Throughout 2017, I was continually finding the smart home discussion to be overly focused on services providers in a cluttered field of competition and so I started to dig into the other unit drivers for wholesale market adoption. In the consumer-facing world, targeting and identifying customer needs can be a daunting task but when the area of interest is consumer home energy consumption, targeting this demographic is even more daunting.
For nearly 10 years, the concept of financing residential and commercial smart home projects with innovative green bonds has been a growing phenomenon. For investors, the technology-driven smart home improvements market is over $60.0B and the market is expected to grow to over $100.0B by 2021.
Emerging digital platforms like PowerScout are pioneers in identifying which homeowners are best suited to leverage these smart home technologies to increase their home value, reap the inherent cost savings and mitigate their carbon footprint. Think Nerd Wallet or Credit Karma for smart home improvement. PowerScout has mastered the customer acquisition process for various home improvement products and services through a targeted digital platform, rich data and machine learning. PowerScout has visibility into billions of data points that map out consumers energy profiles in ways that utilities could only wish to understand.
I recently sat down with Attila Toth, the founder/CEO of PowerScout, a former McKinsey, C3IoT and Sun Edison executive who has built a team of formidable energy and data scientists in Oakland, CA.
Share with us the genesis story of your company?
According to the US Green Building Council, buildings in the United States are responsible for nearly 40% of the nation’s carbon emissions. Despite rapid advances in “Smart Home” technologies that can help solve this problem- rooftop solar, battery storage, energy efficient HVAC, EV charging, and smart home controls- penetration remains painfully low. In the early part of the adoption curve for any new technology, businesses need to take the lead in educating consumers on the benefits of the technology and how to integrate it into their daily lives. That educational process happens as companies spend sales and marketing dollars.
Given that Smart Home technologies have complex qualification requirements and variable consumer benefits, however, a huge portion of those sales and marketing dollars are spent on the wrong consumers. For my co-founders and me, this pain point was summed up as follows: If you buy a Toyota Camry today, a mere 4% of the $25K price you pay is covering sales and marketing costs. Contrast that with 13% for a new roof, or 49% for a home security system. In 2015, the four publicly traded residential solar companies reported customer acquisition costs of 26%- over $5,000 on an average purchase of $20,000- and it became the catalyst for launching PowerScout.
When my co-founders and I embarked on this journey two and a half years ago, our objective was to capitalize on the step change advances in machine learning and the volumes of available consumer data. We set out to build algorithms that would match the right consumer with the right service provider offering the right product at the right time, wrapped in a simple user interface with transparent pricing. While the industry leaders used door-to-door sales tactics from the 1960’s playbook, we wanted to bring the full power of artificial intelligence and digital marketing to bear on driving the adoption of Smart Home technologies.
Today, PowerScout is an artificial intelligence enabled data analytics platform that leverages 100+Bn data points on 45 million homes and their residents. We use the latest advancements in computer vision and deep learning to extract building characteristics, such as, roof geometry, orientation, building distance to vegetation, among other features, with unprecedented accuracy. Leveraging these data we help our partners effectively market products and services related to the home. We started with a focus on rooftop solar and now, after building a strong network of partners on both coasts of the US, we are expanding into additional verticals.
You have a interesting background, where did you grow up?
I was born behind the “Iron Curtain” in a small Hungarian town on the Austrian border. Even as a kid, I wasn’t a big fan of Communism. It was too much ideology that didn’t make sense, with too little freedom of thought and too many long lines in font of meagerly stocked stores. The latter sparked me to start two ventures as a teenager: a currency exchange and a pesticide distribution business. Entrepreneurship was a way to make money by solving problems for others, and I was hooked.
I came to the US in my twenties, and since arriving I have enjoyed a number of formational professional experiences. After earning my MBA from the Kellogg School of Management, I joined McKinsey and Company where I learned how to solve complex business problems and communicate effectively with senior executives. I later became part of a hypergrowth solar company, SunEdison, where I learned about how to scale a global business. The most impactful experience was working closely with Tom Siebel, the legendary founder of Siebel Systems and C3 IoT, the leading industrial internet-of-things platform. Working with Tom, I began to share his conviction that the wave of change unleashed by big data and artificial intelligence will touch every corner of every industry. I am also grateful to Tom for learning a lot about the nuanced process of driving effective enterprise sales.
What ethos do you lead by to keep your team focused?
My personal mantra for the PowerScout team comes from W. Edwards Deming: “In God we trust, all others bring data”. It’s a tongue-in-cheek way to remind ourselves that the future belongs to those who form hypotheses, test the outcomes, and learn from the results. Data will power the next century like oil powered the last century.
Technology is advancing faster today than at any prior point in human history. Cloud computing and machine learning, among other advancements, are making data capture and data analysis possible at a scale never seen before. Being able to aggregate existing data at scale has become an absolute necessity for businesses to remain competitive; you need to be able to generate data that didn’t exist before and derive insights in real time. As machine learning algorithms take over this analysis from humans, the cycle times get even shorter.
I believe PowerScout’s success will be predicated on our ability to synthesize novel and disparate data sets to produce unique insights. For example, we’re combining satellite, aerial, and Light Detection and Ranging (LiDAR) imagery for millions of homes with proprietary data about the demographics and psychographics of their inhabitants to produce a unique understanding of the connections between people, places, and properties. We form hypotheses based on the data, test new marketing strategies, and evaluate the results. The results become part of our data, and we begin the process all over again.
Share with us some your most critical experiences starting out, raising capital and recruiting for your venture?
