From Maryland to Molslinjen
Where retail pricing crosses the line
Retailers have spent years investing in better data, better targeting, and more personalization.
Now a new question is emerging:
When does better pricing become the wrong kind of pricing?
That question is no longer theoretical.
What “surveillance pricing” actually means
Surveillance pricing is simple in principle:
Using data about an individual customer to determine the price they see.
That data can include
- Browsing behavior
- Purchase history
- Location or device
- Inferred willingness to pay
In practice, it means two customers could see different prices for the same product—not because of timing or demand, but because the system predicts one is willing to pay more.
That’s a very different model from traditional dynamic pricing.
Why Maryland matters
In the U.S., the state of Maryland is moving to restrict this exact practice.
Lawmakers are targeting pricing models that use personal data signals to set individualized prices. The goal is to prevent situations where consumers are effectively charged based on who they are—or what the system believes about them.
This is not a broad attack on dynamic pricing.
It’s a targeted move against price discrimination driven by personal profiling.
That distinction is critical:
- Adjusting prices based on demand → widely accepted
- Adjusting prices based on the individual customer → increasingly regulated
And this is how regulation typically starts at state level, focused on a specific use case, before spreading.
For retailers, this is an early warning not a distant scenario.
A contrasting case: Molslinjen
At the same time, companies like Molslinjen in Denmark are being highlighted as successful examples of AI-driven pricing.
They use machine learning to forecast demand and adjust prices across ferry departures. Prices change frequently, sometimes significantly.
The system looks at:
- Departure times
- Booking curves
- Seasonal demand
- Traffic patterns
- External conditions
Everyone looking at the same departure at the same time sees the same price.
This is still highly dynamic. It’s still data-driven.
But it operates on market signals, not personal profiles.
Where the real line goes
Many retailers group all of this under “AI pricing.” That’s where confusion starts.
Because in reality, there are two fundamentally different approaches:
Person-based pricing
- Prices influenced by individual-level data
- Different customers, different prices at the same moment
- High regulatory and trust risk
Context-based pricing
- Prices driven by demand, timing, inventory, external signals
- One price per context, visible to all
- Operationally complex, but widely accepted
The technology behind them can look similar. The inputs and the implications are not.
What retailers should be paying attention to
Most retailers are not deliberately building surveillance pricing systems.
But many are moving in that direction indirectly—because personalization, data, and pricing are starting to overlap.
A few questions worth asking internally:
What data is allowed into your pricing models?
Is it strictly commercial signals or does customer-level data play a role?
Can two customers see different prices at the same time?
If yes, what drives that difference?
Can you clearly explain your pricing logic?
If it’s difficult to explain, it will be difficult to defend.
Are pricing and personalization systems separated or quietly converging?
This is where most risk sits: not in intent, but in architecture.
What capabilities actually matter now
If the safe path is context-driven pricing, then execution becomes the real challenge. Molslinjen’s model points to what’s required:
01
High-frequency demand sensing
Understanding demand in near real time not just from historical sales.
02
Granular forecasting
At the level where pricing decisions are actually made (SKU, route, time window).
03
Fast price execution
The ability to update prices consistently across channels without friction.
04
Clear data governance
A defined boundary around what data is used and what is explicitly excluded.
Most retailers don’t lack pricing ideas.
They lack the systems to apply them precisely and consistently.
The takeaway
Retail pricing is starting to split into two paths:
- Context-driven pricing, which is scaling and becoming more sophisticated
- Person-driven pricing, which is attracting regulatory attention
The difference isn’t technical. It’s philosophical—and increasingly legal.
Retailers who understand that distinction early will have an advantage. Not just in compliance, but in trust.
Because pricing is one of the most visible decisions you make.
And customers don’t just react to the price.
They react to whether it feels fair.