The premier snow-forecasting app for skiers and snowboarders isn’t a product of government agencies or major corporations. Instead, it’s an independent startup called OpenSnow, which masterfully blends government data with its own AI models and decades of firsthand alpine experience to deliver weather predictions—including snow and, soon, avalanche forecasts—that surpass all others. Seasoned skiers and riders, from the slopes of Alpine Meadows to Mont Blanc, Crested Butte to Killington, now defer to OpenSnow and its small team of seasoned forecasters before venturing into the mountains. These dedicated meteorologists, all men, have become cult figures, meticulously analyzing vast amounts of data to craft the highly sought-after "Daily Snow" reports that guide powder-seekers worldwide.

Bryan Allegretto, a founding partner and forecaster at OpenSnow, humorously refers to his status as "F-list famous, not even D-list." This year, OpenSnow’s insights have been particularly crucial, navigating a winter characterized by unusual weather patterns. While the Western U.S. experienced minimal snowfall despite a severe storm that triggered a historically deadly avalanche and a rapid melt, the East Coast has enjoyed a prolonged, deep winter. Allegretto, known to many as BA, shared his journey with MIT Technology Review from the Tahoe mountains, discussing weather, AI, avalanches, and how their humble beginnings evolved into a vital tool for snow enthusiasts. He and his Colorado-based counterpart, CEO Joel Gratz, transformed an email list of 37 into a devoted following of half a million.

Allegretto’s childhood in New Jersey, far from a typical snow-heavy region, was marked by an early fascination with severe weather, particularly Nor’easters. He recalls the 1989 blizzard that blanketed the East Coast, drawing parallels to his father’s role in managing snowplow operations. This childhood exposure fueled a passion for being "in it," leading to him shoveling driveways and later embarking on solo storm-chasing road trips during high school, often sleeping in gas stations.

These expeditions were more than just chasing storms; they were a quest for "natural highs" and happiness, a pursuit stemming from a challenging upbringing marked by his parents’ separation and his mother’s struggles with addiction. With no parental guidance toward a traditional career, Allegretto sought fulfillment. Despite financial constraints and academic hurdles, he pursued meteorology at Kean University, even interning at NBC in New York. His ambition was clear: to merge his love for weather with the ski and snowboard industry. A business degree from Rowan University and a stint teaching snowboarding in Hawaii further solidified his unconventional path, ensuring he wouldn’t settle for a career he disliked.

Unlike many of his meteorology peers, Allegretto identified as a "punk rocker, skater, snowboarder," a "radical storm chaser" with a big personality, a departure from the more conventional meteorology students. He recognized that the traditional "weatherman mold" didn’t align with his aspirations. Back then, meteorology careers were largely confined to government offices, insurance companies, or local news, none of which appealed to his desire for a more rugged, mountain-centric lifestyle. He felt that meteorologists, often based in cities for job availability, lacked the deep understanding of mountain weather that he, living in Tahoe since 2006, was developing. He observed a prevailing skepticism towards weather forecasts in ski towns, a "we’ll believe it when we see it" mentality. He noted that local forecasters often provided vague predictions, failing to address crucial details like wind at higher elevations, which skiers desperately needed. This gap became his opportunity to provide hyper-local, actionable forecasts.

Allegretto balanced a day job at a resort with his burgeoning passion for weather forecasting, waking at 4 a.m. to analyze storm patterns and their interaction with mountain terrain. His early forecasts, initially shared via email to colleagues, quickly gained traction, with requests to be added to the distribution list multiplying. Soon, resorts around Tahoe were seeking to utilize his insights. His forecasting method involved meticulously studying various government weather models—NOAA, GFS, Canadian, European, German, and Japanese—and recognizing their limitations in accurately predicting mountain weather. He understood that these models, with their low resolution, generalized conditions and struggled to account for the microclimates of mountainous regions. His advantage lay in his proximity to the mountains, allowing him to correct model errors. In 2007, he launched "Tahoe Weather Discussion," a website dedicated to his localized forecasts.

Around the same time, Allegretto learned of Joel Gratz in Boulder, Colorado, who was running a similar venture called Colorado Powder Forecast, and Evan Thayer in Utah with his "Wasatch Forecast." Their paths converged, with Gratz proposing a collaboration in 2010. Gratz, having heard of Allegretto’s significant website traffic, suggested joining forces rather than competing. When Allegretto inquired about compensation, Gratz proposed a zero-dollar acquisition of Allegretto’s company, requesting only his expertise. While Thayer, who had built his website from scratch, initially resisted this offer, Allegretto, overwhelmed by his demanding day job, new family, and a struggling website, saw an opportunity for partnership and readily agreed.

For its initial 15 years, OpenSnow relied on human expertise. Allegretto described the early days as "using our brains," meticulously analyzing data from various sources—government models, airplanes, satellites, and balloons—to build spreadsheets and correct forecast errors. This manual process involved hand-keying in snowfall reports from resort websites every morning, a task that consumed hours.

The turning point came around 2018 with the development of their proprietary weather model, METEOS. This AI-driven system, built upon their collective knowledge and experience, automated the forecasting process, allowing them to generate predictions for any GPS point globally. METEOS ingested, corrected, and then outputted forecasts, significantly streamlining their operations.

