Published on August 21st, 2019 | by Michael Barnard0
US ITC Solar Tax Credit Decline Is Being Countered With Machine Learning
August 21st, 2019 by Michael Barnard
Recently, I spent some time working through approaches to refining market segmentation and go-to-market messaging for commercial solar with Abid Saifee, VP, Strategic Marketing at the US firm OnSwitch Energy. There are a few insights worth raising from the discussion.
This is an interesting time for solar in the US. At present, the Investment Tax Credit (ITC) for new solar generation capacity is in the fifth of seven years of decline to a steady state in 2022. As a reminder, the equivalent Production Tax Credit (PTC) for new wind generation disappears entirely in 2020.
As the chart from Congressional records shows, the PTC and ITC have been whipsawed by politics multiple times since their introduction. And now they are reaching yet another historic low, when the US has done nothing to eliminate fossil fuel subsidies.
The decline of the ITC in 2020 and 2021 is less problematic, but a more significant downsizing is approaching in 2022. Residential solar will receive no support, while commercial and utility-scale solar tax credits will diminish to 10% of the investment in the solar facility in the year it starts construction.
“Since the ITC was enacted in 2006, the U.S. solar industry has grown by more than 10,000% – creating hundreds of thousands of jobs and investing billions of dollars in the U.S. economy in the process.”
The substantial reduction in US fiscal support is coming at the same time as fiscal lifelines are being thrown to economically non-viable coal and nuclear plants in the country. The logic for finding a path to a sensible retirement date for nuclear is reasonable, but direct fiscal support isn’t the sensible path for that, as Jenkins et al showed in their 2018 study on minor regulatory changes for US nuclear that would increase renewables use, decrease fossil fuel generation and slightly increase nuclear profits, allowing them to persist until their expect end-of-life.
In that environment, companies such as OnSwitch Energy are trying to thrive. It’s a commercial solar deployer based in California which has built projects across the US.
In addition to headwinds from the ITC reduction, commercial solar has other challenges. OnSwitch indicates that mid-day utility pricing is somewhat more of a challenge than the ITC rampdown right now. While solar panel prices have fundamentally commoditized with little quality variation at low prices, soft costs for deployment have not fallen to the same extent and can vary substantially; NREL indicates that only 35% of solar costs are for the hardware. Many commercial buildings are built by developers for lease or sale, and solar would add to the initial costs without being able to be passed on to purchasers or leasers. In addition, there’s a concern about roof quality of existing buildings in terms of both structural ability and replacement timeframe. There are questions about what motivates buying behavior for commercial solar, and who in the organization is buying it. And finally, there are questions about what constitutes a good quote for multi-building campuses, a more common situation than one would think.
Enter OnSwitch Energy. It is a full-service, commercial solar marketplace and project management team with relationships with pre-qualified solar installers with a business model that supports a committed price per kWh as the only cost for 15- and 25-year terms, as well as an option for full capital costs and ongoing maintenance paid by its customers.
But it isn’t just another solar installer. Founded in 2015, its executive team includes engineers with deep big data backgrounds. The company has been working since on a fast and reliable estimation model leveraging machine learning and Google Maps professional data feeds.
Starting with high-resolution images provided by Google’s professional APIs, the company processes the image to identify the space available for solar, automatically lays out the most likely array on the roof, and prices it for a full-service per kWh offering over 15 or 25 years, as well as the full customer acquisition model. The layout is solid and the quote is contractable immediately, but of course customers may choose to change things, as they are wont to do. At present, the AI model only works for the California marketplace, but the team is considering which US states to add in to the model as they move forward. Manual creation of equivalent quotes takes 3-5 days outside of California vs under 24 hours for full California quotes right now.
This speed to quote is a critical part of its strategic vision, but like many technology centric firms, it is finding that its technical secret sauce isn’t necessarily a sales differentiator.
During our discussion, the reduction in the ITC was obviously a factor, but it’s not impacting sales and deployments yet. It’s a strategic concern, of course.
Similarly, the concerns related to the potential market size given the number of buildings not owned by the occupants are present, but per Abid, only 3% of commercial rooftops in the USA have solar right now. The market is far from saturated, and per Wood MacKenzie’s GreenTech Media, provision of services is fragmented, with multiple vendors often involved in a single project. Perhaps unfortunately for OnSwitch, it is competing with Tesla’s solar division, which is second largest in the market after SunPower.
