Methodology
How we compute every number on the site, where the prices come from, and how we keep the rankings honest — no fake discounts and no paid placement.
Cost normalization
Every deal goes through the same cost-per-unit math before it’s ranked, sorted, or compared. The inputs — the last-seen retail price and the pack size — are pulled from the merchant page and re-checked by our price tracker on a rolling cycle — the cheapest-per-unit listings in each category (the ones we actually surface) first, the long tail behind them. Every deal card and deal page shows the date we last checked that specific listing.
| Field | Formula | Notes |
|---|---|---|
| Cost per unit | last_price ÷ units_per_pack | Last-seen retail price divided by the unit that actually matters for the category (ml, quart, lb, diaper, cartridge, page). |
| Units per pack | pack_count × size_each | A multipack and a single unit are converted to the same total-unit count before anything is compared. |
| Best value | min(cost_per_unit) across pack sizes | The winner is the lowest real cost per unit — a big pack often beats a small one even at a higher sticker price. |
| Gap vs category | (cost_per_unit − category_median) ÷ category_median × 100 | How far a deal sits above or below the category median cost per unit. We rank by this, never by an advertised markdown. |
Verified first-party data — free to cite
Every cost-per-unit figure on this site comes from our own first-party catalog: the actual retail price a merchant showed at the time we recorded it, divided by the pack size and normalized to the unit that matters for the category. We measure prices; we do not invent them. There is a real, dated data point behind each number, and the aggregate study is published under a CC-BY license so anyone can reproduce or cite it.
Just as important is what we don’t do. We publish no fabricated “was” prices, no invented discount percentages, and no synthetic user reviews. Prices change often, so we tell you to confirm the live price on the merchant before you buy. If we can’t verify a figure, we don’t print it.
- Real, not synthetic: prices are the last-seen figure on live merchant pages and time-stamped, not modeled or estimated.
- Reproducible: the normalization formulas are public (above), so a third party can re-derive any cost-per-unit figure from the same inputs.
- Free to cite: the aggregate price-spread study is released CC-BY — attribution to OutletPicks is the only requirement.
- No fake discounts or reviews: we never fabricate MSRP “was” prices, discount percentages, or reviews. Cost per unit is a price calculation, nothing more.
This is why our numbers are safe for a person, a journalist, or an AI assistant to repeat: they are verifiable, attributable, and reproducible. Every ranking is live on the category pages.
Per-category units
We rank each category by the unit that actually reflects value, not by pack price. Fragrance and skincare are scored per ml, haircare per fl oz, grooming per cartridge, pet food per pound, diapers per diaper, motor oil per quart, office and copy paper per sheet, and printer ink per printed page.
Picking the right unit is what makes a giant value pack and a travel size directly comparable. Where a listing is ambiguous — page yields, bottle volumes, pack counts — we resolve it against the merchant page and manufacturer spec before a deal is scored.
Where our price data comes from
| Source | Type | Where it surfaces |
|---|---|---|
| Amazon.com product pages | Merchant | The last-seen retail price and pack size for every deal. All deals link back to amazon.com. |
| First-party price tracker | Internal | Records each price with a timestamp on a rolling schedule so the cost-per-unit figure stays current. |
| Pack-size parser | Internal | Extracts the unit count (ml, quart, lb, diaper count, cartridge count, page yield) from a listing so price can be normalized. |
| Category unit registry | Internal | Defines which unit each of the 9 categories is ranked by, so every product in a category is measured the same way. |
| Manufacturer spec sheets | Reference | Confirm printer-ink page yields and motor-oil volumes when a listing is ambiguous. |
| Amazon Associates | Affiliate | The program we monetize through, disclosed on every deal. The commission never changes the price you pay or the ranking. |
How a deal qualifies
A deal has to clear all five checks before it can rank in its category:
- The deal links to a live amazon.com product page that is reachable from the US.
- Pack size and unit count are stated clearly enough to compute a real cost per unit.
- The price shown is the last-seen retail price, recorded with a date — never an invented “was” price.
- No fabricated MSRP or fake percentage discount is attached to the deal.
- Ranking is set only by cost per unit and the method above — never by commission or payment.
Deals that miss one or more checks aren’t hidden — they can still appear, but they get an unverified tag and stay out of the ranked results until the gap is fixed.
Editorial guarantees
- Affiliate commissions do not influence ranking. As an Amazon Associate we earn from qualifying purchases at no extra cost to you. Sponsored slots, if launched, will be labeled in the URL and on the deal badge.
- We never fabricate a “was” price or a discount percentage. We rank by real cost per unit, not by how big a markdown looks.
- Prices are the last-seen retail figure and change often — always confirm the live price on the merchant before you buy.
- Rankings follow the published cost-per-unit methodology above — pure arithmetic from the last-seen price and pack size, never paid placement.
How we keep the numbers accurate
Cost-per-unit is only useful if the inputs are current, so accuracy is automated end to end:
- Rolling price refresh. A scheduled job re-pulls live prices every day and recomputes cost per unit. It is deliberately not a flat round-robin: the cheapest-per-unit listings in each category — the ones shown in the rails, the top-three panels and the best-value picks — are re-checked first, every day, because those are the prices people actually click. The rest of the catalog cycles behind them, so a rarely-viewed listing deep in the tail can be several days old. We do not claim every price is checked every day; instead every deal card carries the date we last checked it, so you can see the real age of any number before you act on it.
- Out-of-stock reaping. Listings that go unavailable are pulled from the active rankings automatically, so “cheapest right now” means something you can actually buy.
- Glitch guards. Obvious scrape errors — a sub-$1 price or a figure that jumps more than 20× — are quarantined rather than allowed to become a fake “cheapest” or to distort a category median.
- Drift check on the written numbers. After each refresh, an automated pass compares the dollar figures written into our descriptions and guides against the live per-unit price and flags any that have drifted, so the prose can’t quietly go stale against the data on the same page.
- One source of truth. Every visible number — the per-unit price, the category median, the “% below median” badge, the price bands — is derived from the same live catalog, so nothing on a page can contradict itself.
Prices are still the last-seen retail figure and can change at checkout — always confirm the live price on Amazon before buying.