The End of the Sticker Price

FOR over a century, consumers have enjoyed a position of relative privilege in the free market. Privacy and opacity shielded their true willingness to pay – the highest price a particular person would hand over rather than walk away. Today, consumers are no longer protected and anonymous. On the contrary, the dynamic has completely flipped. Sellers use internet browser cookies, location, artificial intelligence, machine learning and consumer data mining to predict and calculate what each individual will bear. The algorithm has many data points and is still imperfect, so a gap in knowledge remains between the two parties. But personalised pricing does not close that gap so much as reverse it, shifting the balance of power decisively to the seller.

Individual price quotes of this kind existed for most of the last century only in theory. Arthur Pigou, writing in 1920, called it perfect price discrimination: charging every buyer precisely what they will bear. Algorithmic pricing has made it something closer to practical. At first glance it looks like an economist’s dream, increasing efficiency and eliminating deadweight loss – the value destroyed when a sale both sides would have accepted never happens because the posted price sat between them. On closer inspection, the shadow of monopolistic extraction emerges. I argue that the net effect of personalised pricing is negative, because it eliminates consumer surplus directly into corporate balance sheets, redistributes wealth away from household purchasing power, and deepens inequality.

How far this has already happened is contested, and worth stating plainly. When the United States Federal Trade Commission examined six firms that build pricing systems for retailers, in a study published in January 2025, it found that a shopper’s location, browser history, demographics, mouse movements and even an abandoned basket were all being fed into the price they were shown; those six intermediaries served at least 250 clients between them, from grocers to clothing retailers.

Against that, a survey of the European evidence prepared for the OECD in 2018 found personalised pricing still rare and the gaps small – prices differed in only about 6% of identical-product comparisons, by a median of under 2%. The machinery exists and is being sold. Whether it is yet used at full strength is the open question, and what follows is about what happens when it is.

The Argument for Efficiency

Technology-infused pricing habitually captures willingness to pay, and so facilitates a steady transfer of wealth from consumer households to corporate balance sheets. Classic welfare theory emphasises the inefficiency of deadweight loss, because consumers whose willingness to pay falls below the market price simply do not buy.

Algorithmic pricing expands the base of people who do buy, since each person’s willingness to pay is matched in the price they are offered, and many of the additional buyers are in lower income brackets or especially price-conscious. As Hal Varian observed in his survey of price discrimination, when different people pay different prices for the same product, total market output grows without sacrificing the yield from consumers willing to pay more.

Airline pricing is the practical example. Historically, if a passenger could not afford the fare, they did not fly: the consumer did not reach their destination and the airline flew the seat empty. Today fares are optimised for a range of consumer profiles, which from the perspective of the free market counts as a success.

There is an alternate perspective. Eliminating deadweight loss also reduces the ability of households to save, so the increase in market output comes at the expense of consumer welfare. The arithmetic is unforgiving: consumer surplus is what a buyer was willing to pay minus what they actually paid, so a price set exactly at willingness to pay leaves a consumer surplus of nothing at all.

The Impact on Consumer Purchasing Power

In the era of algorithmic pricing, corporations benefit entirely from the elimination of deadweight loss. What was once lost value is transferred wholesale into producer surplus, the seller’s side of the same ledger. Historically, consumer surplus contributed to household savings. With that surplus eliminated, a single unexpected setback can leave even a high-earning household in a difficult position. The nest egg disappears, and consumer liquidity – the ability to lay hands on money quickly when it is needed – decreases.

This matters for more than the household. Consumers have a higher marginal propensity to consume than firms do – they spend rather than hold a larger share of each additional pound, a relationship at the centre of Keynes’s account of demand – so consumer surplus flows back into the economy. Corporations are less likely to recirculate captured producer surplus, investing instead in reserve pools and share buybacks. With money pooled in corporate hands rather than household ones, capital moves through the market more slowly.

The Effect on Societal Inequality

Weaponising consumer data is problematic in itself, but there are scenarios in which it is particularly unfair. At first glance, algorithmic pricing may look like an opportunity to tax the rich. I argue that it is regressive rather than progressive. The exploitation falls hardest on consumers with inelastic demand, or caught in high-risk situations. Inelasticity varies by circumstance and is worsened by low digital readiness and by search friction – the effort it takes to find another seller.

Search friction plays out as follows. When a retailer or data aggregator identifies someone searching for an essential good from a residential address with few alternatives nearby, it recognises that their price elasticity of demand is approaching zero. Elasticity means the percentage change in the quantity someone buys divided by the percentage change in price; when that figure falls below one, demand is inelastic and the buyer keeps buying even as the price climbs. An algorithm that can spot such a buyer has every incentive to push.