Financially, OpenSnow’s early days were challenging, relying on advertising revenue. The hiring of Eric Strassburger, who significantly boosted ad income, provided some relief. However, a decrease in Google Ad rates prompted a strategic shift. Recognizing the growing trend of paid content, Gratz proposed transitioning to a subscription model, acknowledging that their daily weather reports were essentially journalism. This pivot proved transformative, leading to substantial revenue growth and enabling Allegretto and Gratz to leave their day jobs and dedicate themselves full-time to OpenSnow. The company experienced explosive growth, surprising them with the willingness of users to pay for their service, even as much of the content remained accessible for free.

In late 2021, OpenSnow introduced a paywall for premium features, a move that initially sparked apprehension about customer loss. However, the predicted exodus of 90% of their user base did not materialize. Despite a dip in traffic during the paywall implementation and the onset of COVID-19, OpenSnow’s growth trajectory remained upward, with users expressing their inability to function without the app.

Allegretto’s assertion that OpenSnow is indispensable for skiers is validated by the fact that it’s the go-to app in ski towns. Following the significant avalanche in Tahoe, OpenSnow actively assisted search-and-rescue operations with their forecasts, solidifying their reputation and leading to official partnerships with organizations like Ski California, Ski Utah, the National Ski Patrol, and the Professional Ski Instructors of America. The company boasts exceptional renewal, retention, and open rates, far exceeding industry benchmarks.

The loyalty of OpenSnow’s users and employees is remarkable, fostering a sense of community akin to a "benevolent cult." This connection stems from their "underground roots" and independent status. With a team of fourteen full-time employees, many with advanced meteorology degrees, and a history of long-term commitment from key personnel like COO Sam Collentine, who started as an intern, the company fosters a stable and dedicated workforce. Allegretto himself has been forecasting in Tahoe for two decades, leading to widespread recognition and even instances of unsolicited gifts and personalized memorabilia from fans.

The "snow god" moniker, while humorous, reflects the profound impact OpenSnow has had on its user base. Allegretto shares anecdotes of fans dressing up as him for Halloween and even finding autographed photos of himself displayed prominently in homes. This level of adoration, while sometimes overwhelming, underscores the personal connection users feel with the forecasters.

Despite the "god-like" status, Allegretto admits to receiving "thousands" of "hate mail" and "mean DMs," with individuals believing he can control the weather and blaming him for unfavorable conditions. He has since deleted his personal social media accounts, never seeking the spotlight and originally signing off forecasts as "BA" to avoid using his full name. The company’s success, he attributes to Joel Gratz’s foresight in recognizing the importance of forecasters in building customer loyalty.

The integration of AI has been a significant evolution for OpenSnow. After years of relying on their proprietary METEOS model, they launched PEAKS in December, a machine-learning model designed to perform forecasting tasks with unprecedented speed and accuracy. PEAKS can ingest decades of historical government data, correcting past errors and providing forecasts for any location globally within minutes.

PEAKS’ accuracy stems from its ability to downscale weather data from broad "blobby" regions to precise, localized predictions. Unlike older models that averaged data over large grid boxes, PEAKS can now differentiate snowfall in a ski resort’s parking lot from its summit. By training the AI on wind patterns, thermal gradients, terrain, and weather phenomena, PEAKS can transform coarse-resolution data into highly specific three-kilometer resolution forecasts, effectively teaching the AI to forecast with human-like intuition, but with a 50% improvement in accuracy. This allows Allegretto to spend more time on crafting engaging written forecasts, emphasizing the personal connection that remains central to OpenSnow’s identity.

Concerns about federal funding cuts to the NWS and NOAA were discussed within the company, but the swift backlash from the weather community and the continued availability of government data have mitigated these worries. Allegretto emphasizes that without this public data, OpenSnow "won’t survive."

The future of OpenSnow includes expansion into severe weather forecasting with the acquisition of StormNet, enabling predictions of lightning, hail, and tornadoes up to seven days in advance. Moreover, they are developing an AI-powered avalanche forecasting feature for next winter. While not intended to replace existing avalanche centers, this tool will provide earlier, location-specific warnings by analyzing slope angles, historical weather data, and current conditions, potentially saving lives.

Allegretto recounts a chilling anecdote of a hiker who heeded OpenSnow’s cautious language regarding an approaching storm, opting to leave the Frog Lake huts before the fatal avalanche. This highlights OpenSnow’s commitment to accurate, unvarnished reporting, devoid of hype or sugarcoating, with their sole incentive being precision.

The recent record-breaking storm in Tahoe, while celebrated for its immense snowfall, also brought tragedy. Allegretto notes the similarity in snowfall amounts to the 1982 storm, which also resulted in significant loss of life, underscoring the destructive power of too much snow too quickly. The rapid disappearance of this snow in just five days further highlights the unpredictable nature of current weather patterns.

Looking ahead, Allegretto acknowledges the challenges posed by climate change and the increasing variability in weather patterns, with wild swings between extreme cold and heat, or drought and heavy precipitation. While a poor snow year in one region can impact the industry, OpenSnow’s diversified markets across the West, East Coast, Japan, and British Columbia have historically provided a buffer. This year, however, marked the worst for the Western U.S. in recent memory. Despite this, OpenSnow’s traffic and business metrics have continued to rise, amplified by their first marketing campaign in 15 years.

Allegretto and Gratz often reflect on their unexpected success, acknowledging that building this empire was never the initial goal. Having never borrowed from investors, they’ve maintained their independence and declined lucrative acquisition offers, driven by their continued enjoyment of the process. Their journey, fueled by a shared passion for skiing and powder forecasting, has unexpectedly led to both wealth and recognition, a testament to their focus on joy and accuracy rather than financial gain or fame. They remain committed to creating something that brings them, and their devoted user base, happiness.