Several PointsCrowd guides include a block that answers one specific question with real numbers: is buying points on this program worth it right now, or is cash cheaper? Those numbers come from our own award search database. This page explains exactly where the data comes from, how we clean it, and what the numbers do and don’t mean.

Where This Data Comes From

Every number in these blocks comes from the same award search backend that powers search on this site — real hotel listings, real dates, real points prices and real cash prices for the same room and the same night. We query it directly, pull every available listing for the program in question, and work from that full dataset rather than a handful of hand-picked examples.

What “Cents Per Point” Means

For every night in our data, we divide the cash price of the room by the number of points it costs, then multiply by 100. A room that’s $300 in cash and 40,000 points works out to 0.75¢ per point. We run this calculation per night (dividing multi-night stays by the number of nights first), so a 3-night stay and a 1-night stay are directly comparable.

The 15% Margin: When We Call a Redemption a Win

We don’t call buying points “worth it” the moment the math breaks even. A redemption has to beat the purchase price by at least 15% before we count it as a win. That margin exists because an award night usually doesn’t earn points or count toward elite status the way a paid stay does, and points you’ve bought can’t be refunded if your plans change. A thin, breakeven edge isn’t a real edge once you account for that — so we require a real one.

Cash Prices Include Taxes and Fees

The cash price we compare against is the full price of the room, taxes and fees included — not a pre-tax rate that looks better than what you’d actually pay at checkout.

How We Clean the Data

Raw search data is messy, so before any number makes it into an article, we remove:

  • Rates that aren’t priced in points at all.
  • Listings where the points price is inflated by a credit card sign-up bonus rather than a standalone award rate — these would make buying points look artificially better than it is.
  • Rows missing a points price, a cash price, or a length of stay.
  • Duplicate listings for the same hotel, date and stay length.
  • Dates less than 7 days out, since you couldn’t act on them by the time you’re reading the article.
  • Statistical outliers — a value per point far outside the normal range for that program almost always means bad or mismatched data, not a real deal. We set a sane range per program and review anything outside it by hand rather than publishing it automatically.

We only publish a block once the cleaned sample clears a minimum size — at least 500 nights across at least 30 hotels. Below that, the numbers move around too much from one snapshot to the next to be useful.

How Often We Update

We recalculate every block whenever the program in question runs a new buy-points sale, using that sale’s actual price per point. Between sales, we refresh roughly monthly using the program’s standard, non-bonus price — which often means the honest answer is “don’t buy right now, wait for a sale,” and we say so when that’s what the data shows. Each snapshot is dated, so a block always tells you how current its numbers are.

What This Isn’t

This is a snapshot of real listings at the moment we pulled them, not a live price guarantee. Points and cash prices move — sometimes daily — so always confirm the actual price before you buy points or book an award stay. Treat our numbers as a reliable starting point for the question “is this generally worth it,” not as the exact price you’ll see when you search yourself.