"...we've decided to end Rewards soon, as we shift our focus to our new Uber One membership program." (EDIT: received in an email notification)

It looks like they're trying to drive loyal customers into a paid membership who may have been otherwise happy with the benefits they were receiving under Uber Gold or other level.

First they nerfed it. Used to be able to redeem 500 points for $5.

Earlier this year they replaced it with discounts, so 400 points gets you 10% off. My most expensive ride in the past year was $30, so I view this as a 25% reduction in point value.

I guess that wasn’t enough?

Loyalty programs also cost money in the form of the free stuff you give away as benefits. For a company that perennially has trouble turning a profit on razor-slim (and most often negative) margins, this would be a problem.

There's also a question of if the retention ability of the loyalty program outweighs of the cost of benefits you're giving out.

Additionally the shape of the business matters. For some industries (see: hotels, airlines) a lot of the benefits are in the form of things they would've lost money on anyway if they didn't give to the loyal customer. Upgrading room at check-in? That room was pretty much guaranteed to go empty that night anyway. Upgrading airplane seat 24h before departure? That seat was pretty much guaranteed to go empty if they didn't. So the opportunity cost of giving the user the reward is ~zero-ish. This is definitely not true for Uber, where free rides = lost money.

This is a great point, I was wondering why this seemed to work well for airlines and not Uber. And for airlines, I mean it's not even 24hrs sometimes, it's literally at the gate when there is 0 downside for them (other than the cost of a slightly more expensive meal and maybe some cheap champagne?)

Good for profitable companies, yes.

I don't see the difference? Like if the program was providing value, perhaps in the form of retention, then the value exceeds the loss.

They must have data that their loyalty program did not improve retention to justify the costs to run the program or (as others pointed out) they must see that force loyal customers to Uber One will increase their revenue/more.

> if the program was providing value, perhaps in the form of retention, then the value exceeds the loss.

This looks like a syllogism, but it is not. If something provides value, it does not mean that the value exceeds the loss. Otherwise everything useful would be profitable, like my gold toothbrush.