CleanJoule’s Three-Question Framework for Spending Three Years in “Dark Alleys”

Most startup advice tells you to build fast and validate quickly. Ship an MVP in weeks, get customer feedback, iterate toward product-market fit.

In a recent episode of Category Visionaries, Mukund Karanjikar, CEO of CleanJoule, a sustainable aviation fuel producer that’s raised $55 million, described a radically different approach: his team spent three years not building anything. Just exploring, learning, and going down what he calls “many dark alleys, coming back, restarting.”

Those three years from 2009 to 2012 weren’t wasted time. They were the minimum viable exploration period to frame the problem correctly. And they were guided by three questions that prevented CleanJoule from spending a decade solving the wrong problem.

The Four Walls and a Roof Phase

When Mukund arrived in Salt Lake City in 2009 after leaving his oil and gas job in Houston, CleanJoule had no defined plan. Just a chemical engineer with molecular transformation expertise, a serial entrepreneur in his eighties, and conviction that human civilization shouldn’t be built around petroleum.

“All we had was four walls and the roof,” Mukund recalls. “Then the first three years were lot of ideation as to what to go after, what’s important.”

This is the phase most founders skip or rush through. The temptation is to pick something—anything—and start building. Market research says there’s demand for X, so build X. Investors are excited about Y, so pivot to Y.

CleanJoule resisted that pressure. They spent three years in exploration mode, guided by three core questions.

Question One: What Can We Meaningfully Contribute?

The first question wasn’t “what’s the market opportunity” or “what’s trending.” It was about honest assessment of capability.

“With my sort of a mental frame of those asset of expertise,” Mukund explains, describing his approach to the problem. “Being a chemical engineer, I know how to build molecules, right? So I take feedstock and we do what we call in our jargon, molecular transformation. So a becomes b, and b does not look like anything that a started out with.”

This is the expertise inventory phase. What can you actually do that others can’t? Not what sounds interesting or what seems lucrative, but what sits at the intersection of your technical capability and hard unsolved problems.

For Mukund, that capability was molecular transformation at scale. He knew how to take feedstocks and create entirely different compounds. That was his competitive advantage, built through years in oil and gas.

The trap most technical founders fall into is assuming their expertise is common. They think “everyone knows how to do X” when X is actually rare capability that took years to develop. Mukund didn’t make that mistake. He inventoried his molecular transformation expertise and used it as the foundation for problem selection.

Question Two: What’s Relevant?

Capability alone doesn’t determine direction. Plenty of technical problems are solvable but don’t matter. The second question was about timing and market relevance.

“Those three years were lot of going down many dark alleys, coming back, restarting, and we were confident,” Mukund says. “That most important piece of those three years, I remember, is we were learning every day there was something new every day. And that’s what kept us going.”

The learning wasn’t random exploration. It was testing hypotheses about what problems were becoming relevant in ways that would create tailwinds for their solution.

In 2009, right after the financial crisis, sustainability wasn’t the dominant narrative it is today. But CleanJoule could see regulatory pressure building, climate commitments emerging, and industries beginning to think seriously about decarbonization.

“I was always driven by sustainability and impact of human operations on sustainability,” Mukund explains. “So we decided to jump into what is hard to decarbonize and long haul transport is very hard to decarbonize.”

The relevance question forced them to think not just about current demand but about where the world was heading. Aviation fuel in 2009 didn’t have urgent demand for sustainable alternatives. But looking toward 2030, 2040, 2050, it was clear that aviation would need to decarbonize and had no obvious solution.

Question Three: What Would Be Impactful?

The third question was about scale and significance. Not just relevant problems, but problems where solving them would meaningfully change outcomes.

“Energy dense hydrocarbon building for mature industries is the most impactful aspect of pursuing sustainability from the viewpoint of what is that we are going to disrupt,” Mukund says. “And that would be building molecules in a new way from different and sustainable feedstock.”

