Solar opportunities: Theory and projections for entrepreneurs

79
picture of solar innovation curves.

Green opportunities

I’ve been tracking the annual drop in the price of solar power and what I keep seeing amazes me, and gives me great comfort. Let me explain.

It seems that the price per watt for solar keeps getting better from year to year. Much of the product innovation continues, but that is no longer the main driver. The reduction in cost now has more to do with economies of scale and the implementation of process innovations at the manufacturing and installation stages.

Not surprisingly the solar industry has been growing. Nowadays about 260,000 Americans work in this industry, which constitutes a 17-fold increase since 2010. Just in 2016 alone, the solar industry added 51,000 jobs, a 24.5% increase over 2015, and accounting for 2% of all new jobs created in the US.

Innovation theory

The innovation cycle starts with product innovation. We’ve had those gains from solar already—the yields on new ‘commodity’ panels are better than those placed on early satellites.

The second curve, that of process innovation, started around the time that process innovations made it feasible to try to achieve economies of scale and climb the associated learning curves.

The arrow in the figure denotes about where we are today: right near the end of the product innovation curve, but beyond the point where product and process innovations were essentially equal contributors. Process innovations have accounted for the majority of recent gains as huge factories in the U.S., but especially in China, have upped their production and learned to make better panels more cheaply.

The demand keeps growing as lower prices make whole new regions viable for solar installations. The more sunlight a place gets, and the further away from cheaper alternatives, the faster the diffusion. But with each additional place, there come more process innovations, increasing the pace of the feasibility wave.

The bet

This leads me to place a big rhetorical bet. Given that no other energy technology comes close to the improvement curve in solar, and that process innovations are going to continue driving demand, we should bet on solar. Technology scholars often note that people are not very good at making estimates for non-linear curves. Early improvements seem slow when they start because they are starting from such a low point. They still look slow right now, but if we are at 1% global capacity, and we go to 2% next year and 4% the following year, then the jump to 8% will perhaps be the first one to really get attention. But by then, the time to invest will be over because by then everyone will believe it, and valuations will reflect it.

Solar enterprise opportunities

Entrepreneurs can get into every level. They can learn from the gains of big production facilities and build new facilities with newer, cheaper technologies. They can get into the business of installing and maintaining panels, repairing them, and optimizing them. They get into the distribution or sale of the panels. They can resell panels with marketing startups. They can finance projects in search of returns; even buy options on their neighbor’s roofs. Now seems to be the time.

Some opportunities we scored

While only one type of solar venture, Launchscore estimates the potential yearly earnings of Solar Energy Systems Dealers in 750 U.S. cities. The statistical model suggests that there are over 100 cities where the entrepreneurs could make over $100,000 in net profits within three years of entering the business, and also 100’s of cities where it is not yet viable as a business choice. Not surprisingly, some of the worst places are cloudy, like Bridgeport, Connecticut with 99 sunny days a year, and some of the best are in sunny places like Cupertino, California, with 265 days a year. Note that some of the sunniest places of all like Phoenix (AZ) are not in the top 100 because opportunities have been strong enough there for long enough that dealers are already ubiquitous (sorry but you need to sign in to see this one). But we didn’t just factor in sunny days. We used 30 other control variables, including demographic factors (e.g., median age and income), elevation, location (East, West, North, South), weather more generally, and cultural factors (e.g., red versus blue states) to make our estimates.