Hello! What's your background, and what are you working on?
I'm Andy Qin, software engineer and clean energy advocate. Photon is an iOS application that helps solar installers measure shade loss on solar panels. Solar installers (the folks who climb on rooftops) want to know how the shade from surrounding trees and buildings will affect the power output of their solar panel arrays. They also use this data to apply for government rebates. So instead of using legacy measurement devices (e.g. Solmetric SunEye) that cost thousands of dollars apiece, why not use the supercomputer in your pocket?
My co-founder Canzhi and I launched Photon in July 2016. Since then, 11 solar companies have become paying customers. We've also recently partnered with Soligent, the country's largest distributor of solar power equipment. Altogether, we're doing around $1,000/month in recurring revenue.
How did you come up with the idea for Photon?
I was volunteering to install solar panels through a nonprofit called SunWork, when my project manager pulled out a Solmetric SunEye. I learned that solar installers everywhere rely on them, even though they're wildly expensive and their firmware crashes often. Hardware-wise, they consist of a camera attached to a computer, not unlike my iPhone. The only differences were that my iPhone doesn't have the software, and it doesn't have a fisheye lens camera. The latter was no problem, since fisheye lenses for selfies are cheap. The former I took on as a side project.
At the time, I was supposed to be finding a full-time job. I had about 5 months left before finishing my undergraduate education at UC Berkeley. But you know that feeling where you're so obsessed with your project that it's all you daydream about? That was me. The only goal was to ship. I saw the project as a small way to help ease our transition to clean energy generation. It took priority over things like homework and interviews.
What was involved in building and launching the initial product?
Photon was accepted into the third cohort at Powerhouse, a renewable energy-focused accelerator based in Oakland. This was huge for us. We had no connections to speak of in the solar industry. They equipped us with a network, $10k in pre-seed funding, and office space for 6 months to get started. Perhaps most importantly, Powerhouse gave us the excuse we needed to work on Photon full-time. There were no distractions because college was over for me and postponed for my younger co-founder, Canzhi. We shipped V1 in just a month after starting at the accelerator program.
Deciding what features went into V1 was tricky because neither of us were solar installers — our intended user. Actually, neither of us had ever been employed for any job that wasn't white collar. So to compensate, we asked solar installers for what they wanted in user surveys on r/solar.
To further validate our market, we'd make attempts to reverse engineer the Solmetric SunEye, then present our prototypes to local solar companies on the weekends. Coincidentally, a design engineer at one of the first companies we spoke with told us that he'd been tasked with finding a modern alternative to the Solmetric SunEye earlier that month. They were thrilled to be beta testers.
What strategies have you used to grow Photon and find customers?
The single most effective thing we did to acquire customers was attend solar industry trade shows. One of the biggest is Intersolar, which was held at the Moscone Center in San Francisco last year. At Intersolar, you had established giants like Lockheed Martin and Huawei presenting in huge, flashy booths, operated by dozens of employees in clean-cut uniform serving complimentary refreshments. By contrast, we had a witty poster and an iPad running Photon. It felt very David and Goliath:
One of countless demos from that day.
But it turns out that's all you need to get some interest. Our first customers came from those small crowds we drew at Intersolar. We also put together a short tutorial video, which is table stakes for any software used by solar installers:
Since our product is essentially a measurement tool, the most frequently asked question was, "How accurate is your shade analysis?" We needed certification from a respected third party to show that our app wasn't a gimmick. NREL, the National Renewable Energy Lab, typically does the certifying for commercial products like ours. The problem was that getting NREL approval could take months, and we were trying to move fast.
The solution here was to do it ourselves. We took it upon ourselves to rigorously and honestly test our product on 20 different rooftops against the Solmetric SunEye. The result was a white paper that we published ourselves, and distributed it as part of our marketing materials. As software engineers, we're considerably more comfortable typing at a desk than climbing a 2-story ladder perched against a steep roof. But when duty calls, we answer:
"How did we get up here?"