Impact has two dimensions in this framework. First, the magnitude of the problem. Aviation contributes significantly to carbon emissions and is considered one of the hardest sectors to decarbonize. Solving it matters at civilization scale.

Second, the leverage of the solution. By focusing on drop-in compatible fuel that requires no infrastructure changes, CleanJoule maximized the potential impact of their technology. Airlines could adopt it without redesigning aircraft or rebuilding distribution systems.

“We wanted to turn the paradigm of fossil based human life to sustainable, raw material based human life. And aviation is right in the middle of it,” Mukund explains about choosing aviation as the target industry.

How the Framework Eliminated Options

The power of this three-question framework isn’t just what it helped CleanJoule choose. It’s what it helped them eliminate.

Could Mukund have built something in software? Sure, but that fails question one about meaningful contribution. He couldn’t write code to save his life.

Could they have focused on easier-to-decarbonize sectors? Yes, but that fails question three about impact. The whole point was tackling hard problems that would uplift civilization.

Could they have targeted consumer applications instead of industrial scale? Possibly, but that fails question two about relevance. The meaningful opportunity was in decarbonizing mature industries buying fuel by the billions of gallons.

The three questions acted as filters, eliminating paths that might have been interesting but wouldn’t align capability, relevance, and impact.

The Dark Alley Navigation Strategy

Those three years weren’t linear progression. They were iterative exploration with lots of dead ends.

“Those three years were lot of going down many dark alleys, coming back, restarting,” Mukund recalls. But even the dead ends served a purpose: “We were learning every day there was something new every day. And that’s what kept us going.”

This is the psychological shift required for deep tech problem framing. In software, failed experiments mean lost weeks. In deep tech, failed experiments are data about what won’t work, which narrows the search space for what will.

The framework prevented them from getting lost in the dark alleys. Every exploration could be evaluated against the three questions. Does this leverage our capability? Is this relevant to where markets are heading? Would solving this be impactful at scale?

If the answer to any question was no, they could abandon that alley and explore another.

The 2012 Breakthrough

By 2012, the three questions had converged on a clear answer: sustainable aviation fuel built through molecular transformation of biomass feedstocks.

“By then in at least framing the problem, we thought we hit gold,” Mukund says about reaching that three-year milestone.

Notice the language: “framing the problem,” not “finding the solution.” Those three years weren’t about building technology. They were about understanding what question they were trying to answer.

Once the problem was correctly framed, the next 10+ years could be spent solving it. But if they’d picked the wrong problem in 2009 and spent a decade building toward it, they’d have wasted far more time than the three years spent in exploration.

The First Validation

In 2012, CleanJoule won their first government contract. It wasn’t a massive deal, but it validated the problem framing.

“It wasn’t a very large contract, but it was an instant gratification and instant ratification that other people” saw value in what they were pursuing, Mukund explains. The Department of Defense agreed that sustainable aviation fuel for military applications was worth funding.

That external validation confirmed what the three-question framework had suggested: they’d identified a problem at the intersection of their capability, market relevance, and potential impact.

Applying the Framework

For deep tech founders facing the dark alley phase, CleanJoule’s three-question framework offers a navigation strategy:

Start with honest capability inventory. What can you meaningfully contribute based on years of accumulated expertise? Not what sounds interesting, but what you can actually do better than others.

Layer in relevance assessment. Where are markets, regulations, and customer needs heading? What problems are becoming urgent that weren’t five years ago?

Filter for impact. Of the problems you could solve, which ones matter at scale? Where would success meaningfully change outcomes rather than create incremental improvements?

The intersection of those three questions is where to focus exploration. And if you spend three years going down dark alleys before building, that’s not wasted time. It’s the minimum viable exploration period to avoid wasting the next decade solving the wrong problem.

“By 2012, we thought we hit gold in at least framing the problem,” Mukund says. Fifteen years later, CleanJoule is on track for commercial manufacturing by 2029, having avoided the costly mistake of building the wrong thing quickly.