Hacker News Reader

Comment by JumpCrisscross | Hacker News Reader

> I had Gold status (30 nights) with lounge access, which was super-useful. Now that got shifted to Platinum (50 nights)

Marriott seems to be running into the same problem as American airline rewards: counting nights and miles, only, doesn't adequately differentiate between high-spend and low-spend nights and miles. So you have to set the nightly rate such that someone staying at your cheapest property can access perks at the most-expensive property. That, in turn, ends up making the perks less valuable for those who frequent your pricy properties. (This is Gresham's Law [1].)

The loyalty programs should have been kept separate, with the benefit of high-value programs getting credit at low-value ones (but not the other way, at least not at parity).

[1] https://en.wikipedia.org/wiki/Gresham%27s_law

View on HN

This sounds like a misinterpretation of Gresham's Law. That's specifically talking about currency, which I suppose in this case would be the reward points. There is no "good" vs "bad" version of the currency though. Certain types of points are not driving other types of points out of circulation. The value of the points depends on how you use them, with people more eager to spend points on high-value redemptions.

> There is no "good" vs "bad" version of the currency though. Certain types of points are not driving other types of points out of circulation

Imagine Marriott has two brands: Fairfield Inn and Ritz-Carlton. They each have a rewards program which grants one free night for every night stayed.

Now suppose, in a fit of idiocy, the rewards programs are made cross-compatible. So a Ritz-Carlton reward night can be spent at Fairfield Inn and vice versa. Where do you think the rewards will be earned? Where do you think they'll be spent? The "bad" (Fairfield Inn rewards) drove out the "good" (Ritz-Carlton rewards).

I accentuated the number of brands, magnitude of their value difference and rewards program juiciness. This was to bring into relief the resulting effect. In reality, these terms--and the effect--are smaller. Still, Gresham's law applies if one treats rewards nights earned at different hotels as different currencies. They've simply, now, been brought into a currency union.

I'm having trouble seeing this as anything new or different.

These rewards programs were based on on nights or stays prior to the merger. SPG had no minimum spend requirements, and was widely considered the best hotel rewards program out there. And SPG had "low-end" properties as well, and there was no differentiator. I could earn all my rewards at an Aloft or Element and redeem them at a W or St. Regis.

If anything, Marriott had a higher quantity of "low-end" hotels and unless I'm significantly misremembering, you could always have earned points at the Fairfield Inn and used them at the Ritz-Carlton.

I don't understand what you think changed with this merger in that regard.

The issue I'm having with this comparison is that you are equating the "nights stayed" with the points earned (the more accurate currency).

>They each have a rewards program which grants one free night for every night stayed.

It's a lot more fair to say for every $1000 spent, you have enough points to redeem a free night. In which case, some people might choose to spend more nights at the cheaper Fairfield Inn, and some might choose to stay fewer nights at the luxury Ritz. The bad money is not driving out the good.

It's true when it comes to qualifying nights, people will try to hit elite night requirements at cheaper properties. But that's not driving out the business for higher-end properties. Especially because award redemption is (mostly) based on how much you spend, not how many nights you stay.

Nights and miles are the metric to separate out "high spend" customers, up until you establish a nights/miles metric and you encourage customers to game it...

I can see where he was going with the analogy to Gresham's, but maybe another G's law is a better fit:

https://en.wikipedia.org/wiki/Goodhart's_law

Reply on news.ycombinator.com