The other day, I went hunting for a new television series to watch while waiting for a new episode of the popular PBS series "Downton Abbey" to air on Sunday night. A search for Downton Abbey on Netflix turned up two recommendations — "Bleak House" and "The Way We Live Now" — two BBC dramas about wealthy English families in the 19th century. Terrific, I thought. I prepared my queue accordingly.

Remember when you used to get movie recommendations from the film buff who worked behind the counter of your neighborhood movie rental store? Before the Internet Age, people relied on friends, family members, and knowledgable staff members at various businesses for recommendations. Now, getting recommendations online is just a matter of doing a simple search.

When customers are searching for a product on Amazon, for example, the site can tell you what "customers who bought this item also bought." Readers who bought the bestselling book The Girl with the Dragon Tattoo by Stieg Larsson also bought The Hunger Games by Suzanne Collins — voilà, instant recommendation.

You can also get recommendations based on your favorite restaurants, clothing stores, grocery shops and more. Bundle's "Related Places" feature works in a similar way to Amazon's "customer's also bought" feature in that it can tell you where customers of well-loved businesses also like to dine and shop. The feature is located on the left side of every merchant page.

Let's take fine dining restaurant Per Se as an example, which former New York Times food critic Sam Sifton described as "the best restaurant in New York City."

Our data shows that diners typically spend $749 per visit at Per Se, which sounds like a lot, but makes sense given that a chef's tasting menu costs $295. Throw in two of those, a nice bottle of wine, tax and tip, and you'll easily hit $749.

Obviously, not every Tom, Dick, and Harry can afford to drop $750 for dinner, so let's take a look at where else Per Se diners like to eat:




Bundle's list of related places are determined using conditional probability, or the chance that a person will visit one business based on a person visiting another business. And the businesses with the greatest odds of also being visited are listed at the very top.

In the list above, we know that people who eat at Per Se have also been known to eat at restaurants like Jean Georges, Bar Boulud and The Modern. So even if Per Se is too expensive for a diner, we know that Per Se patrons also like to go to Bar Boulud, which not only has a high Bundle rating, but a typical cost of $77 — nearly 10 times less than the average check at Per Se. Obviously you won't be able to receive the same culinary experience that you would at a revered restaurant like Per Se, but the "Related Places" feature is a great way to look for alternatives places to go to within your budget.

The feature is also a great way to discover new favorite places to go based on your current favorites. Say, for example, that you like to shop for designer goods at consignment shops, and your favorite shop is Cadillac's Castle on New York City's Lower East Side.



Cadillac's Castle's related places shows that customers have also been known to shop at places like Amazon and Kmart, which makes sense since consignment shoppers are also likely to be bargain hunters. But we also see that shoppers go to Tokio 7, a consignment shop with a higher Bundle rating, and where shoppers are willing to spend about $34 more dollars per visit. Shoppers are also going to Housing Works, a highly rated thrift store in New York where the typical spend is under $30, compared to Cadillac's Castle's $66 typical spend.

Knowing that there are other discount shops where customers spend less is like finding gold for bargain hunters, because is there's anything that they like, it's a better bargain. Eventually, new places will become old favorites, and you'll be able to get more recommendation as your list builds.


Try it out for yourself. See where else people shop and eat based on your favorite places.


Previously: "Ratings, reinvented: Bundle's business ratings can tell you things no other rating site can"