There are two opportunities there that OnSwitch intends to capitalize on. The first is the untapped market itself, obviously. The second, however, is to provide one-stop shopping and rapid quoting for commercial buyers, reducing the complexity of the purchasing, construction and utility-integration process substantially. The quoting solution is core to that.
Roof quality is less of a concern than I had understood, as roof replacement can be bundled into the ITC submission, with obvious economic benefits. However, this means that the cost factor increases for a portion of the market as the ITC diminishes as well.
Where the conversation started to become interesting was in the related questions of which people within corporations were the primary buyers and what would motivate them to buy.
This is a screenshot of the company’s online quote example. A few things emerged as Abid and I discussed it.
The first is that the most compelling argument on this entire page is that California’s commercial price of electricity is 13.8 cents USD and its 25-year solar electricity service is only 51% of that, while its 15-year plan is 62%. For buyers who are motivated by operating expenditure optimization, these can be big numbers. The most compelling argument from the quoting system is buried in the presentation right now.
But the question that immediately sprang to mind was which customers that might be true for.
Many companies are in a steady state, low-growth phase, and in companies like this where electricity is a high-percentage of expenses and expense optimization is the primary concern, the CEO would be very interested in a substantial expense reduction. Many mid-tier electronics manufacturers of commodity computer components and computer systems would fit into this category. They are often using labor arbitrage into low-cost markets as a key wedge and are concerned enough about electricity costs that they optimizing lighting and climate control for their staff using IOT sensors integrated to environmental control systems, but many of them have legacy assembly lines in the US still in fully depreciated buildings. These potential customers are fairly well served by the price-focused, rapid-quoting system OnSwitch has developed.
The sample quote is for Infinera, a California-based vertically integrated manufacturer of Wavelength division multiplexing optical transmission equipment for the telecommunications service provider market. Simple customer research shows that the company grew revenue by US$200 million, roughly 27% year over year from 2017 to 2018, and that it is in a high-margin portion of electronic manufacturing. This implies a couple of things. The first is that it is a growth-focused company, and so more interested in above the line increases in revenue more than below the line reductions in expenses at this point in the company’s trajectory. The second is that it is in the electronics fabrication business, which has a very high electricity consumption due to the ovens that bake-in components on every assembly line. The combination means that for that type of customer, OnSwitch would reasonably target not the CEO but an operations executive owner in the corporation. They would have targets for cost efficiencies as part of their KPIs and likely know the rough percentage of electricity as a portion of overall expenses, whereas the CEO would likely not given the strategic stance of the company. Finding the right contact in these firms is more of a challenge.
Many corporations have multiple buildings, either in a campus or spread across different sites in a city or state. These buyers have much bigger potential annual savings if all buildings were considered, something that OnSwitch currently doesn’t do automatically, but which has great potential for large customer acquisition. It’s made easier by the solution that the company has developed, but it’s not flawless. A subset of these multi-building owners are a high-value potential target: those which already have solar on one or more of their buildings, but have untapped rooftop potential in their building portfolio. They have already internalized solar as a solution, but may not have been presented a simple path to extending them. OnSwitch’s technology allows much more rapid identification of this segment, a significant boon in focusing their marketing and sales, and it has a portal specifically for larger customers which they are finding beneficial in the sales process.
Other companies have a very different buying behavior for commercial solar, where the cost of the electricity is not the point, but the brand value of reducing their carbon footprint. Whether it’s a company like REI where the environment is baked into their brand, or a fossil fuels or plastics company which needs to improve their environmental image, the cost of rooftop solar is less important than the marketing value and the carbon emission reductions, but there’s currently no direct line from OnSwitch quoting that assists these customers. It’s certainly doable, but once again, manual assembly of quotes is required, hindering the sales process somewhat.
The larger points of this solar supplier case study are three-fold. First, the market is far from saturated with significant potential growth, but also a lot of fragmented providers who have to assemble complex bids. Second, the shift in economics of the ITC rolldown need to be addressed by efficiencies in commercial solar sales companies. Solar panels are not the majority of the cost, so process efficiencies such as OnSwitch’s AI quoting module are necessary to maintain competitive prices. Third, the market’s buying behavior is varied, and go-to-market strategies have to be aligned to that.
Commercial solar has significant untapped potential in the USA. Innovative companies such as OnSwitch are working to maximize that, regardless of federal policies.