My biggest lesson learned throughout this journey is that timing is everything. The most effective startups and leaders need to know how to adapt to a changing environment and when a pivot can help access untapped stores of value. As I described earlier, we picked residential solar as the first industry where we wanted to apply our artificial intelligence platform to slash sales and marketing costs. When we started in 2015, the residential solar market was growing at 50% annually with multiple billion-dollar companies rushing to solve the customer acquisition cost problem. Since then the market has gone through a correction, with profitable super-regional installers taking market share away from the big national companies that were built on a model of growth-at-all-costs. Many of the big companies were unable to adapt to shifting market expectations, and a string of bankruptcies hurt investor confidence and the industry’s broader reputation.
Despite the unfavorable timing of the market shift, we were able to maintain our growth trajectory by developing products that catered to the ascendant super-regional installers. We were also able to navigate the difficult fundraising environment at the time thanks to two U.S. Department of Energy SunShot grants and a handful of private investors, who helped us to raise capital when many others were shut out. Timing has had an outsized impact on PowerScout, and I am proud of the way our team has adapted to the changing market.
What are some of the technology advances and changes you envision in the next 5,10 and 20 years in your space?
The most dramatic change is going to be the exponential growth in data collection at the household and consumer level. Sensors are being deployed at a rapid pace within the built environment, measuring everything from building occupant activity patterns to sub-minute electrical usage. The Internet-of-Things is moving into the home, and every device that a consumer touches will be connected and producing data within the next 10 years.
The second will be driven by the availability of technology to process that data, not in gigantic mainframe computers but at the “edge”. Without a need for central data processing, the cycle time of hypothesis, test, learn can be dramatically shortened. Self-driving vehicles gather data as they drive, process it in conjunction with the data from millions of other vehicles, and adapt their behavior in real time. The new iPhone X has a processor dedicated to machine learning, processing and training. That capability could let it can learn to send you an early reminder if you’re chronically late to meetings, automatically reply to incoming text messages while you’re driving, or suggest restaurants in a new city based on your hometown favorites. Device level-analytics and processing will become standard fare within the next five years.
Third, there is a war for who will provide the operating system for the smart home. Google, Amazon, Apple and others are pouring their considerable resources into becoming the hub for the ever-growing ecosystem of devices that will make up the Smart Home of the future. The winner will enjoy a powerful positive feedback loop, as their greater access to data expands their ability to hypothesize, test, and learn such that they become an indispensable part of daily life.
I predict that within the next few decades these converging trends will fundamentally alter the way we interact with the built environment around us. For example, imagine a future where Amazon could become the largest energy provider without owning a single power plant. Homeowners could buy a subscription for say 10,000 EV-miles and 10,000 kWh from a trusted brand like Amazon. Amazon would provide electrons through a distributed network of solar systems owned and operated by a third-party provider, and transportation services through a fleet of self-driving cars. The key will be the data-driven understanding of consumer’s needs and how seamlessly services can be tailored to meet those needs.
What are the characteristics you look for in talent that you recruit into your company?
The most successful PowerScout employees occupy the intersection of humility, intelligence, and curiosity. These are the people that find new problems to solve, attack them with analytical rigor, and never fear the possibility that they might be wrong. The environment produced by these types of employees is one where the team is obsessed with finding the best solution and values the participation of each individual member. No job is too big or too small, and the right answer is always to roll up your sleeves and approach the problem with an open mind. If the above sounds like you, give us a call!
What are your most rewarding moments as a founder?
External to PowerScout, I love seeing our customers reach their business objectives and attribute their success to our data. Helping a client to slash their customer acquisition costs, thereby earning the right to grow our business with theirs, is tremendously rewarding.
Internally, the most rewarding moments have always been watching the achievements of our team. Going from a whiteboard to a revenue generating artificial intelligence platform with a bicoastal client base has been an incredible journey. I am very fortunate to have walked this path with some of the most intellectually talented, curious, and humble individuals that I have ever worked with, and consider it one of the most rewarding experiences of my career.
What causes you the most stress and how do you manage these episodes?
Entrepreneurs live their lives one step ahead of failure. As one of our investors sagely pointed out, a founder can never be sure he has made it until the moment he actually makes it. Until that moment, all of your energy is focused on earning the right to work towards your vision for one more day. If you can build your momentum, be deliberate in every choice, and spend each dollar as if it were your last, then you have a chance of stringing together enough days such that you reach the moment when you’ve finally made it.
The most stressful part of this process for me is that many people prefer security in their personal and professional lives, and thus don’t deal well with the uncertainty and quick pivots inherent in the startup journey. The concept of joining a startup is easily romanticized, and the reality can make for a rude awakening. This startup thing is really hard!
I had to come to terms with the fact that a startup career is not for everyone and that different stages of a startup’s evolution require people with different skillsets. While churn is disruptive to the organization and it results in a never-ending recruiting cycle, I learned that it is inevitable. No founder enjoys making difficult personnel decisions, but at the end of the day if an employee isn’t an organizational fit I have learned that the best way to manage it by being transparent and honest with yourself and the employee.
What big things are in store next for you and the company?
I am continually in awe of the amount of data our team has assembled over the last two and a half years. It has allowed us to establish a beachhead in the Smart Home core competency through our residential solar business, and we have big plans for new states, new partnerships and new Smart Home verticals in 2018.
We have been exploring other areas where can leverage our unique data sets and artificial intelligence algorithms. We’re really excited about disrupting industries that touch the built environment and residential building stock but haven’t yet participated in the artificial intelligence revolution. We think that the property insurance and consumer mortgage markets are really attractive use cases for our technology.