"The Best" is Subjective

Trying New Restaurants Doesn't Have To Be Such A Gamble.

- Josh Sapienza | Course Restaurant Guide

The Big Picture

Widely recognized as the world's first influential restaurant critic, Alexandre (Balthazar)-Laurent Grimod de La Reynière (1758-1837) originated the double genre of food critic and restaurant guide by creating eight volumes of the famed "Almanach des Gourmands” (the first public restaurant guides c1803-1812).

"Reynière founded a group of critics that put out a monthly journal that discussed dishes prepared by top Parisian restaurants. Later, he would be accused of accepting bribes in exchange for positive reviews, leading him to cancel his publications and move away from Paris" -KaTom

People often ask why we decided to build an AI restaurant recommendation engine
powered by private reviews. The answer is simple. What was true in the 18th century remains true today: Once reviews or endorsements are made public by an influential party, they become a form of advertising and thus susceptible to commercialization. In other words, publicizing data that impacts the profitability of someone else’s business inherently creates an incentive to manipulate that data.

Existing recommendation engines like Yelp and GoogleMaps, that crowd-source public opinions and then present them as unbiased third-party reporting or search results, are the modern versions of this type of advertising.

Although these review platforms have policies that prohibit (and “systems to detect”) fake and solicited reviews, the data they rely upon is as corrupted as their business model. They rank and assign one-size-fits all grades to restaurants based on their level of popularity - a methodology that is now as archaic as elections in Ancient Greece where the loudest voices were given the most weight.

Curating public opinions make these platforms relatively useless for a small yet increasingly significant percentage of the population whose preferences fall outside the median.

While the practice of giving the loudest voices the most credibility may have proven incredibly profitable during the height of the “keyboard warrior era”, these types of social recommendation apps have been gradually losing credibility with the next generation of users who are looking for reliable information and more personalized service. Yelp (with a declining market cap of $2.94B) is experiencing domestic attrition that outpaces the total number of active users globaly*.

Whether it’s recommending restaurants, lodging accommodations or healthcare practitioners, a recommendation engine that provides most of it’s value on the back-end can be monetized in a way that avoids the conflicts of interest on the front-end that eventually erode a platform’s integrity.

Maintaining a guest-centered approach to every aspect of platform design, including monetization, ensures integrity of the platform and provides an unprecedented level of security for members. We firmly believe building trust and maintaining integrity is integral to maximizing usability and increasing marketability across a multitude of verticals to increase, diversify and sustain revenue streams in the future.

Course Restaurant Guide is an AI Recommendation Engine that doesn’t promote or endorse any specific restaurant. Instead, it uses machine learning to calculate the compatibility between each person’s taste profile and the food & beverage outlets available to them. Unlike current offerings, this technology was developed with foundational security that safeguards data against internal manipulation. We ensure that the only person able to see or change a rating is the person who originally entered that rating.

As an additional layer of security, we’ve built the entire system to be free of any incentive to externally alter or artificially generate data. Simply put - even if an account was artificially created or data was amended, the only impact of that corruption would be contained and confined to the compatibilities calculated for that specific account.

Our methodology and technology are unique assets that put us just ahead of the curve.
The zero-party data we collect is far more actionable than: category-level filtering, extrapolations derived from curated public reviews, “cookies” or online history data that drives much of the targeted advertising we see today - and generally consider “spam”.


*Yelp’s attrition has been attributed to two main factors:

1. Curating reviews of companies while simultaneously selling B2B services to those very same companies is, at best, a conflict of interest...and one that has become more and more obvious to today’s largest group of consumers: Millennials & Gen Zers. 

“According to a Bloomberg report from last year, these young students and professionals command more than $360 billion in disposable income...If a member of Gen Z doesn't agree with the morals of a company, many of them will boycott the products completely and get their friends to do so as well. What advertisers and brand managers learned in their marketing classes years ago is outdated. The best way to learn what Gen Z is looking for from businesses is to ask them.” - Jeff Fromm | Forbes

2. As technology advances, the data curated by these companies (public ratings, “Likes” and online reviews) have, according to experts like those at Harvard Business School, become increasingly less reliable since, online content like this is now bought, sold and artificially generated every day . In fact, the commoditization of ratings has resulted in them determining the system is more hackable than many realize: “1 in every 5 Yelp reviews estimated to be fake”. And, not all algorithm manipulation is limited to the benevolence of members in a facebook group who agree to give each other “Likes” and positive reviews in order to off-set the negative ones. If you have $100 and a Fiverr account, you can buy five hundred 4 star ratings and positive reviews for your own business - or five hundred 1 star ratings and negative reviews for your competitor.