How Silicon Valley uses data science to help legal weed grow

Forget hippies and D.A.R.E. campaigns, forget Cheech and Chong and Nancy Reagan. For over a century, assumptions about cannabis have informed everything from incarceration rates to medical classification. Not any longer. 

As laws and regulations relax around the world’s most profitable plant, and as waves of “ganjapreneurs” move from the shadows into the grey market, Silicon Valley is taking tentative steps into an industry that didn’t exist twenty years ago (legally, anyway).

New venture-backed tech companies are using data science to tackle the enormous amounts of newly created data and the relentless, contradictory, and uncharted roadblocks facing the “green rush” of legal marijuana startups. It’s an industry that needs to be built and shaped, and a new crop of founders are ready to do it.

As the commercial cannabis industry has grown, so too has the quantity of data that companies, consumers, and investors want to track. A trickle of information has become a full-on firehose of data, with everything from brand preferences to plant growth becoming trackable and actionable, right down to the precise ratios of molecules in the cannabis plant itself.

What’s being created is a brand new vertical: marijuana.

Marijuana startups are using data science to create products that help boost consumer confidence, streamline and simplify supply chains, and fight back stigmatization. These startups are building the data foundation that the industry will use now and in the future. We are at the dawn of quantified cannabis.

It feels like exciting times in tech again. Legal marijuana is so new that there’s no disruption to be had. Everyone is a startup, and they’re roughly divided into two factions: companies that grow and sell legal cannabis, and “marijuana tech” – those pick-and-shovel (or dare I say “trowel-and-shovel”?) startups that are using data to bring legal marijuana products into a 21st-century market.

“There’s this really rare combination of having this old school industry, this physical agricultural product that is consumed, being reborn in an era of technology,” says Steve Albarran, co-founder and CEO of Confident Cannabis, a Y Combinator- and StartX-backed legal cannabis tech startup whose overarching mission is to bring greater transparency to legal marijuana technology. “That’s never really happened before, so all the rules are being completely rewritten.”

The opportunities for “quantified cannabis” are numerous. With every new state and country that permits the industry to flourish, the data accumulates. California-based Eaze is one such company that predicted on-demand delivery of medical cannabis being a key opportunity for growth.

Often billed as “the Uber of marijuana”, Eaze facilitates the delivery of medical marijuana to patients from dispensaries in 20 minutes or less. In tandem with its EazeMD app, which helps streamline the process of obtaining medical marijuana recommendations from physicians, Eaze leverages its database of medical marijuana patients and anonymized customer data to change the mandates of lawmakers at all levels of government.

“We have access to aggregated medical marijuana usage data that has been unaccessible to regulators for years,” says Scott Dunlap, CMO of Eaze. “We’re uniquely able to work with policy makers in local and state government to help them understand the industry better and thus create regulation which enables patients to access their medicine safely and conveniently.”

Continuing shifts in public opinion bode well for marijuana; the once haphazard “wild west” of the legal marijuana industry is now a “green rush” of goods and data. But what is being quantified? How do growers and producers, the overwhelming majority of companies in the space, cut through the unparalleled noise?

That’s exactly how Silicon Valley enters the picture, where legal marijuana tech companies are moving legal marijuana producers from the black market and trial-and-error into an era of digital convenience and predictive modeling. For a “cannatech” startup, data has been its saving grace to ensure that margin for error is even slimmer than usual. It also helps these newly legal companies grow, push more product, and become far more mainstream than ever before.

Valley insiders are predicting that “vice tech” startups will yield the next crop of unicorns, representing a market with a combined (predicted) value of over $341.8 billion.

Marijuana tech startups are one subset of a subset of this market, and they are increasingly grabbing the attention of mainstream venture capitalists and erstwhile tech founders. Marijuana tech strikes the perfect balance of risk vs. reward among those who see the massive potential of using data science to capitalize on what is, essentially, a brand new industry.

But the stakes are high, as flaunting federal law can have serious business implications. When it comes to state-legal marijuana, there’s precious little room for error, but in turn there’s almost unlimited opportunity.

According to the ArcView group, the commercial cannabis market is predicted to swell from $5.7 billion in 2015 to $22 billion by 2020, which would be nearly double the revenue of the NFL. The first state to legalize the recreational use of marijuana, Colorado, has already reached $270 million in sales of marijuana products in Q1 2016 alone. Given how public opinion is shifting, legal marijuana has all the appearance of a sure thing, and commercial cannabis micro-VCs are giving way to funding from Y Combinator and Peter Thiel’s Privateer Holdings.

“One of our angel investors came from the tech investment world,” says Cy Scott, CEO and co-founder of legal cannabis SaaS company Headset, a startup co-founded by the team who also co-founded Leafly, the world’s largest cannabis website.

“I asked him, ‘What are the differences that you see when you’re investing in cannabis versus when you’re investing in technology?’ He said with technology, you’re betting on something that you think is really good. You hope that the mass market will adopt it and see the value in it at some point in the future. He said with cannabis investments you know the demand is there. It’s a product that already sells.”

