Why Two Canadians May See Different Online Prices
Matthew Leo · Published August 4, 2026 · Business & Policy
Two Canadians can search for the same product at the same time and still receive different offers.
Sometimes the explanation is ordinary: one person qualifies for a loyalty discount, lives in a different delivery area or opened the page after inventory changed. In other cases, software may use customer or behavioural data to predict what each shopper is willing to pay.
Those situations are commonly flattened into one alarming phrase—“AI pricing”—even though the legal and consumer consequences depend on what the system is actually doing.
Dynamic and personalized pricing are different
Dynamic pricing changes an offer in response to conditions surrounding the transaction. Demand, remaining inventory, competitor prices, time and location can all influence the result. Airline tickets and ride-hailing fares are familiar examples.
Personalized pricing uses information associated with a particular customer or customer segment. That could include account history, loyalty status, device information, previous purchases or an estimate of price sensitivity.
| Pricing method | What changes the price | Example |
|---|---|---|
| Dynamic pricing | Demand, time, inventory or market conditions | A hotel room becomes more expensive as availability falls. |
| Segment pricing | Membership in an eligible group | A verified student receives a discount. |
| Personalized pricing | Data associated with the shopper | An account receives an individually targeted offer. |
| Algorithmic coordination | Pricing systems react to competitors or shared signals | Competing sellers’ software may repeatedly move prices together. |
The last category creates a competition concern when algorithms facilitate agreements or coordination that would be unlawful if people performed the same conduct manually. Software does not provide an exemption from ordinary competition law.
Canada has not simply banned different prices
A price difference is not automatically evidence of unlawful conduct. Businesses routinely offer negotiated prices, geographic rates, subscriptions, coupons and loyalty promotions.
The harder questions involve deception, competition, privacy and discrimination. Was the displayed price represented as available to everyone? Were mandatory fees hidden until checkout? Did competitors coordinate? Was personal information collected or used in a way that violated Canadian privacy requirements? Did the system produce discrimination connected to a protected characteristic?
The Competition Bureau’s algorithmic-pricing consultation found both potential benefits and risks. Personalized discounts can make a product more affordable, while opaque price discrimination can transfer more money from customers who receive less favourable offers.
The Bureau did not conclude that every personalized price is illegal. Its work instead highlights how existing competition rules apply when algorithms support misleading practices or anti-competitive coordination.
Data makes the difference difficult to see
A physical store usually displays one shelf price to everyone standing nearby. Online services can create a different storefront for each session without shoppers knowing that another version exists.
The inputs may include:
- Account and loyalty history
- Approximate location
- Device and browser information
- Previous searches or purchases
- Referral links and advertising campaigns
- Time, demand and remaining inventory
- Coupons or experiments assigned to an account
This does not mean every retailer uses all of those signals to set prices. It means that a visible difference alone cannot reveal the cause. A browser cookie may matter, but so may an expired promotion or a delivery fee calculated for another postal code.
Run a useful comparison instead of guessing
When a price looks suspicious, preserve enough information to reproduce the comparison.
- Record the exact product, model, quantity and seller.
- Take a screenshot showing the price, date and time.
- Check whether taxes, shipping or mandatory fees are included.
- Compare the price while signed in and signed out.
- Open a private browsing window.
- Compare another device on a different connection when available.
- Use the same delivery location and currency.
- Proceed to the final checkout screen without completing the purchase.
Change one variable at a time. If the account, device, location and time all change simultaneously, the experiment cannot identify which difference mattered.
Do not create several fake identities or provide false eligibility information. The goal is to compare ordinary offers, not violate service terms or claim a discount that does not apply.
Private browsing is not an invisibility switch
A private window prevents the new session from using many existing cookies and removes its local history after it closes. It does not conceal the Internet address, approximate location, device characteristics or information entered during checkout.
A VPN also changes only part of the picture. It may alter the visible IP address, but signing into the same account reconnects the activity to the customer. It can additionally trigger foreign currencies, fraud controls or offers that are not actually available at the delivery address.
Mapletechie’s VPN and iCloud Private Relay comparison explains why network privacy tools do not make an authenticated shopper anonymous.
Look at the final price, not only the first number
A retailer may display the same product price while charging different shipping based on location. Another may advertise a low starting amount before adding mandatory fees.
Record the complete payable total immediately before purchase. Canada’s Competition Act addresses deceptive representations and drip pricing involving mandatory fixed charges, apart from government-imposed amounts such as applicable sales tax.
A fluctuating headline price and a hidden compulsory fee are different problems. Keeping the complete checkout trail helps regulators or the business understand which one occurred.
When to report the difference
Contact the business first when the offer may reflect a coupon, account benefit or obvious technical mistake. Ask which condition produced the price and preserve the reply.
Consider reporting the matter to the Competition Bureau when the evidence suggests misleading advertised prices, hidden mandatory charges or coordination between competitors. Privacy concerns can be raised with the relevant federal or provincial privacy commissioner, depending on the organization.
A provincial or territorial human-rights body may be more appropriate if credible evidence connects discriminatory treatment to a protected ground. One unexplained screenshot is unlikely to establish that connection, so preserve repeated comparisons and communications.
Price transparency should travel with the algorithm
Algorithmic pricing is not inherently abusive. Lower off-peak prices, targeted discounts and rapid responses to shortages can benefit shoppers and businesses.
The problem begins when the customer cannot tell whether the price reflects the product, the market or a private judgment about that particular person’s willingness to pay.
Canadian regulators should require businesses using material personalization to explain the main categories of data involved and provide a meaningful way to challenge an incorrect classification. Shoppers should not need a laboratory and three borrowed phones to learn whether everyone was offered the same deal.
Tags: algorithmic pricing, consumer data, Competition Bureau, online shopping, privacy