitor price cannot become the strategy.
When a retailer pegs itself to a competitor, it is implicitly assuming that every other part of the customer decision is comparable: brand strength, convenience, assortment, service, store experience, delivery promise, loyalty program, digital experience, trust, and habit.
That is rarely true.
Two retailers can sell the same product at the same price and produce very different outcomes because customers are not choosing price alone. They are choosing a total proposition.
A competitor’s price should be a feature in the decision system.
It should not become the decision system.
If the only pricing intelligence a retailer trusts is someone else’s price, then it does not really have a pricing strategy. It has a reaction function.
Pricing and Promotions Are the Same Problem in Disguise
The deeper issue is organizational.
Most retailers staff pricing and promotions as separate teams. They create separate processes, calendars, analytics, approval flows, dashboards, and success metrics.
The pricing team may be measured on margin rate, price index, competitiveness, or architecture.
The promotion team may be measured on lift, traffic, vendor participation, event performance, or promotional revenue.
Those objectives can easily conflict.
One team protects margin. Another creates demand. One manages the base. Another manages the exception. One sets the regular or everyday price. Another overrides it.
But the customer does not see the org chart.
The customer sees the price.
That is why the future cannot be a better pricing program sitting beside a better promotion program.
The future is one integrated pricing capability with one shared mandate: determine the best price at each point in time, under the right constraints, for the right business objective.
Dynamic Pricing Is Not Chaos. It Is Discipline.
Dynamic pricing is not chaos. It is discipline.
Dynamic pricing is often misunderstood.
It does not have to mean changing every price every hour. It does not mean copying Amazon blindly. It does not mean removing human judgment. It does not mean creating customer confusion.
Dynamic pricing means treating price as a time-aware decision.
Instead of setting a regular price and occasionally discounting from it, the retailer defines a planning horizon and solves for the optimal price path across that horizon.
The unit of time could be a week. For some categories, it may be a day. For others, it may be a month. The right cadence depends on the category, customer mission, operational constraints, competitive environment, and the retailer’s ability to execute price changes across channels.
The important shift is this: regular price stops being a resting place.
It becomes one point on a price path.
That is the language of the future state. Not a permanent regular price with occasional exceptions, but a managed price path that reflects demand, inventory, competition, customer trust, and strategic intent over time.
The retailer is no longer asking, “What should the regular price be, and when should we discount it?”
It is asking, “What price should this product carry in each period over the next 52 weeks so we can maximize total margin and sales while respecting inventory, category strategy, customer trust, competitive guardrails, and financial goals?”
That is a better question.
It recognizes that not every week plays the same role. Not every category needs the same level of investment. Not every product should be harvested or funded at the same time.
Demand conditions change. Customer urgency changes. Inventory positions change. Competitive intensity changes. Seasons change.
Pricing should change with those realities.
The Forecasting Requirement
To make this work, retailers need stronger multi-period forecasting.
A dynamic pricing system requires the ability to forecast demand, units, margin, and customer response across multiple future periods. Traditional approaches can support parts of this. ARIMA-style models, gradient boosting models, and other machine learning methods can be adapted for multi-step forecasting. More advanced sequence-data models, including neural-network-based approaches, can forecast across time horizons more naturally.
But the algorithm is not the point.
The capability is the point.
Retailers need to understand how demand is likely to behave across time, how customers respond to price under different conditions, and how today’s price decision affects future outcomes.
That last part is critical.
A price decision this week can pull demand forward. It can train customers to wait. It can clear inventory. It can protect margin. It can create a new reference price. It can shape future promotional effectiveness.
This is why single-period optimization is not enough.
Retailers need to stop optimizing isolated moments and start optimizing commercial trajectories.
Dynamic Pricing Builds a Better Learning System
Dynamic pricing also creates a major data advantage.
When prices are static for long periods and then interrupted by promotions, the retailer is always trying to learn from imperfect history. It has limited price variation, limited clean response, and too many confounding factors during promotional periods.
Dynamic pricing changes the learning loop.
It builds pricing experimentation into the operating rhythm of the business.
Over time, the retailer creates a richer dataset of price points, demand conditions, customer responses, category roles, competitive contexts, and margin outcomes.
That data is a strategic asset.
It helps improve elasticity estimates. It helps separate true price response from promotional noise. It creates a stronger feedback loop between model recommendations and market behavior. And it makes the pricing engine smarter over time.
Dynamic pricing is not only a better decision system. It is a better learning system.
That may be one of the most underappreciated benefits. The retailer is not just choosing better prices. It is building the intelligence to choose better prices in the future.
The Process Shift Is Non-Negotiable
This is where many retailers underestimate the transformation.
Dynamic pricing is not an analytics project with a business case attached. It is an operating-model redesign.
