Les avis sur les produits vous mentent-ils ? Nous avons analysé 2,4 millions d'avis

Canadians spend an estimated $25–30 billion every year on the six product categories HomeGadgets tracks: TVs, appliances, laptops, monitors, vacuums, and unlocked phones. These are not impulse purchases — they are the refrigerator that runs for 15 years, the laptop a student carries through four years of school, the TV a family watches every night.

Getting the price wrong by even 10% on a $2,000 appliance means leaving $200 on the table. Getting it wrong by 76% — the gap we found between the cheapest and most expensive Canadian retailer for the same Samsung TV — means paying for two of them.

The Prices Are Already Different

Before we get to reviews, here's what the price data shows — because the review story and the price story are connected. For the same product, sold by multiple Canadian retailers, right now:

These are not sale prices versus regular prices. These are all current, regular retail prices for the identical product, on the same day. Notice that no single retailer wins across all six products. The retailer with the best price depends entirely on what you're buying.

Now, About Those Stars

You're standing in front of a $1,800 washing machine at Best Buy. Four and a half stars. 800 reviews. Sounds solid. But what if that same machine is rated 3.7 stars on the manufacturer's own website?

That's not a hypothetical. We found it in our data.

The Compression Problem: Almost Everything Is 4 Stars

In a universe where nearly every product rates between 4.0 and 5.0, the difference between a 4.6 and a 4.8 is not a meaningful signal. It's noise. A 0.1-star difference in TVs separates the 25th percentile from the 75th percentile.

Do Manufacturers Inflate Their Own Ratings?

The short answer: sometimes, and not always in the direction you'd expect.

• GE refrigerators rate 9% higher on GE's own website than on Best Buy — the clearest case of inflation in our dataset.
• Shark vacuums rate 7.1% higher on Shark's own site than at retail.
• Samsung dryers show a 5.9% gap in Samsung's favour.

But here's where it gets interesting:

• LG washing machines rate 14% lower on LG's own website than at Best Buy.
• Samsung washers rate 10.9% lower on Samsung.ca than at Best Buy.

A likely explanation: LG and Samsung's appliance sites attract buyers who already invested in the brand and have higher expectations. Retail sites average across a broader, less brand-loyal audience.

The Dyson Paradox

If you asked most people which vacuum brand makes the best products, Dyson would rank near the top. Our data tells a different story.

Dyson is the most-reviewed vacuum brand in our dataset by a factor of 85x. Their 12 tracked products have accumulated 130,092 reviews — an average of 10,841 per product.

Their average rating: 4.09 out of 5. That places Dyson last among major vacuum brands:

With 130,000 reviews, Dyson's 4.09 average is one of the most statistically reliable numbers in our entire dataset. Dyson makes premium-priced vacuums. Buyers have premium expectations. When a $900 vacuum doesn't meet those expectations, they say so.

The Most Disliked Appliance Category

We expected microwaves to rank lowest. The data surprised us.

Fridges are the most complained-about major appliance. TVs are consistently the highest-rated. A TV either works or it doesn't. A refrigerator has dozens of failure modes — and buyers live with those failures for 10 to 15 years.

The Planted Review Signal

A perfect 5.0 rating is essentially impossible to sustain at scale. When a product has 3 reviews and all 3 are 5 stars, one of two things is true: the product is genuinely exceptional, or the listing has been seeded.

That is why HomeGadgets does not display ratings on any product until it has crossed a minimum review threshold calibrated by category.

What Actually Makes a Review Trustworthy

1. Volume matters more than score. The Electrolux washer with 10,000 reviews at 4.70 is far more informative than a 5.0-star product with 8 reviews.
2. Retail sites are more neutral than manufacturer sites — but not because manufacturers suppress negative reviews. The audience is different.
3. The category matters. A 4.4-star fridge and a 4.4-star TV are not equivalent. Fridges average 4.35; TVs average 4.60.
4. Age of the product matters. Early reviews skew positive — early adopters are enthusiasts. A year-old product with 2,000 reviews tells you more.

How We Score Products at HomeGadgets

Rather than showing a raw star average, we show where a product ranks among its peers — same category, similar size, similar price range. We also show the review count prominently, because a 4.5 from 3,000 reviews and a 4.5 from 12 reviews are not the same thing.

We are transparent about where ratings come from. When the source is Best Buy Canada, we say so. When there aren't enough reviews to say anything meaningful, we say nothing.

Methodology

Data collection period: March 20–27, 2026. Prices scraped from retailer websites. Ratings from Best Buy Canada (reviews API), manufacturer websites (Samsung.ca, Sony.ca, Bosch.ca), and Visions Electronics (PowerReviews). Only aggregate scores collected — not individual review text. Category claims require minimum 25 reviews/product. Individual callouts require 50+ reviews. Brand claims require 5+ rated products.

Data sourced from HomeGadgets.ca — 3,800+ rated products, 11,000+ products tracked, 20+ Canadian retailers. Analysis conducted March 2026.

HomeGadgets.ca carries no advertising and accepts no fees from retailers or brands for placement. The site earns referral fees through affiliate programs — disclosed on every page. Those fees do not affect rankings. Full legal disclaimers at homegadgets.ca/terms.