Before we had SolarCity and Vivint, there was Apple and Commodore.
In separate industries and decades apart, they all pursued a similar strategy: vertically integrated development and distribution of an emerging technology. That "do it in-house" approach helped advance personal computing from its earliest stages, but it ran into trouble bringing down costs.
The transition from that model to a more horizontal and specialized industry holds valuable lessons for the residential solar industry at its current inflection point, according to Will Anderson, founder and CEO of Complete Solar, a project management platform.
The national, vertically integrated installers blew open the rooftop solar market by combining panel procurement, design, financing, installation and maintenance under one roof. They used economies of scale to nudge down costs.
That growth spurt has ended. The three most successful national installers have begun losing market share to the so-called "long tail" of local and regional installers that can acquire customers more competitively than the out-of-town behemoths with higher overhead costs.
Market leader SolarCity is slowing installations as part of Tesla. Runner-up Sunrun employs a two-track system, running its own installations and partnering with an installer network; that company posted growth and profits last quarter. No. 3 Vivint has pulled back on growth to try making more money.
Several top 10 installers went bankrupt or quit the business this year.
All of this raises questions about what’s coming next and who will succeed in the long run. Amid that uncertainty, the record of early computing offers surprising parallels and hints at what the future may hold.
Vertical integration gets the market started, but hits a limit
The personal computing industry started to gain steam in the late 1970s. The technology had been developing in government labs for years, but it eventually jumped into homes and businesses.
The dominant companies controlled each step of the value chain: the box, the hardware in the box, the operating system to run it all, the software that uses the operating system and the distribution channel to the customer.
These pioneers, like Apple and Commodore, succeeded in pushing adoption to around 8 percent by 1984. The cost of doing everything in-house, though, kept the PC out of reach for most households. If the industry didn't change, it would be a while before mass consumption became possible.
"The big shift and the big step-function change in the industry happened when Microsoft came in and offered an operating system to be licensed to anyone who wanted to use it," Anderson said. "It was very hard to build a good OS. It removed a tremendous barrier to entry for other companies that would otherwise want to be PC manufacturers."
Opening up the industry to specialization lowers costs
Once Microsoft removed that barrier, it allowed hundreds of new entrants to try their own approach to building a PC that runs Windows.
Some focused on optimizing particular components, like Intel with microprocessors. Gateway and Compaq found ways to reduce the cost of hardware compared to the incumbents. Dell tapped into the direct-to-consumer market.
The open-platform model encouraged innovation among a vast number of companies that no longer had to do everything in order to compete.
“Specialization is going to enable companies to focus on their core competency, and it’s going to improve their ability to really take advantage of that core competency without having the benefits of that muted by weaknesses in other areas,” Anderson said.
While Apple chased specialty markets like graphic designers and education, Windows computers got cheaper and more accessible for everyone else. Apple might have faded slowly from history, if it weren’t for a game-changing new product years later.
This change is happening now in residential solar
The PC industry transitioned from vertically integrated to decentralized and specialized in about four years, Anderson noted.
“It was fast and it happened before people really realized that it had happened,” he said. “As soon as it gains momentum, the structural advantages delivered by this model make it self-propagating. We’ve already started down that path in solar, and in a few years we’ll see most sales and installs going through this model.”
SolarCity has already begun to fade. To play out the Apple comparison, the top solar installer needs to invent an iPhone to make itself relevant again. Maybe the solar roof does that; maybe it remains a niche product and fails to remake consumer demand in its image. A technology innovation approach seems likely under Tesla.
As for the platform that opens itself to all comers, there isn’t a clear heir apparent yet. In its final months, Sungevity positioned itself as a solar platform company, but went bankrupt before it could execute on that vision.
Sunrun incorporates elements of an open platform in its installer partner business, but it holds onto direct installations as a major revenue driver.
Many smaller companies are beginning to use both inside and outside sales teams, said Allison Mond, a solar analyst at GTM Research.
“Without a clearly defined process and software to manage customers, there will be friction as companies make this change,” she said. “I expect that in five years, the market will look very different, with a large portion of the market doing either sales or installations, and not both.”
It’s that transition that spurred Anderson’s fascination with the business history of computing.
“The big question is who’s Microsoft in all this, and that’s what I’m striving to become,” he said.
Let the experimentation begin
Complete started in 2010 as traditional solar company that focused on fast installations and keeping customers happy, and it built a software platform to manage that.
When the company tried to grow by expanding its sale team, costs shot up too. Anderson decided that the firm was more competitive in project management and fulfillment. He pivoted to focus on those strengths.
Now grown to 60 employees, Complete provides a platform for solar specialization. Sales teams that don’t want to bother with designing, installing and financing systems can log on and hand off their deals to Complete staff to fulfill.
Acquiring closed customer contracts as opposed to open leads keeps customer-acquisition costs low, Mond said.
Installers sign up by showing credentials and licensing, and then start getting assigned projects. They document their work in the app, and the quality of their performance influences the future assignments they receive.
Financiers use the system to connect with customers, after which the platform collects the necessary disclosures and documentation.
Close to 50 partners have joined across California, Utah and New Jersey, with Massachusetts coming soon, Anderson said. Some have experimented and failed, but close to two dozen partners encompassing some 500 sales reps and installers are actively producing deal flow and building a business around it.
Sales teams on the platform have already started experimenting with different strategies, he added. Flex Home Living, a division of Flextronics, is trying to expand its smart home platform, called Wink, by selling solar systems.
Another partner, Garantia Solar, caters to Spanish-speaking customers with ads on Spanish language radio and a call center in Mexico. Its customer-acquisition costs are one-third of what public companies face, and it’s scaling, Anderson said.
Some partners have something that works locally and doesn’t need to scale, like a community networking firm that gets solid leads within its territory.
"It’s not a pendulum swinging back to a previous industry state where things were less efficient," said Anderson. "It’s an evolution that looks different from the fragmented market that existed seven or eight years ago.”
If he pulls off his Microsoft-for-solar ambitions, the industry won’t simply become more localized. It will outsource the tasks that can be done more cheaply and efficiently by tech-savvy software startups, so that the hometown installers can focus on the tasks that they do best.Lead image courtesy of Alessandro Grusso, license available here.