Flexible dates: what "shift two days and save $140" actually compares

If you gave us a window instead of one exact date, some alerts now carry a second line under the fare: something like "Shift 2 days to Mar 14 instead of Mar 16 and save $140." It's new, it's easy to misread as a bigger claim than it is, so it's worth being precise about what that sentence compares.

Where the number comes from

Every time we check a watch with a flexible window, we don't look at one date. We sample up to eight dates spread across the window you gave us and price all of them in the same run. The alert has always been about the cheapest one we found — that part hasn't changed. What's new is that we now also look at the other dates from that same run before we throw them away.

Specifically: the fare we alert on, and the next-cheapest date we saw alongside it. If the gap between those two is at least $20, we say so. Under that, we say nothing — a $6 difference isn't a reason to change your plans, and a line that showed up every time would stop meaning anything.

Why the next-best date, not the worst one

We could have built this to look more impressive. Instead of comparing against the closest-priced alternative, we could scan all eight sampled dates and show the gap between the cheapest and the most expensive one — same window, bigger number, better-looking email.

We didn't, on purpose. The point of this line is to tell you something true about the choice you actually had. If you're flexible across a window, the real alternative you were weighing was never "book on the worst possible day" — it was "book on this day, or the next reasonable one." Comparing against the worst date would be technically accurate and functionally misleading, which is the same category of mistake as reporting a fare as its full price when bags and a seat aren't included. The score on this product is meant to be something you can act on, not a number optimized to look good in a subject line.

It's a rendering choice, not a new number

Nothing about this required a new data source or a new column. The eight sampled dates were already being priced and written to our observation history every time a flexible watch ran — we were just discarding seven of them the moment we picked the cheapest. This line is what's left over when you stop throwing that away.

That matters for the same reason it matters everywhere else on this site: every figure you see from us traces back to a fare we actually recorded, not something inferred or estimated after the fact. The savings line either points at two real prices we saw in the same run, or it doesn't appear.

What it doesn't tell you

It isn't a prediction that the cheaper date will still be cheaper next time we check — fares move, and the next run might rank those same two dates the other way. It also isn't a claim that no cheaper date exists outside the eight we sampled; a wide window with a narrow sample can still miss something. And it isn't a substitute for the percentile comparison the score itself is built on — that's about whether this fare is good against this route's history. The shift line is a narrower, simpler fact: of the specific dates we checked for you this time, here's what flexibility was actually worth.

We said back when we turned down hotels that date-shift savings was one of the few additions that still fit what this product is for, because it extends a claim we can already stand behind instead of adding one we can't. This is that claim, built out of data we were already collecting for a different reason.