How To Scale A Weather Forecasting App To 500K Users


Powder days are like heaven to skiers. The hill becomes like butter, and there’s a magical feeling in knowing your skis are the first to grace the snow.

The toughest part of skiing power, though, is finding powder. Joel Gratz, a meteorologist in the Boulder, CO area, knew this first hand. No one in the Colorado area was paying particular attention to powder - or the backcountry of the state for that matter - so each week Joel would send out a forecast to people in his network about upcoming powder days.

The newsletter, which started off as 100 people, soon grew through word of mouth to 400. It was at that point that Joel quit his job as a meteorologist and OpenSnow was born.

I had a chance to chat with Joel about bootstrapping the business and how finding a niche within a niche helped him succeed.

This interview has been edited for clarity.

Early Start Of The Business

Joel: Most meteorologists know that they want to be a meteorologist very early in their life due to some type of weather event - snow, hurricane, tornado - whatever it might be. I am no different. At four years old, I became interested in meteorology when I first went skiing in Pennsylvania. I knew basically since elementary school that I wanted to be a meteorologist at Penn State.

Fast forward to grad school - I’m at the University of Colorado studying environmental studies - and I realized that powder was my calling. I loved trying to forecast and pinpoint powder days. At the time, there weren't really many people laser-focused on powder days, so it took me some time to test my predictions.

I was wrong often, but eventually honed in my skills and decided to try to make a business out of it. The analog business that had been around for a couple of decades was a company called Surfline. They do what we do for skiing, but they forecast waves for surfers. They'd been around well before the internet so I just said heck, if they can produce a pretty good website and app, make money on advertising and subscriptions build a business off for surfers - I could do something similar for powder for skiers.

Building The MVP

Joel: I hired contractors, but I'll go one step back before that, which is setting up a rudimentary website. I am not a good developer but I knew enough to start a WordPress blog. It cost around 50 bucks for hosting per year. So, I did that while I still had a job writing forecasts for a couple of weeks and it was just getting out to people organically.

It was only after doing that for a couple of years that people had started to email me to place ads in the blog, which is when I decided to quit my job and try this out. I had saved some money from working my job and at that point, in my late twenties, I did not have a family. I didn't have a significant other or kids. I had a roommate paying half my mortgage. Financially, it was the best time for me to go out and start the business.


Meeting His Co-Founder

Joel: I met a meteorologist in Boulder who was a developer while looking for a technical co-founder. We went to the same community meetup one day and I was shocked that there was another meteorologist in the room. He helped to build the first iteration of the site.

Early Mistakes

Around this time, I wasn't making any money. I didn't know how to charge people money for advertising. OpenSnow started out with me sending out a powder forecast on a Wednesday night Colorado powder forecast to 37 people on Gmail. After a couple of years it was 400 people, but Gmail only let you send a to a hundred recipients, so I would copy and paste my email in four different Gmail windows and then insert a hundred contacts into each one because I couldn’t send it to more than a hundred people at a time.

This was just purely fun. I didn't really think it was going to be business until maybe two or three years in when the media started calling me for interviews. Then I made it a blog and it took off from there, but I didn't make any money those first couple of years.


Joel: People started emailing me to pay for advertisements in my emails and on the blog. That's when I first started to make money. When I did make money, I hired a contractor. When he first started I asked him to quit his job. It was something like $35,000 that first year, but he loved to travel and with a nine to five job he couldn't really travel.

In this job, though, he never had to be in Boulder - he could be anywhere you wanted. That's how it started, I went to every meetup in town. There's tons of meetups, but I still struggled for months trying to find a person that I could trust. I didn't want to just hire somebody to code. I needed a technical co-founder, and so it took a long time to find that person.

Freelancing While Building OpenSnow

Since it was all bootstrapped, I consulted for years in the early part of the business for ski publications that wanted me to write articles about different weather forecasts. So I did that for years, just an additional way to make revenue while building up my brand.

It was in 2013 when I decided I wanted more control over the business. Advertising is tricky, you might one client one year and then lose then the next to no fault of your own. So we started a subscription product that was the next iteration of the financial side of the company. Just this past year, actually, that subscription product is now the majority of our revenue and growing quite fast.


Building Multiple Streams Of Revenue

Joel: We had two ways of making money: subscriptions and advertising. These are still our two revenue streams. The advertising supports the free version of the product while subscription gives you a premium experience. This year we're still doing advertising and subscription, but the advertising will be fully in an email newsletter that we have. So we’ve basically removed almost all advertising from the site.

It's cleaner from a product decision standpoint. If you focus too much on generating subscriptions, you lose the readership or at least the page views on the ads. If you go really ad heavy, and you make a few more bucks, but the revenue per user is much greater for subscription than it is for advertising. So there are trade offs.

We are now focused on growing our subscriber base, but advertising still plays a role.