Price recommendations must flow seamlessly from models into the systems that execute prices across channels. Digital channels need to update prices quickly, consistently, and accurately. Store channels need item files, point-of-sale systems, shelf labels, signage, and customer communication to support the new cadence.
For many brick-and-mortar retailers, electronic shelf labels will become increasingly important.
Not because they are a shiny technology.
Because they are part of the price execution layer.
Retail leaders need to stop telling themselves that dynamic pricing is only possible for e-commerce-first players.
That is an excuse.
Yes, the transformation is harder for store-based retailers. The complexity is real. Legacy systems, store labor models, governance routines, and operational constraints will create friction.
But the answer cannot be to keep running a slow pricing model in a market where customers, competitors, costs, and channels move much faster.
If a retailer wants margin to become a true resource for growth, the pricing execution layer has to modernize.
Governance Should Be Guardrails, Not a Parking Lot
Dynamic pricing does not eliminate human judgment.
It makes governance more important.
Retailers need business review processes that ensure model recommendations align with strategy, customer trust, brand positioning, competitive posture, legal requirements, and financial goals.
But governance cannot become the place where dynamic pricing goes to die.
A model that produces recommendations quickly, only for those recommendations to sit in manual review queues for weeks, is not a dynamic pricing capability. It is an analytics prototype trapped inside an old operating model.
The role of governance is to define decision rights, constraints, escalation rules, approval thresholds, exception handling, and performance monitoring.
Humans should shape the system. Humans should set the boundaries. Humans should intervene when context requires it.
But humans should not have to manually re-litigate every price decision.
That is not control. That is friction.
Marketing Cannot Be the Brake
Pricing does not live only in pricing systems.
It also lives in the customer’s mind.
That means marketing channels have to support dynamic pricing. If the organization still needs long lead times to create, approve, print, schedule, and distribute every savings story, the pricing model will always be constrained by the marketing calendar.
This is another reason the traditional promotion model is so limiting.
It forces the business into pre-planned events, fixed savings claims, and long-cycle execution.
Dynamic pricing requires more flexible customer communication.
The savings story does not disappear. It evolves.
Instead of always saying “20% off regular price,” a retailer can communicate value in more dynamic ways:
- Lowest price in 52 weeks
- Best price this season
- Member price today
- Price drop on top-rated essentials
- Limited-time value on high-demand items
The customer still receives a value signal.
But the retailer is no longer trapped in the artificial structure of regular price versus promotion. The business keeps control of the broader price path.
The goal is not to eliminate the customer’s sense of savings.
The goal is to stop letting the promotional calendar dictate the economics of the business.
One Pricing Capability
The organizational implication is clear.
Retailers do not need a pricing program and a promotion program operating as separate systems.
They need one pricing capability.
That capability should bring together pricing strategy, promotional strategy, forecasting, optimization, marketing communication, category management, finance, store operations, digital execution, and customer analytics.
The shared question should be simple:
What is the best price for this product, in this channel, at this moment, given the role it plays in the broader business?
Sometimes the answer is to harvest margin.
Sometimes it is to invest margin.
Sometimes it is to stimulate demand.
Sometimes it is to protect price perception.
Sometimes it is to clear inventory.
Sometimes it is to defend against a competitor.
Sometimes it is to support a broader category, loyalty, or customer strategy.
But those should not be separate decisions made by disconnected teams.
They should be coordinated decisions made through one integrated commercial decisioning layer.
This Is How Margin Becomes a Resource
The first article in this series, Stop Pricing Products. Start Allocating Margin., argued that retailers should stop pricing products and start allocating margin.
This article is the operating-model answer to that thesis.
You cannot truly allocate margin if pricing and promotions are managed separately.
You cannot use margin as a growth resource if the business only thinks in terms of regular prices and temporary discounts instead of dynamic price paths.
You cannot build a long-term margin strategy if the organization is structurally dependent on promotional events to drive demand.
And you cannot modernize commercial decision-making if teams, systems, data, models, governance, and marketing channels are still designed for a slower world.
Dynamic pricing is not just a pricing tactic.
It is the mechanism that allows retailers to decide when to harvest margin, when to invest it, where to protect it, and where to redeploy it for growth.
That is the shift.
Stop running promotions as exceptions to pricing.
Start orchestrating price as a dynamic enterprise capability.
Dynamic pricing is not chaos. It is discipline.
Margin is not simply what is left after the sale. Margin is the fuel. And pricing is how you decide where that fuel goes.
If you are a retail, analytics, data, merchandising, loyalty, or commercial strategy leader thinking about how AI should improve enterprise decision-making, I would welcome the conversation. Start a conversation, explore Speaking & Advisory for executive briefings and leadership sessions, connect with me on LinkedIn, or subscribe for future insights on AI, advanced analytics, and commercial decisioning.
Views are my own and do not represent my employer.