Why We Started ForkFox: The Food Intelligence Story Behind the Algorithm

The Problem No One Named

We kept getting burned. A 4.7-star sushi place where the hamachi tasted freezer-forward. A Michelin-starred tasting menu where the bread course was the only thing worth ordering. A Mission taquería with 400 reviews whose carnitas earned a 92 and whose al pastor earned a 67 — and the reviews never told us which was which.

The problem wasn't the restaurants. It was the way we were discovering them. Review sites rank the room — the ambiance, the service, the one time the waiter was rude, the one time the dessert surprised you. A restaurant is a fuzzy average of unrelated things, which is the same as no information. The scoring felt random. The patterns invisible. And everyone called it taste instead of calling it what it actually was: a system that had no way to measure what mattered.

One of us was in Philadelphia. One was in San Francisco. Different cities, same problem. We kept ending meals more interested in why the plate worked or didn't than in what we'd just eaten. That's when the real question emerged: what if you scored the dish instead of the room? Not as a replacement for knowing a place or loving a place — as a tool for knowing exactly where to look.

How We Started Scoring Dishes

The first prototype was exactly as crude as you'd expect. A scoring framework built in 2024, starting with a Google Sheet and a simple question: what actually makes a dish work? No sample strategy. No pretense of scientific rigor. Just two people in two cities eating, arguing, and trying to articulate what they'd just experienced.

The first version of the framework had too many variables. Flavor, yes. But also: consistency across dishes. Ingredient integrity. Value math — what the plate costs relative to what it delivers. Context fit — how the dish compares to the canonical version of that cuisine anywhere in the world. We cut it, cut it again, and landed on five attributes that actually predicted whether a plate was worth ordering.

Philadelphia became one laboratory. Fishtown. West Philly. The Italian Market. Han Dynasty in Old City. Vetri in Midtown Village. Counter spots on East Passyunk with no website and reservation policies that amounted to whoever showed up first. Ethiopian food along Baltimore Avenue. San Francisco was the other. The Tenderloin's Indian corridor. The Mission's taquería wars. Chinatown dim sum counters that out-score half the city's Michelin-starred dinners. We weren't building a list. We were building a dataset of decisions, not opinions.

The algorithm started noticing things the existing review ecosystem entirely missed. A restaurant could be beloved and mediocre on the metrics that actually predicted longevity. A restaurant could be almost entirely unknown and score in the high eighties on every attribute that mattered. The gap between reputation and performance was the whole point.

What the Data Revealed (and What It Hid)

The first surprise was that consensus was not correlation. A restaurant could be beloved by the city and mediocre on the metrics that actually predicted quality and consistency. A restaurant could be almost entirely unknown and score in the high eighties on every attribute that mattered. The algorithm was identifying things that the public discourse had no name for.

The second surprise was tougher to articulate. The framework was working — but it was working in a way that didn't feel like measurement at all. It felt like translation. We were taking the language of food — flavor, technique, integrity, intention, context — and converting it into a system that could be studied across hundreds of restaurants, across neighborhoods, without losing the actual thing we cared about. Which was: Is this plate real? Does it know what it's doing?

The data started to tell stories. Fishtown wasn't a trend — it was a structural shift in how young chefs could afford to cook. West Philly wasn't nostalgic — it was where the city's actual culinary heritage was being maintained and evolved by the communities that created it. The Italian Market wasn't retro — it was a system of provisioning that still worked better than anything the internet had invented. San Francisco's Tenderloin had dosas scoring 93. Its Chinatown had dim sum counters that out-scored half the city's Michelin-starred dinners. None of those places led with a tasting menu. None had a national press mention. They were the heart of how actual people in those cities actually eat.

The question nobody wanted to ask: was the algorithm any good? The answer turned out to be yes — but only if we were honest about what it actually was. Not a replacement for experience. A way to organize it. Not a truth. A language. A tool for seeing patterns that human attention couldn't quite hold all at once.

What COVID Did to Restaurant Discovery

The pandemic broke the restaurant industry in ways that took years to fully understand — but it also broke the discovery ecosystem in ways that never fully recovered. When restaurants closed en masse in 2020, it became clear which ones people actually missed. Not the most-reviewed. Not the most-photographed. The ones that had understood what they were actually for.

The restaurants that survived weren't always the ones that scored highest on traditional metrics. Vetri weathered through. Beloved institutions decided they couldn't make the mathematics work. Neighborhood spots that nobody had reviewed became essential infrastructure. The pattern that emerged was striking: the restaurants that held on were the ones with real connections to their customers — places where execution wasn't a feature, it was a way of being.

What COVID exposed was that the restaurant discovery problem had been accumulating for years. Review volume had exploded. Attention had fragmented. The actual signal — is this plate worth ordering? — was drowning in noise. The influencer economy rewarded the photogenic over the delicious. Yelp ratings averaged unrelated things into a single number that meant almost nothing. The data on what Philadelphia lost and what remained tells the story better than any argument does.

That's the landscape we built ForkFox into. Not as a COVID project — but with a clear understanding of what the discovery problem had become. The restaurant industry had rebuilt itself. The tools for finding the best plate in a neighborhood had not.

Why Two Cities

We chose San Francisco and Philadelphia on purpose. Not because they're the obvious markets — they're not, really. Los Angeles has more restaurants. New York has more food media. Chicago has a deeper fine-dining scene on paper. But SF and Philly share something the other cities don't, and that something is the reason ForkFox exists.

Both cities have dense immigrant food cultures that the mainstream discovery ecosystem systematically ignores. Philadelphia's Italian Market has Vietnamese corners that out-score destination restaurants. Baltimore Avenue's Ethiopian stretch has been running high eighties on consistency for decades without a single national press mention. San Francisco's Tenderloin has South Indian dosas scoring 93. Its Chinatown has dim sum counters that outperform half the city's Michelin-starred dinners. None of those places lead with a tasting menu. They're the heart of how actual people in those cities actually eat.

The framework had to work across both cities simultaneously, which meant it couldn't be tuned to one place. What "value" means in Philadelphia is not what it means in San Francisco. What "consistency" looks like in Fishtown is not what it looks like in the Mission. The algorithm had to understand context, not impose it. A 92 on a Neapolitan pizza is not the same creature as a 92 on a pho. That cross-cuisine contextualization is the part we're patenting. And it only works if you're building it in two cities that are genuinely different from each other.

Why Transparency Matters More Than Accuracy

The biggest decision we made as a company was to never hide the algorithm. Every score we publish comes with the attributes behind it. Flavor, eighty-four. Value, ninety-two. Context, eighty-eight. Consistency, ninety. This is not how recommendation engines usually work. Usually they hide the logic to appear smarter. We did the opposite.

The reason was simple: if the algorithm is wrong, you need to know why. If a restaurant scores high on our framework but doesn't work for you, you need to be able to see exactly where we diverged. You need to be able to say: "You're right about the flavor. You're wrong about the value. You didn't understand the context the way I do." That is not a failure of the system. That is the system working exactly as designed.

This approach has cost us. Some people think the scores are too complicated. Some think we're overthinking it. Some think a restaurant should be rated on a single metric and called done. But the people who use ForkFox understand what we're actually offering: a way to think about restaurants that's more honest than sentiment and more human than a pure algorithm. A framework that notices things. Not a replacement for experience. A tool for organizing it.

We've also learned that the algorithm notices things that don't fit neatly into the stories we like to tell about food. It notices that a neighborhood can be gentrifying and still have great food. It notices that a very expensive restaurant can have mediocre execution. It notices that an unpretentious spot on a corner can understand flavor at a level that should be unthinkable at the price. It notices what we'd rather not see sometimes. And that's exactly why it matters.

What Comes Next (and What It Means)

We launched our public beta late last month. Two cities. Ten cuisines. Thousands of dishes. The framework has held. The patterns are real. The algorithm is doing the thing we built it to do: make visible what was always there but too distributed for human attention to hold.

But here's what we've learned that matters: the algorithm is not the product. Understanding is the product. We built ForkFox because we got tired of discovering restaurants the way everyone else discovers them — through tourism boards and influencers and review sites that had every incentive to praise and almost no incentive to be honest. We built it because food deserves better than sentiment. Because restaurants deserve better than being judged on a curve that nobody understands. Because you deserve to know not just whether a place is good, but exactly why, and whether that reason matters to you.

The founding story of ForkFox is not a story about technology. It's a story about two people — one in Philadelphia, one in San Francisco — who kept getting burned by high-rated spots that couldn't deliver on the actual plate, and stopped accepting that this was normal. It's a story about building a tool that lets you see what you're looking at instead of just accepting what you're told to see.

That story started with a dish that should have been good and wasn't, and a review ecosystem that had no way to tell us the difference. It ends — if it ends well — with you ordering exactly what you actually wanted.