Early Marketing Strategies

Joel: We did almost nothing because we are in a passion driven industry and frankly, we took it for granted. Skiing in some cases is a niche - it's not football or soccer. Then beyond that, we're getting niche of niche, right? Its people that are looking for powder skiing. So we're in this pretty small niche as it is, and people were talking about us. We just grew largely by word of mouth and it kept our costs down. As a small team, we also just didn't have the bandwidth to focus on marketing projects.

We almost never bought ad words - this was even before Apple had search ads in the store. People are super passionate about what we're doing and so word travels.

Number Of Annual Users

Joel: As a private company, we don't give out everything, but we've got about 3 million people that come to the site or app annually. We have close to 500,000 registered users - a small percentage of those pay us as well. In my opinion, we can do a better job of communicating the benefits of paying that subscription, because we have a lot of people that use us, but don't pay.

The world was vastly different when we started this in 2011. Paying for things was not the norm. Now, people are far more used to paying for something that they find valuable. For us as a business, just trying to get over that hump, we had this worldview that was rooted in ideas from 10 years ago. I'm not saying everybody has to pay for everything. There's a reason some things are free and that's fine, from a business model standpoint, but the world is just different, and so we're needing to evolve to match that.

I didn't necessarily start the whole business based on Joel Gratz. There are plenty of people making plenty of money doing that - it's a one trick pony right and it's hard to hand that over to somebody else. Whereas building out the whole platform might take a little bit longer, but it's a better service to customers. It gives the business legs. God forbid I get hit by a bus one day, OpenSnow can continue.


Building The Tech Behind OpenSnow

Joel: All of the computer models that you hear about are generally all available to everybody. The difference between any weather service is how you take that data and craft it into a forecast and how you display it and how you alert on it. We are creating our own model from multiple models and we're making adjustments based on our experience. We're also making data available in pretty remote places that you generally just can't go to a standard weather app and access.

You can easily get a forecast for the town of Stowe, but it's a lot harder to get forecasts for the locations where you can chase powder. We've tried to focus our forecasts on these remote locations that are not easily accessible. The other thing that we're doing is presenting the data in a way that makes it easy to compare against multiple locations to figure out what area has the best powder.

Most weather apps are not designed in any way to allow you to wake up and look at six locations and figure out where you're going to go in 20 minutes. It was a way for us to scratch our own itch, but there are hundreds of thousands to millions of people around the world that want to scratch that itch powder.

Biggest Challenges

Joel: You need a good strategy and idea. But at some point, it becomes finding great people and figuring out how to all work together. Also, as a bootstrap person, you don't have a lot of ability to recover from mistakes.

As a founder, your business is a massive part of your life. I've spent a lot of time trying to figure out what I want out of this business. We've had acquisition offers but I’ve always said no, because we have a lot of fun trying to achieve our vision. There are plenty of nights I'm up until two in the morning working, because I'm super excited to do it, but that's my choice. We like being in control of what we're doing.

Launching A Mobile App

Joel: App and web usage was roughly 50-50 because we have a lot of people that check us at their work computers when they are going to work. When COVID hit, the majority of our traffic was now through the app. But again, the marketing I shared earlier is the same. We've toyed with Apple search ads for installs and it works fine, but we haven’t poured a bunch of money into it.

We were also working with a couple athlete ambassadors, so we pay them and these are people we respect. We didn't just reach out to anybody. That gets a little bit tough to measure - you can measure click backs from social media posts, but not everybody clicks from a social media post to download your app, but we have seen some good success with that. Ultimately, when people find something that they love, they'll often tell their friends, which is responsible for a lot of our growth.


Joel: From a business perspective, I could have been more aggressive, invested more. But it's hard when you start out with that scarcity mindset to think about business from a growth mindset. I've heard this from a number of people that they wished that they had put together a board of advisors or board of directors earlier in their business lifetime. That's just something I am starting to work on now. Even people with massive successes have all doubted themselves. But the faster you can do things and learn from them, the better position you will find yourself in.

  1. 2

    Happy to chat and answer any questions about bootstrapping and finding powder:-) -joel

  2. 1

    The Regret and Biggest Challenges sections - great words. Feel this all the time.

  3. 1

    So we started a subscription product that was the next iteration of the financial side of the company. Just this past year, actually, that subscription product is now the majority of our revenue and growing quite fast.

    Huge. Congrats! Did you initially try to convert free users over to the subscription product, or did you have to target totally new customers for it?

    1. 2

      Good question. We had a traditional freemium model where our most engaged users were most likely to become paid subscribers. The quicker growth of the subscription product during the last year was somewhat due to focus (marketing it more and improving the product) as well as converting some previously free features to be limited to a certain number of views.

Trending on Indie Hackers
My SEO experience 28 comments How long did it take to build your MVP? 12 comments Free e-mail countdown timer that doesn't suck 💩 5 comments My first year of making money on the internet 🤑 This is how it went... 4 comments Dealing with lack of motivation due to failed soft launch 4 comments Is open-source really free to use? 1 comment