On a side note, we eventually were accepted into NREL's Commercialization Assistance Program, so their official assessment is pending.
What've you done to grow your revenue to $1000/month?
Our revenue comes from two sources: our regular subscribers and our distribution partnership. To speak to the former, we run our subscriptions on Stripe. They make it super easy to configure three subscription plans for us: a 1-month plan at $34.99, a monthly recurring plan at $29.99/month, and a yearly recurring plan at $299.99/year.
Our partnership with Soligent entails that we build them a white-labeled version of Photon with their logo and brand colors, due for launch in April 2017. In return, they're paying us a flat annual fee to distribute this version of the app to the thousands of solar companies they regularly serve. This way, we don't have to do sales, and Soligent has a way to reward their loyal customers. It works out nicely.
Our business model is tricky because we can't do a true freemium model. The app can be downloaded right away, but the fisheye lens takes Amazon a couple days to ship. So the best we can do is let users download and get a feel for Photon, but we're unable to give them the full experience until they choose a subscription. Because each subscription plan brings in more revenue than the cost of the lens, we ship the lens for free with the purchase of any subscription plan. That keeps our unit economics in the green.
What are your goals for the future?
I want to keep Photon reliable and useful to our customers. The biggest thing on our roadmap is how our traffic is expected to rise dramatically when we launch our joint iOS application with Soligent in April. So if there are say, a few thousand solar installers simultaneously flooding the server with requests, we want to make sure that server responds to all of them in a timely manner.
What are the biggest challenges you've faced?
We learned the hard way that while Photon is desired by small and medium-sized solar companies, the large solar companies are less interested.
SolarCity, SunPower, and others like them are investing in fully remote shade analysis technologies. Whereas tools like Photon require their users to be physically onsite, the future is being able to closely approximate the same measurements from aerial imagery captured by drones or low-flying planes. This shift represents a many-orders-of-magnitude decrease in costs for big players, with their big overhead costs. Our market was then limited to the long tail of solar companies, which in aggregate is attractive, but in practice is difficult to sell to.
Funny enough, I was reading The Innovator's Dilemma by Clayton Christensen when we were first evaluating the threat of fully remote shade analysis tools to our business. Everything had an uncanny resemblance to Christensen's described pattern of how new tech (remote shade analysis) that initially costs less but perform poorly sometimes improve its performance until it eventually displaces the old tech (onsite shade analysis).
Get a Beautiful Mockup. No Photoshop required!
Choose from 5000+ templates.
What were your biggest advantages? Was anything particularly helpful?
The solar industry is in dire need of software engineering talent. Because we were a nimble engineering team of two, we could respond to feature requests like lightning. I was on the phone with a user who requested a feature and asked if it could be done in the next 2 to 3 months. We shipped the feature within a week. I think that speaks to the quality of software support that the industry is used to. It significantly differentiated us from our competition.
I've always been a believer of Y Combinator's mantra for founders: either be talking to users or building product. Their class at Stanford years ago remains a timeless resource of good principles to follow. I found Emmet Shear's talk about interviewing users to be particularly useful for us.
What's your advice for indie hackers who are just starting out?
The first version can be more minimal than you think. Only in retrospect did we come up with ways that we could have shipped without deploying a single line of code. For example, we could have had users take pictures with the lenses, then email them to us for analysis. No app needed!
If you're still brainstorming, look for network effects. Whether you're the first user of Photon, or the millionth, the experience is more or less the same. I wish this wasn't so. If you're in the process of finding a project, ask yourself how the nth user signing up will improve things for the existing n - 1.
Stop trying to start unicorns. Not every internet business needs to be a world-changing, venture capital-backed, headline-making company. I get the appeal of having your name in the news after a successful fundraising round. But the biggest critique I have of the Silicon Valley startup culture is for founders to question harder whether they really need institutional money. Could your business be really compelling and lean? I'd conjecture that many technology businesses here would've been better off not having raised the millions of dollars. Now those founders bear the weight of investor expectations on their shoulders, when they perhaps didn't need to.