How Lowvyn Scores Work
Every badge and number you see on Lowvyn is calculated from real price history. No guesswork, no opinions. Just data.
Lowvyn Score (0 to 100)
The Lowvyn Score is a single number that tells you how good a deal you are getting right now. Higher is better. A score of 85 means this is a great time to buy. A score of 30 means you should probably wait.
The score is built from four weighted components. Their raw weights are 45 : 15 : 15 : 15, and the total is scaled so the score uses the full 0 to 100 range. A product does not need to be sold at multiple retailers to reach the top band: a single-retailer product at a genuine, well-tracked low can score in the 80s.
Price position (biggest weight)
Where is the current price compared to the lowest and highest prices we have recorded? If the price is at or near its all-time low, this component contributes its full weight (half the score). If it is near the highest price we have seen, it contributes close to zero.
Guard: For products whose all-time range is under 5% (i.e. the lowest and highest prices we have seen are basically the same number), this component is capped at half its weight. A newly-listed item with two prices $5 apart should not be able to claim a lowest-price full score.
Freshness
Has the price changed recently? A price that dropped in the last week is more relevant than one that has not moved in a month. If the price changed within seven days, this component contributes its full weight. After 30 days without a change, it drops to zero.
Stability
How predictable is this product's price? Products with stable, consistent pricing score higher here because you can trust that the current price reflects the real value. Products with wild swings score lower because today's price might not last.
Guard: Stability needs at least 14 recorded prices before it can grant full credit. Below that threshold, this component stays neutral. A handful of flat readings on a freshly tracked item is coincidence, not evidence.
Cross-retailer comparison
When the same product is sold at multiple Canadian retailers (Amazon, Walmart, Best Buy), this component rewards the fact that you have alternatives to compare against. The cross-retailer matching system identifies these duplicates by UPC, MPN, brand, and title similarity.
- Full credit: Confirmed at 2 or more retailers. You can see "Also available at" alternatives on the product page.
- Half credit: Grouped with similar items but only available from one retailer (e.g. multiple SKUs of the same product on the same store).
- No credit: No cross-retailer matches found yet. Either the product is genuinely unique to one retailer, or our matcher has not connected it yet.
Example: A car seat priced at $299 with an all-time low of $279 and an all-time high of $399, tracked for two months across Amazon and Walmart, might score 88: strong price position (lower third of its range), full freshness (it dropped this week), high stability (consistent prices), and full cross-retailer credit (available at both retailers). That is a genuinely strong signal.
Why a product appears where it does in our lists
The Lowvyn Score is what we show on every product. The order products appear in on category and store pages (like /amazon or /deals/car-seats) is set by a related but separate number called the Lowvyn Rank.
Lowvyn Rank starts from the Lowvyn Score and then applies a few small adjustments that are appropriate for ranking but not for the per-product score:
- Affiliate yield by retailer. Amazon currently pays a meaningfully higher affiliate commission than Walmart or Best Buy, so an Amazon listing earns a small tie-break boost. This is purely a ranking factor; it does not change the price, score, or signals on any product page.
- Demotions for missing data. Products without a brand or image are pushed lower so the listing pages look consistent. Quarantined products (those we have flagged as having unreliable price data) are pushed much further down.
The Score still does the heavy lifting. A product with Score 85 ranks above a product with Score 50 regardless of these adjustments. The adjustments only matter when two products have similar Scores.
Price Signal
The price signal is a quick label that tells you where the current price sits in the product's historical range. It is based on the price percentile, which measures how close the current price is to the all-time low versus the all-time high.
| Signal | What it means | Price range |
|---|---|---|
| Buy Now | Price is near the all-time low. This is a rare opportunity. | Bottom 15% |
| Good Price | Price is below average. A solid time to buy. | 15% to 35% |
| Average | Price is in the middle of its range. Not great, not bad. | 35% to 65% |
| Wait | Price is above average. It has been cheaper before. | 65% to 85% |
| Wait for Drop | Price is near the highest we have recorded. Not a good time to buy. | Top 15% |
Deal Verdict
The deal verdict compares the current price to the average price over the last 90 data points. It answers a simple question: is this price actually lower than what this product normally costs?
This is different from the price signal. A product can be at a "Good Price" (relative to its all-time range) but still be at "Average Price" (relative to its recent average). The deal verdict focuses on what you would typically pay, not the extremes.
| Verdict | What it means | Discount vs average |
|---|---|---|
| Real Deal | This is genuinely cheaper than usual. Buy with confidence. | 20%+ below average |
| Modest Deal | Slightly below the typical price. A small saving. | 5% to 20% below |
| Average Price | Close to what this product normally costs. No rush. | Within 5% of average |
| Above Average | More expensive than usual. Consider waiting. | 5% to 15% above |
| Bad Time to Buy | Significantly overpriced compared to recent history. | 15%+ above average |
Lowest Tracked Badge
When a product sits at the lowest price we have tracked, it gets a green "Lowest tracked" badge. We only show it when the claim is backed by real evidence: at least 5 price checks, at least 30 days of tracking, and a meaningful observed price range. A product we started tracking last week does not earn the badge just because we have not seen it cheaper yet. If you have been waiting for a good price on a well-tracked product, this is your signal.
Why our price can differ from what a retailer shows you
Retailers, especially Amazon, do not always show everyone the same price. When a product has no first-party offer, the "buy box" rotates between marketplace sellers, and two shoppers can see two different prices for the same product at the same moment. We have watched a single product show one shopper a fleeting "only 1 left" listing at 66% off while showing another shopper the regular price.
Lowvyn records the verified price. When a marketplace offer looks suspicious, priced far below the product's recent history from a seller we cannot vouch for, we hold it instead of counting it. If the same price is still there on our next check, it was a real deal and we record it. If it vanished, it was noise, and it never touches the price history, the score, or the lowest-price badge. So if the retailer's page briefly shows a different number than we do, that is usually the reason: we wait for prices to prove themselves.
How much data do we need?
Scores improve as we collect more price history. Here is how the experience changes over time:
- 1 to 2 data points: We show the price signal (based on lowest and highest recorded) but mark the deal verdict as "Checking" since we do not have enough history yet.
- 3 or more data points: The deal verdict becomes active. We can now calculate a meaningful average and tell you whether the current price is above or below it.
- 14 or more data points: The stability component unlocks its full range. Below this threshold, stability stays at a neutral 7.5 points so coincidental flat readings on a newly tracked item cannot claim a high score.
- 30 or more data points: Full confidence. The score, signal, and verdict all reflect a solid baseline of price behaviour.
What we do not do
Lowvyn scores are based entirely on historical price data. We do not factor in reviews, popularity, brand reputation, or personal preferences. We also do not accept payment from retailers or brands to influence scores. If a product scores 90, it is because the price data says so.
Scores are recalculated every time you view a product. They are never cached or manually adjusted. The same product will get different scores on different days as prices change.
Questions about how a specific product scored? Check the product's price history chart on its detail page. The chart shows every price point we have recorded, so you can see exactly why the score is what it is.