Bank Data Is the Key Ingredient for Bundle's Restaurant Ratings

Register now

Bundle Corp. is catering to a different appetite with its newest personal financial management tool: restaurant ratings.

By mining credit card data to determine how much and how frequently consumers spend money at specific restaurants, Bundle is making a significant departure from what PFM sites typically provide. Bundle and its rivals primarily help consumers plan budgets, track spending and compare their spending habits against those of peers.

The New York company's restaurant rating tool opens the door for banks to give their customers another reason to frequent their websites — creating cross-sell opportunities. Observers also see the possibility for Bundle to sell the data to more established restaurant review sites like or Zagat's.

"These are the kinds of things that go toward giving PFM tools more depth and a greater sense of value," said Mark Schwanhausser, a senior analyst for multichannel financial services at Javelin Strategy and Research in Pleasanton, Calif.

Jaidev Shergill, Bundle's chief executive, said selling the restaurant data to others and cross-selling with banks are not in the works right now, though he did not rule either out.

Bundle plans to rate restaurants in all major U.S. metropolitan areas, but its pilot program, to begin in December, will focus on New York.

Instead of using subjective customer feedback to compile its rankings, the way consumer sites like do, Bundle will use the aggregated MasterCard Inc. and Visa Inc. credit card spending data of 20 million people, provided by Citigroup Inc., an investor in

"We are experimenting with data so we can help people make better decisions on spending in New York City," Shergill said. He added that the pilot program will start in New York because it has the highest per capita rate of restaurant spending of any U.S. city.

Bundle will rate restaurants based on the amount of repeat spending by consumers, and it will recommend other restaurants that may be cheaper. Higher amounts of spending and repeat spending at particular restaurants will correlate to higher scores.

Mohammed Khalil, director of finance data and strategic partnerships at Bundle, said that in contrast to other restaurant ranking sites, where the majority of ratings fall in the 3.5-to-4.5 range out of a high score of 5, less than a tenth of the restaurants that Bundle rates receives a top score.

Bundle already uses Citi's data to power the main feature of its website, a comparison engine that shows users how much they are spending in particular categories compared to peers. Users filter data by income level, ZIP code, whether they have kids and other factors.

Though Bundle initially offered its tools for only the aggregated data it gets from Citi and other sources, in August it added the capability for consumers to analyze their own data for more accurate comparisons. Financial data that users add for themselves is not mixed in with the aggregated data visible to the public.

Not everyone considers adding restaurant ratings a logical step for a PFM site.

"I see this perhaps as a step removed from services more required [of Bundle] in the shorter term, like integration with financial services sites or other financial-related sites," said Jacob Jegher, a senior analyst with Celent.

Nonetheless, Jegher said, Bundle is exploiting keen customer interest in restaurant reviews.

"The whole raison d'etre of is about data and the value around data, so no doubt they are clearly demonstrating prowess in that area," Jegher said.

"Having something different and fun to use is a big part of this," he added, "but I would rather see differentiation on educational components."

Cathy Graeber, founder of the consulting firm Swimming Upstream in Monterey, Calif., said the restaurant-review offering was far down on the list of must-haves for PFM sites.

"It is interesting, but not compelling, and it wouldn't resonate in consumers' minds to choose versus or a bank offering," she said, stressing that consumers are turning to PFM primarily to get a handle on their finances.

At the same time, Graeber said Bundle could potentially sell the data to more established names in the restaurant rating space.

"There could be a way for Bundle to monetize the data," Graeber said. Other ways to make money might include giving customers coupons for similar restaurants when they visit the site, with the restaurants paying Bundle for the opportunity to present the offer, she said.

Shergill said that Bundle has already been approached by a number of restaurant review companies, though he would not specify which ones.

"We don't look at the data as a restaurant review site; it can help consumers make smarter decisions on spending well and saving well," he said.

Bundle will present three rankings. One is a "confidence score," which represents the likelihood that a consumer would dine at a recommended alternate restaurant, based on the initially selected restaurant. The second is a "loyalty score," which is diners' share of total spending at a particular restaurant. A third component will show "local favorites," or restaurants in the same neighborhood that earn a high loyalty score.

Bundle uses nine months of credit card spending data at each establishment, Khalil said. "We are able to determine, based on comparisons to peer restaurants, whether it has a high frequency of return from the same people," he said.

The pilot program ranks all restaurants in New York that accept MasterCard and Visa. Bundle does not compile rankings for restaurants that are cash-only or that accept other cards exclusively.

Chris Kay, managing director and head of growth ventures at Citi, said Bundle's use of consumer data could address other needs or interests.

"It's an opportunity to compare social behavior and spending that is meaningful to consumers," Kay said. After restaurants, Bundle "could do this for health care or any other topic."

For reprint and licensing requests for this article, click here.