In the summer of 2012, the Wall Street Journal reported that the travel booking website Orbitz had, in some cases, been suggesting to Apple users hotel rooms that cost more per night than those it was showing to Windows users. The company found that people who used Mac computers spent as much as 30 percent more a night on hotels. It was one of the first high-profile instances where the predictive capabilities of algorithms were shown to impact consumer-facing prices.

Since then, the pool of data available to corporations about each of us (the information we’ve either volunteered or that can be inferred from our web browsing and buying histories) has expanded significantly, helping companies build ever more precise purchaser profiles. Personalized pricing is now widespread, even if many consumers are only just realizing what it is. Recently, other algorithm-driven pricing models, like Uber’s surge or Ticketmaster’s dynamic pricing for concerts, have surprised users and fans. In the past few months, dynamic pricing—which is based on factors such as quantity—has pushed up prices of some concert tickets even before they hit the resale market, including for artists like Drake and Taylor Swift. And while personalized pricing is slightly different, these examples of computer-driven pricing have spawned headlines and social media posts that reflect a growing frustration with data’s role in how prices are dictated.

The marketplace is said to be a realm of assumed fairness, dictated by the rules of competition, an objective environment where one consumer is the same as any other. But this idea is being undermined by the same opaque and confusing programmatic data profiling that’s slowly encroaching on other parts of our lives—the algorithms. The Canadian government is currently considering new consumer-protection regulations, including what to do to control algorithm-based pricing. While strict market regulation is considered by some to be a political risk, another solution may exist—not at the point of sale but at the point where our data is gathered in the first place.

In theory, pricing algorithms aren’t necessarily bad. Prices that are more responsive to market forces and beyond human intervention could make some of the buying process more efficient and tailored to individual buyers. It might also result in pricing being clearer. And, technically speaking, by using data profiling to accurately assess a consumer’s willingness to pay (that is, the maximum amount someone would spend on something), personalized pricing could meet people where they are financially, creating more opportunity for some people to buy things they might otherwise not have been able to afford.

Again, that’s in theory. In practice, it sometimes works differently. Strange things can happen on the way to maximizing profits (the goal of pricing algorithms writ large). Researchers concluded in 2019 that “relatively simple pricing algorithms systematically learn to play collusive strategies” as they adjust to constantly meet changes made by others—even if they’re not designed to do so, nor are able to communicate with other pricing algorithms. Researchers noted in 2021 that, under this scenario, “the largest gains accrue to a dominant firm with the most advanced technology and the largest market share.”

A 2016 study that tracked for four months the top twenty Amazon sellers of over 1,600 products came to a similar conclusion. In some cases, algorithms were changing the prices of items “tens or even hundreds of times per day” (a frequency difficult for a human to replicate), creating “a largely winner-take-all marketplace.” The study showed these same sellers received more positive feedback, which put them at an advantage when it came to page ranks on Amazon. In other words, a site that appears to offer vast choices may, when driven by algorithms, end up offering only limited options from a few top sellers. Algorithms may also make things more expensive overall, as lowering prices might simply prompt competitors to undercut, decreasing the incentive for anyone to drop prices. If you feel like you’re spending more for things all the time, it might be because you are.

There is also a concern regarding biases inherent in big data. In 2015, ProPublica revealed that prices for the Princeton Review’s online SAT tutoring packages varied depending on US ZIP codes, creating the “unexpected effect . . . that Asians [were] almost twice as likely to be offered a higher price than non-Asians.”

All of which makes regulation necessary but simultaneously difficult—particularly if the issue is addressed at the point of sale. For example, a government could implement price controls for consumer goods, ensuring that we never pay more than a set rate (an idea that’s also been discussed recently as a temporary solution to rising grocery prices). But its implementation is not the norm, and its history would make it a hard sell politically. In fact, market regulation of any kind is likely to spark a backlash both from businesses that use tools like personalized pricing to increase profits and from opposition parties looking for ideological leverage. Given the recent history of the government’s update to the country’s broadcasting act—which also had at its core a question about the commercial value of data and which created a hyperactive discourse about censorship—a foray into personalized pricing regulation could prove too perilous to attempt.

The 2022 federal budget bill made updates to Canada’s Competition Act, but those amendments didn’t deal with algorithmic pricing directly, and a discussion paper released by Innovation, Science and Economic Development Canada noted that the act needs improvement due to “[t]he new challenges posed by how data-driven and digital markets operate.” There are “valid reasons to limit grounds for intervention in private commerce,” the paper went on but conceded that the question of how is growing increasingly complicated. “The public interest is not well-served if competitive harm is identifiable but the [Competition] Bureau is not sufficiently empowered to intervene.”

But there may be a way to limit the harms of algorithmic pricing other than by direct intervention against a seller. To get to the root of the issue, policy might be aimed much earlier in the shopping process and could focus on what personal data is available to sellers in the first place.

When it comes to our data, consumers are getting a raw deal in many instances, says Pascale Chapdelaine, associate professor at the University of Windsor Faculty of Law, who specializes in privacy and copyright and e-commerce law. For example, the idea of consent around personalized pricing “is very dubious,” she says, as it’s almost never explicitly brought to shoppers’ attention before checkout. Just by virtue of our being online, it’s possible that information about us is gathered without our informed or explicit opt-in and is used to dictate how much we pay at checkout. “How can you consent to something you don’t even understand in advance?” In many instances, the access we unknowingly grant to our personal data is “disproportionate . . . to what we’re going to get out of this” as consumers, Chapdelaine says.

One way to begin addressing the problem might be more transparency in stating that prices are being personalized by algorithms. But more transparency might simply raise awareness without fixing anything. Besides, people already know the computers are up to something. Instead, Chapdelaine advocates for legislating limits to the use of personal data that could be used to personalize prices, along with more enforcement from the courts and privacy commissioner.

For the government, the decision to curtail technology that businesses know can improve profit margins—particularly as many retailers and services are still recovering from pandemic losses—could be a risk, both politically and practically. But having been accused of losing touch with ordinary people—and specifically the bills they pay—the government may find that the benefits of tackling algorithmic pricing and introducing more fairness into our online marketplace outweigh the costs.

Colin Horgan
Colin Horgan is a writer and communications professional in Toronto.