Digital readiness serves as a protective layer, because consumers can use virtual private networks and incognito browsing to slip out from under the algorithmic lens – but that readiness is a privilege some low-income and less-educated consumers do not possess. The model can also exploit the wealthy in vulnerable moments, charging the maximum in every situation of inelastic demand including the most desperate.

This is precisely the mechanism by which, as Thomas Piketty and Joseph Stiglitz have each argued in different registers, everyday transactions harden into structural inequality.

The Impact on Markets and Competition

The consequences of algorithmic pricing are not restricted to individual consumers. Over time, intelligent pricing models alter the fabric of the competitive market itself. When personalised pricing becomes widespread, economists envision two opposing trajectories: poaching or entrenchment.

As Mark Armstrong has argued, personalisation can increase competition, because several firms can each offer targeted discounts to poach customers loyal to a rival while charging their own loyal customers more; Varian’s work suggests this can lower average prices and restore some surplus to consumers. The result is a version of the prisoner’s dilemma: uniform pricing across the market yields stable profits, but a single entrant that personalises can extract a disproportionate share of producer surplus.

This poaching matrix assumes symmetric capabilities among a wide field of players. The long-term structural reality shifts towards entrenchment instead. In practice, only a few producers have the technological foundation to implement algorithmic pricing successfully: it requires data warehouses, computational power, hardware capable of supporting machine learning and the infrastructure to run artificial intelligence models.

Large incumbents have all of that. Nascent start-ups may not be able to compete with a moat of that depth, as Lina Khan argued in her account of Amazon’s market power. Consumers are left in a stagnant, oligopolistic market that lacks healthy competition and innovative breakthroughs, and where a few players own the data, consumers may fall victim to predatory pricing.

In a conventionally priced economy, start-ups can disrupt their vertical by entering the market and earning customer loyalty – the contestability that William Baumol argued was the real discipline on incumbent behaviour. When personalised pricing is supported by large data moats, start-ups may go out of business rather than grow.

In economic theory, an incumbent had to cut prices across the board to price out a new entrant, which was expensive enough to deter it. With algorithmic pricing, it need only price out the particular consumers looking to shop elsewhere. This is how algorithmic pricing circumvents the standard guardrails of a competitive market, and innovation stagnates when large data-powered firms channel their energy into extracting revenue through pricing rather than through research and development.

The Trickle-Down Fallacy

Free-market enthusiasts may claim that capturing producer surplus instead of destroying it as deadweight loss is a success story, because the captured surplus will eventually trickle down from corporations back to customers. Competition will force costs to pass through; lower prices will widen access for marginal buyers; corporate reinvestment will organically improve wages, arriving as corporate tax, new jobs, dividends or share appreciation in pension accounts.

On this view, algorithmic pricing is a redistribution of wealth rather than a reinstatement of social inequality, and classical economists have long seen perfect price discrimination as a feasible way of reallocating capital.

In reality, algorithmic personalisation bypasses those channels. Personalised pricing requires vast data moats, so the downward competitive pressure on price never materialises. Low-willingness-to-pay consumers are technically new entrants at lower price points, but their discount yields them no surplus at all.

Many low and middle-income households own no equity, so even if corporations intended to redistribute their gains the route would be regressive and would miss most of the consumers the pricing had extracted from. Digital-first companies are in any case likelier to spend excess gains on technology than on staff.

In the twentieth century, the success of large companies was shared with a larger workforce; in the twenty-first, as Erik Brynjolfsson and Andrew McAfee have documented, profitability can be achieved without one. Corporate leadership benefits personally from bloated balance sheets while the trickle-down effect remains a theoretical dream and a practical fallacy.

Conclusion

As of 2026, personalised pricing carries high stakes for consumer welfare worldwide, and the law is only now beginning to reach it. In August 2026 the United States Federal Trade Commission proposed an enforcement policy statement warning that a firm which implies a price is fixed when it in fact varies by individual, or which uses personal data to set prices without saying so, risks acting deceptively under the Federal Trade Commission Act. That is a disclosure remedy, and disclosure does not undo the transfer of surplus; it only means the consumer knows it is happening.

On paper, increased market output with no deadweight loss is an economic success story. In practice, this market structure is far from healthy. When efficiency, extraction and exploitation go hand in hand, there is a serious paradox at play.

Algorithmic pricing flies under the radar of the watchdogs because it enlarges the overall consumer base rather than restricting any segment of it: by the standards Robert Bork set for antitrust half a century ago, a market can still look competitive while a firm extracts the maximum from every buyer in it.

Regulatory policy needs to place its guardrails at the earliest stage of the pricing pipeline: the data sources themselves. And it is worth reconsidering our definition of success as the maximisation of market output. Should economic success be measured by efficiency or by equity? The free market was championed as a paradise for consumers. It is rapidly becoming an obstacle course of corporate monetisation.