Key Takeaways
- Surveillance pricing gives businesses in Australia new ways to use customer data, but raises transparency concerns.
- Dynamic pricing, concealment, and surveillance create different forms of information asymmetry.
- Ethical fading can make questionable pricing decisions feel like routine optimisation.
- Australian businesses need stronger pricing governance as regulatory scrutiny increases.
The Amazon case isn’t a one-off. It’s the sharpest edge of a pattern already taking shape in Australian supermarkets, and one our own regulator has just named as a priority. Here’s the psychology behind why smart, careful businesses keep walking into it, and what it means if you’re running one here.
I’ve spent sixteen years watching pricing decisions get made, and I’m watching something new take shape right now: three different mechanisms, converging on the same idea, that a business can see more about what you’ll pay than you can see about how that price was set. Not one trend. Three, and they’re moving fast enough that regulators are now naming them separately.
Read This CEO Pricing Strategy To Improve Margin & EBIT
Three mechanisms, one pattern
The first is extraction. Woolworths has installed roughly 17 million electronic shelf labels across more than 600 Australian stores, with more than 170 stores in New Zealand. Coles is trialling the technology too.
The technology itself is neutral. A digital tag is simply a faster way of changing a displayed price. But speed changes what’s commercially possible. A price that once took hours of labour to change can now move almost instantly in response to demand, inventory, time of day or other commercial signals. Former ACCC chair Alan Fels has said plainly that this is exactly what the technology now makes possible.
Surveillance pricing in Australia is not inherently problematic.
Dynamic pricing has legitimate uses everywhere from airlines to energy. The question worth asking is what a business decides to optimise for once changing the price becomes almost frictionless: how much more can we charge given the commercial conditions? That’s extraction. It’s the mildest of the three, and the easiest to defend, provided the business is honest that it’s happening.
The second is concealment, and it’s the Amazon case readers of this newsletter already know. The FTC and 22 states allege Amazon secretly inserted a “soft reserve price” into its advertising auctions from 2019, telling more than a million advertisers their price was set by open competitive bidding when, the complaint alleges, it wasn’t. The FTC puts the extraction at more than $20 billion. Amazon denies the allegations and is contesting the case. The underlying question here is different to extraction: how much can we change the mechanism without the customer understanding what changed? That’s concealment, and it’s the most serious of the three, because unlike extraction, it can’t survive being disclosed. The mechanism only works if the customer believes something is true that isn’t.
The third is surveillance, and the evidence behind it is arguably the strongest of the three. The FTC’s January 2025 study found pricing intermediaries used precise location, browsing history, shopping activity, and online cart data. These factors shaped individualised prices and offers.
The intermediaries it examined worked with at least 250 clients across industries.
The question underneath surveillance pricing is the sharpest of the three: how much can we infer about this customer’s willingness to pay?
What if customers don’t realise that information influences their price?
Put together, that’s a real taxonomy, not just three examples of the same thing.
And it produces a governance question that matters more than any of the three individually: the technology isn’t the ethical decision. What you allow the technology to know, infer and act on is.
These aren’t the same thing, and it’s worth being precise about that. Dynamic pricing can be entirely legitimate. Personalised pricing can create genuine customer value. Data can make pricing considerably better than a flat, one-size-fits-all number ever could. The problem begins when the commercial advantage depends on an asymmetry the customer wouldn’t accept if they understood it. And that raises a harder question than “is this legal”: how do sophisticated, well-run businesses cross that line without ever noticing themselves doing it?
Why smart businesses in Australia build unethical surveillance pricing without meaning to
Here’s what interests me more than deciding whether these are stories about good actors or bad ones. You don’t need consciously unethical people to build an ethically questionable pricing system.
Researchers Ann Tenbrunsel and David Messick studied this gap between unethical behavior and ethical self-perception. They found a connecting idea called ethical fading.
The ethical dimension of a decision doesn’t get consciously overridden. It fades from view entirely, and their research, along with the pattern I keep seeing inside pricing teams, points to a few recurring ways that happens. One is language: nobody approves “a system to charge people more than they’d agree to if they understood it.” They approve “personalisation,” “dynamic pricing,” “demand-responsive optimisation,” language that’s technically accurate and morally weightless.
Another is incrementalism: a business starts with genuinely reasonable dynamic pricing tied to supply and demand. Each subsequent refinement, adding a location signal, browsing history signal, or cart-abandonment signal, looks like a small, defensible step.
It never looks like the leap it actually represents from where the business started.
A third is simple repetition: once a team has justified the first data-driven pricing decision, the fortieth one doesn’t get re-examined with fresh eyes. It gets waved through by the momentum of the thirty-nine before it.
An organisation doesn’t necessarily encounter the decision labelled as an ethical one at all. It arrives as a KPI improvement, a conversion initiative, or an optimisation project, and by the time anyone asks the harder question, the mechanism is already live.
What does this mean if you’re a business in Australia thinking of surveillance pricing
Here’s the part I want every Australian CEO reading this to sit with. And if you’re wondering whether this is really coming here, the government already answered that question. In July 2026, it specifically named “retail surveillance pricing” as an AI consumer risk it intends to examine under Australian Consumer Law, alongside agentic commerce, as part of a broader look at automated decision-making. That’s not a think tank raising a concern. That’s the government of the country you operate in telling you what it intends to look at next.
Australian law is already moving on this from another direction too. From 1 July 2026, a new excessive-pricing prohibition applies to very large supermarket retailers. Currently, Coles and Woolworths fall under this regime.
It assesses prices against factors including cost of supply and a reasonable margin. The maximum penalty is the greater of $10 million, three times the benefit obtained, or 10 per cent of annual turnover.
Separately, under the general Australian Consumer Law, misleading conduct about pricing carries its own significant penalty exposure. I won’t turn this into a legal briefing because the two regimes differ, and your lawyer can tell you which one applies to you. But both point in the same direction: regulators are showing less tolerance for pricing mechanisms that customers don’t understand.
〉〉〉 Get Your FREE Pricing Audit 〉〉〉
Woolworths and Coles are not hypothetical examples either.
The ACCC alleges, in ongoing and unresolved Federal Court proceedings, that both misled customers over “Prices Dropped” and “Down Down” promotions. Claims both companies are contesting.
Separately, and without suggesting that either company has done anything wrong, both companies are deciding how to use their electronic shelf labels.
If you run pricing for an Australian business of any size, you’re operating in the same regulatory environment they are. With the same euphemisms available to make an extraction decision feel like an optimisation decision.
So here’s the proactive version of this, not the defensive one. Before your business adopts the next piece of pricing technology, ask what data it uses, and then ask the harder question. Could you explain that use to the customer it applies to, in plain language, without it sounding like a confession? If the answer is yes, you probably have a mechanism capable of surviving scrutiny. If the answer makes you reach for a euphemism, stop there. That may be ethical fading doing its work in real time.
I’m going deeper on this in the next podcast: three pricing mechanisms, the psychology that allows them to develop. And what Australian CEOs should be asking before the technology gets ahead of the governance.
If you’re already grappling with where that line sits in your own pricing, message me directly.
Read This CEO Strategy for Surveillance Pricing in Australia, Increasing Prices & Improving EBIT
Are you a business in need of help aligning your pricing strategy, people, and operations to manage surveillance pricing in Australia and deliver an immediate impact on profit?
If so, please call (+61) 2 9000 1115.
You can also email us at team@taylorwells.com.au if you have any further questions.
