Last year I had a cocktail in the extremely fancy Ritz Hotel bar in Paris with my friend Rohan. I was absentmindedly taking (bad) pictures with my rather conspicuous camera:

We'd hardly touched our €35 drinks when the elegant waiter whisked over and gave me a quick "non monsieur -- pas de photos." Bashful, I put the camera away, returning to my cocktail and wanting to vanish.

I figured that was the end of it. But five minutes later, another Frenchman, this one beefier and dressed in a black suit, demanded that I take my camera out and delete the photos as he watched. I stammered that I was a tourist and not a journalist, and that I wasn't even sure how to delete the photos from my camera, but he was insistent. He explained to me that it was his job to maintain the clientele's privacy -- if someone brings their mistress there, he needed to make sure they were treated with discretion.

He eventually relented when I looked helpless enough, and so the above photo is actually contraband.

His comment fascinated me. I'd never thought of discretion being a value proposition for a hotel, but it obviously made sense when I thought about it. Covert business dealings, spies trading secrets, romantic liaisons -- all of these fit my notion of what happens at exclusive hotels.

The single podcast episode I have recommended to the most people is this one with Nick Kokonas, the cofounder of Alinea, a very fancy & innovative Chicago restaurant, known for its theatrical presentation of food. I love this idea:

Know what you're selling and then actually sell it, which sounds incredibly obvious....I'm telling you what, there should be a business course on that. If I ask Gramercy Tavern what they're selling they'd say food. And I'd go, "No. You're actually selling seven different experiences plus some merchandise."

So, I wanted to understand what hotels are actually selling. With that in mind, I pulled data on New York City hotels from SerpApi. From it I dug into how hotels in the city are priced and reviewed.

Price Distributions

We can see that most of the hotels in NYC are in the $100-$400 range, with some notable outliers at the high end. This distribution is a good example of something I wrote about several months ago: I am continually surprised by how skewed most datasets I see are -- how much bigger the extreme values are than the median values.

These hotels are spread out across the five boroughs and east Jersey, but most of them (and the most expensive ones) are in lower and midtown Manhattan:

What do reviews tell us about what hotels are selling?

Hotels at different price points are selling very different things.

Ratings of hotels in New York (and, I suspect, in most places) tend to center around 4.25/5, with some really poor outliers on the low side. The low outliers are all priced below $300 -- once you are paying more than $300, you can be pretty sure you'll have an okay experience. So, for anything much over that -- say, $1,500 at the Four Seasons -- you're buying something well beyond a comfortable place to sleep.

This data includes the attributes that google extracts from reviews, like the below for the Aman (as an aside, if you want a reminder that some rich people are deeply unhappy, sort their reviews by "lowest score," and click into some profiles).

So, we can see which attributes come up at different price points -- that is, what people talk about when they review hotels at different price levels:

For the small handful of extremely expensive New York hotels, the ones charging $1,300 or more per night, service was far and away the most frequently mentioned thing in reviews. I suppose this includes brusk security guards who keep camera-toting plebeians in check, but also kind of cute things like this for the Four Seasons Downtown. Must be nice!

As the price declines, service continues to be important, but the property becomes relatively more important. I wonder if this is because beyond a certain price range, you can't really compete on property -- how much more dazzling can the lobby of a $3,000 hotel be than that of a $1,000 hotel?

Under the $300 price point in New York, the basics become important: factors like cleanliness and the ability to sleep.

I'd summarize it like this:

• Low-end New York hotels are selling a clean place to sleep.
• Mid-range New York hotels are selling location and clean rooms with good service.
• Fancy New York hotels are selling beautiful spaces with good bars & dining and great service.
• Super fancy New York hotels are selling all of the above, but above all else, they sell pristine service.

I'd love if someone who has more knowledge of this could weigh in on how accurate this is.

Speaking of hotel bars, I happen to have this week off 😎 , and so for my reference here's the distribution of the ratings of hotel bars in the city:

I walk by the Wall Street Hotel a lot and it looks nice, but I don't want to go to FiDi when I'm on vacation. Maybe I'll head to Nine Orchard, which looks beautiful, even though the neighborhood is over, or the Hotel Chelsea, which has the best song written about it (edging out Margaritaville).

But of course I am a diehard Casablanca fan and so maybe I'll make my way up to Times Square for a drink with some tourists at Rick's. Everybody comes to Rick's.

If you enjoyed this, you'd probably like this post tracking cocktail popularity in the New York Times. Subscribe to my writing below: