Conversation
Highlights

How ClearCOGS Discovered Their Product by Asking “Why” Until Every Feature Request Collapsed Into One Problem

Five assistant managers across five Jimmy John’s locations, all guessing how much bread to bake. Matt Wampler watched thousands of dollars disappear into dumpsters while his 21-year-old self worked 110-hour weeks—sometimes pulling 24-hour shifts three times per week—just to keep the operation running.

“I don’t scale,” Matt realized. The inefficiency wasn’t just costing money. It was the fundamental constraint preventing any restaurant operator from growing beyond their personal capacity to be everywhere at once.

That constraint eventually led Matt to co-found ClearCOGS, a demand planning platform that’s raised $3.8 million to create an entirely new category in restaurant technology. In a recent episode of Category Visionaries, Matt shared how building ClearCOGS required systematically unlearning everything he thought restaurant operators needed—and discovering what they actually needed instead.

The Discovery Process That Revealed the Real Problem

Matt entered product discovery with strong opinions. Years running Jimmy John’s franchises gave him a mental backlog of features he wanted as an operator. But his technical co-founder approached customer conversations differently.

Restaurant operators would request dashboards for specific metrics. Matt would nod along—these made sense to him. Then his co-founder would ask: “Why do you want that dashboard?”

They’d explain they needed visibility into certain data. “Why do you need to see that?”

This interrogation repeated across dozens of conversations with different restaurant brands. Every time, the questioning led to the same answer. “It would always come down to the same thing,” Matt explains. “I don’t know who is going to walk in the door tomorrow. And because of that, I don’t know how much food to prep. I don’t really know what to order and I don’t know who to staff.”

The inefficiency wasn’t abstract. “I make inefficient decisions on those,” Matt notes. “And it either, if I get them right, means we make money, and if I get them wrong, I struggle to make payroll.”

This discovery process revealed something critical about building B2B products in industries you know well. “It’s this interesting yin and yang of you need to have experience and domain expertise, but you also can’t be too close to the problem,” Matt says. Domain expertise earned him credibility and access. But his co-founder’s relentless outside questioning cut through the feature requests to expose the core constraint.

Reframing a Complex Category Through Cognitive Dissonance

Once ClearCOGS identified uncertainty as the core problem, they faced a bigger challenge: the category didn’t exist. “Restaurants are still reactive,” Matt explains. “When we think about software in restaurants or reports or how do I manage it’s all retrospective. There’s no proactive prescriptive analytics.”

Every existing restaurant software tool showed what happened yesterday. Nothing helped operators make better decisions about tomorrow. But “predictive analytics” didn’t resonate. “Doesn’t resonate with them at all like it is. It’s very nerdy,” Matt admits.

His breakthrough came from observing what restaurant operators already did in their personal lives. “The average guy out there running a restaurant probably got a fantasy football team,” Matt notes. These operators obsess over probabilistic forecasts for their fantasy lineups. They cross-reference injury reports, weather conditions, historical matchups, opponent defenses.

“It’s very mainstream in how we make decisions, but it’s not when it comes to how do we run our business,” Matt says. “And it’s really ironic that we have more analytics and forecasting for our fantasy football than we do for our business.”

This wasn’t category education through explanation—it was category creation through exposing absurdity. The fantasy football analogy made the gap visceral. Operators immediately understood the cognitive dissonance of having sophisticated forecasting for leisure while running six-figure operations on Excel sheets and gut instinct.

Building Infrastructure That Processes 100 Million Variables Daily

Today, ClearCOGS’s technical architecture processes “about 100 million data points every day per decision, per restaurant.” The platform ingests granular point-of-sale data and cross-references it with dozens of external variables to generate forecasts.

The system analyzes POS data at 15-minute intervals going back years. For a brisket that needs smoking tonight for tomorrow’s service: “We’ll look at how much brisket you’re using every 15 minutes and have for the past five or 10 years,” Matt explains. “Then you cross reference that with things like weather. How’s weather affect my brisket usage?”

The data reveals patterns that operators would never discover manually. One Florida-based restaurant brand discovered dew point—not just temperature or precipitation—significantly impacted their demand. “If it wasn’t humid out, people would go outside. But if it was really humid, they didn’t,” Matt says. High humidity kept customers off the patio, creating predictable shifts in ordering patterns and prep requirements.

But the platform goes beyond prediction to optimization. “The real question is not tell me a number to do tomorrow, but what’s my goal as an operator?” Matt explains. Different restaurants have different objectives. Some want to run out of items daily to maximize freshness. Others want to minimize stockouts at all costs. Some want profit maximization, accepting occasional waste as the cost of never missing sales. “There’s a magic number there for every business.”

Converting Forecast Errors Into Customer Intelligence Touchpoints

No probabilistic system achieves perfect accuracy, and ClearCOGS’s predictions sometimes miss. The team’s initial response was technical: “We used to constantly stress out about this. We got something wrong, let’s go. You re optimize the models and look at this, that the other, you know, and just frantically spin around for a week.”

Then they tried something operationally simpler but psychologically harder. “One day it was, how about we just call them and apologize and tell them we’re on it and working on a fix,” Matt recalls. “And it meant so much to the customer that they had somebody on their team that was watching and cared.”

This shift reframed errors from technical failures to relationship opportunities. “How do you use that as an opportunity to let the customer know that you’re paying attention because so many softwares don’t,” Matt says.

The approach works because of baseline expectations. Restaurant operators are “really just guessing using Excel spreadsheets,” Matt notes. “Going from that to probabilistic machine learning, time series forecasting, you know, like, it’s like going from the horse and buggy to the automobile. It’s already so much better than they’re used to.”

In a software category where vendors typically disappear after contracts are signed, proactive acknowledgment of misses became differentiation. Every error became a touchpoint to demonstrate attentiveness that competitors structurally couldn’t match.

Segmenting by Incentive Structure Rather Than Department

ClearCOGS discovered their messaging framework wasn’t based on typical B2B segmentation. “Interestingly enough, it’s less to do with department and more to do with the organization,” Matt explains.

The key split wasn’t finance versus operations—it was franchisors versus franchisees. These groups exist in the same organizations but operate under fundamentally different incentive structures.

“You’ve got a brand, the franchisor whose job it is to vet all of these technologies. Those franchisors make their money off top line revenues,” Matt says. Franchisors earn royalties as a percentage of sales, making consistency and scalability their primary concerns.

“Whereas there’s the on the ground managers and owners, you know, they make, they live on the bottom line,” Matt continues. Franchisees and operators earn profit after expenses, making waste reduction and margin improvement their primary concerns.

Same product. Same problem being solved. But different value propositions required based on how buyers were compensated, not which department they sat in. “We slightly different messagings towards hey, are you somebody that lives on the bottom line or are you a franchisor that just needs consistent operations and scalable solutions?”

Building in Public as Enterprise Lead Generation

Matt started ClearCOGS with zero social media presence. “No Instagram, no Facebook, no Snapchat, no TikTok. I had like 500 people on LinkedIn I was connected with, and I never logged on,” he admits. “I did not like social media.”

Building in public felt unnatural. “It was really hard to put myself out there, and I second guessed a lot of it,” Matt says. But he started posting about ClearCOGS’s struggles and failures on LinkedIn. The impact came fast: “Within six weeks a giant brand had reached out and said, I’m interested in what you’re doing and set up a call.”

Later, Matt posted ClearCOGS’s first website draft publicly, asking for brutal feedback. “I must have had 50 people reach out to people, record videos, going over things. Lawyers would come in and be like, by the way, make sure you put this disclaimer in this spot for this purpose.”

The vulnerability created asymmetric enterprise access that paid channels couldn’t replicate. “In this glossy world where everything is overproduced, when somebody gets on social media and says something real and is vulnerable and shares failure,” Matt observes, it cut through competitor noise in ways polished marketing never could.

His co-founder initially resisted. “My co founder being like, why would you ever let anybody see you down or when something went wrong, like, no, we’re in startups. You have to always be positive and you know you’re taking over the world and nothing can stop you.”

But Matt found the personal benefits transcended business metrics. “I have found more real relationships with people who really matter in my life because I did put myself out there, and I will be forever grateful for that.”

Investing in Category Education Infrastructure Without Clear Attribution

ClearCOGS launched the Restaurant AI podcast despite having no clear ROI model. “The biggest debate was whether or not we should do a podcast,” Matt recalls. “It’s a lot of time, it’s a lot of energy and you know, frankly, there’s not a direct, clear ROI on it.”

The decision came down to category positioning over lead generation metrics. “Everybody is so hungry for information on AI right now. So like, we wanted to own that one specific niche,” Matt explains. “We’re trying to help define this category. There is no label right now, so we’re working towards that.”

The podcast became long-term infrastructure investment rather than demand generation channel. Matt views it as practicing a muscle that compounds over time. “If you just go do an interview once every two weeks, you know, you keep the juices flowing, you stay in shape,” he says. “Start working the muscle now because no one’s listening to you when you first start a company. So go make a bunch of mistakes, fail fast, learn, put yourself out there.”

The Long-Term Vision: Automating Operations, Not Operators

Matt’s vision for ClearCOGS extends beyond forecasting to fundamentally restructuring how restaurants operate. “Nobody gets into restaurants because they want to manage the numbers. They don’t want to do the inventory or work on spreadsheets. This is the people business,” he says. “It’s people managing people, serving people, trying to create experiences.”

Current restaurant management requires dual expertise: exceptional people skills and analytical operational capabilities. Most operators excel at one but struggle with the other. “Right now we ask our managers in restaurants to have to do both,” Matt notes.

AI shifts this constraint. “AI really does—I mean, it’s never going to create those magic experiences for the guest, but man, it’s good at the numbers,” Matt says.

The opportunity is role specialization at scale. “If we can automate the financial metrics, basically give me that fractional CFO COO that sits on top of my data so that I can start bringing in general managers that are great with people and can spend their time there, I think fundamentally we’ll just be a better industry.”

It’s the same insight that started ClearCOGS: Matt himself didn’t scale across five locations because he couldn’t be everywhere simultaneously making both people decisions and analytical decisions. The solution isn’t better operators—it’s separating responsibilities that AI can handle from responsibilities that require human judgment, allowing each to operate at their highest use.