Questions worth asking.

It’s when companies use software (often machine learning) to adjust prices in real time based on supply, demand, what competitors are doing, and data about you specifically. Prices aren’t really fixed anymore. They get calculated for you in the moment you look, based on everything the company knows or can infer about who you are.

They get used interchangeably most of the time. Technically, dynamic pricing refers to any system where the price moves in response to market conditions, while algorithmic pricing emphasizes that software (not humans) is making the adjustments. Personalized pricing is a subset of both: it’s specifically when the algorithm is pricing you individually rather than pricing the market as a whole.

In the United States, mostly yes. Federal law prohibits pricing based on protected characteristics (race, religion, sex, national origin, disability), but it does not prohibit pricing based on your device, location, browsing history, or inferred behavior. The EU is moving toward stricter disclosure rules through the Digital Services Act. The FTC has opened multiple inquiries but has not set firm nationwide rules as of 2026.

Usually you can’t tell directly, which is kind of the point. The reliable way to check is to compare the same product across incognito mode, a different device, a VPN, or a different account. If the price shifts meaningfully between any of those, you’ve been personalized. Tools like the Mozilla Price Observatory and some browser extensions can help automate the comparisons.

Somewhat. Incognito blocks persistent cookies, so sites can’t recognize you as a returning visitor. But it doesn’t hide your IP address, your device fingerprint, your general location, or your login state if you sign in. It’s a real tool, not a silver bullet. For serious testing, combine incognito with a different device or a VPN.

Travel (airlines, hotels, rideshare) is the most aggressive because those industries have been doing yield management for decades. E-commerce and online retail are catching up fast. Insurance has used risk-based personalization for a century. Streaming and subscriptions use a softer version: A/B-tested landing pages and retention offers tailored to your cancel risk.

Not legally, no. Price gouging usually refers to charging unconscionable prices during emergencies for essential goods, and most states have laws against it. Algorithmic pricing is routine, ongoing personalization that isn’t tied to any crisis, and it isn’t regulated the same way. The ethical questions overlap a lot, but the legal ones really don’t.

Yes, in some industries. The famous 2012 Orbitz study found that Mac users were being shown hotels about $20 to $30 more expensive than what Windows users saw for the same rooms. Amazon’s device-based variance is narrower now, but rideshare, hotel, and e-commerce platforms still use device as one signal among many. Mobile vs desktop also matters on most booking sites.

Device type and version, operating system, browser, IP-derived location, referrer URL, cookies, session history, account status, past purchases, email engagement, how intensely you’ve been searching, time of day, and day of week. If the company buys data from brokers, add credit signals, demographic inferences, and cross-site behavior to that list. Pricing models typically weight 15 to 25 inputs at once.

Surge pricing applies to everyone opening the app in a given market at a given moment. If the surge is 2x in downtown Boston at 9pm on Friday, every rider sees the 2x. Personalized pricing is different: it sits on top of the base fare and quietly shifts the number for you individually based on your history. You can see the surge. You usually can’t see the personalization.

In the U.S., the FTC monitors deceptive pricing practices, and several states (California, Vermont, Colorado) have data-privacy laws that give you some rights over your data. The EU’s Digital Services Act and Digital Markets Act are pushing harder on transparency. But no U.S. federal law currently requires companies to disclose that a price has been personalized.

No. Price discrimination is the economic term for charging different prices to different customers for the same product, and it’s mostly legal. Proxy discrimination is a narrower concept: it’s when a pricing model uses inputs that correlate with a protected characteristic (ZIP code for race, for example) without asking about the protected characteristic directly. The outcome is the same as direct discrimination, with a legal workaround on top.

It’s the pattern where long-term customers end up paying more than new ones for the same product. It’s well documented in insurance, broadband, and phone contracts, and has been flagged by state regulators as a real consumer harm. The short version: companies raise rates quietly on customers who haven’t shopped around recently, because those customers are modeled as less price-sensitive.

A few things actually help. Compare the same product across incognito, a different device, and a clean account. Use a VPN when you’re checking travel prices. Clear cookies between searches. Check comparison sites before trusting a quoted price. Remember that loyalty isn’t always rewarded (sometimes it’s quietly penalized). And support regulation that requires disclosure, which is the single most effective long-term fix.