Even faced with hazy legalities, the possibilities appear endless when one realizes they don’t have to grow and sell legal marijuana to be a marijuana startup. So, why haven’t more investors started to dip their toes in the water?  

It’s simple: marijuana regulations affect everything.

Imagine building software for an industry where the rules can change from one month to the next, where you can’t write off your business expenses, where you can’t use AdWords to promote, and your customers’ customers can’t use credit cards. That’s what “cannabusiness” is up against.

Founders have to get things right the first time in very, very fundamental ways. Shipping product isn’t so clear cut when the usual “shoot first, aim later” ethos could land you, or your customers, in jail. How, then, does a legal marijuana startup position itself when the stroke of a lawmaker’s pen could wipe out your entire business model?

“It’s difficult to really scale or be profitable in this industry,” says Christian Nitu, co-founder of Weave, a software company based in Boulder, Colorado that offers supply chain management and product analytics for dispensaries and POS menus for consumers.

With their mission of “simplifying cannabis through data science,” Weave is focused on helping shed light on these uncharted waters, creating an app ecosystem that streamlines solutions for inventory management, customer feedback, and regulatory compliance.

“Startups don’t know where things are going to be in a couple of months or a couple of years,” Christian continues. “Data integrity, from the very first time we entered the industry in 2014 to today, is still a huge issue.

“We thought that if we could tackle the burden of data integrity, we could make these businesses a lot more efficient, and in turn maybe we could help bring down the price of marijuana for consumers and patients, and we could actually help industry progress forward.”

The drive to move things forward for the legal marijuana industry via tech is a common thread among the founders I talked to for this feature. It’s because everyone in the legal marijuana product ecosystem must rely on one another to ensure laws and opinions change for the better. And their best lever is data.

If one were to mention data science and the legal marijuana industry in the same sentence five years ago, there wouldn’t be much to talk about.

The norm for “analytics” in weed at that time was (and sometimes continues to be) wild guesses, numbers crammed into spreadsheets, and a lot of crossed fingers. Data was barely being collected, much less being put to good use.

Founder and CEO Keith McCarty made data science a priority at the outset of Eaze. He sees data as critical to every decision that they make. Eaze considers itself a logistics and technology company first and foremost. Fully one-fifth of their growing team are data scientists, led by Shri Ganeshram, their VP of Growth. It’s a staggering challenge from both a compliance and logistics perspective, but by keeping real-time data analysis front and center, Eaze is able to tackle it with agility.

“Data is a first class citizen internally,” says Keith. “We see the investment in our data team as critical to us sustaining our position as the best cannabis technology company in the world. Whether it be user interface changes, enhancements to the logistics engine, or enhancements to the menu to fulfill patient needs, every decision starts and ends with data.”

“In this industry, there is no benchmark for what we are doing because it is so novel,” says Shri. “We rarely make a decision without first intensely focusing on the data.” Just like all good entrepreneurs.

But there’s a huge piece of the legal cannabis industry that needs the most help from data science: marijuana itself.

As the proliferation of legal marijuana-related data explodes, there is similarly massive demand for data that can transform marijuana from a “fringe” product to a legitimate consumer product.

For example, there’s no consensus on what “quality” means when it comes to legal marijuana. Neither the USDA nor the FDA regulate legal marijuana or marijuana products, and there’s no RDA for THC.

There’s no industry standard quality control, and no assurance that you’re getting what you think you’re getting.

Before legalization in Colorado, for example, only three strains of marijuana (out of hundreds) had their genetics sequenced; that number has gone up 300x and will only continue to increase, unleashing mountains of data on an industry made up of mostly small businesses that can barely keep up.

In response to the heaps of data, one startup has risen to the challenge to make this data more transparent.   

“When we were thinking about starting a company in the industry, we intentionally didn’t come up with an idea,” says Steve Albarran at Confident Cannabis. The Palo Alto-based SaaS startup is working to make it easier for dispensaries, producers, and laboratories to measure, track, and share product information and ingredients.

“We didn’t say, ‘We’re going to go build this and see if someone wants it.’ Instead what we did was just very quietly ask questions about problems. We found out that data transparency was a big problem we felt we could solve. Nobody knows what’s being made, who’s making it and what it’s made of. The quality expectations are totally opaque and nobody can communicate it in a common language.”

Their goal is to help consumers have a more consistent experience, give regulators the information they need to monitor commercial cannabis quality and lab data, and give customers the information they need to make informed purchasing decisions.

Already, Confident Cannabis has received strain data from thousands of legal marijuana producers in California, Washington, Colorado, Nevada and Oregon, helping map and compare strains for quality control. They now have the largest database of legal cannabis in the world.

Using analytics to improve legal marijuana products or create new niches simply wasn’t a possibility even five years ago. A company like ClubM, a legal marijuana subscription startup that delivers a monthly selection of premium products by courier in California, grew out of category analysis that revealed that there were segments of legal marijuana users who wanted to sample new products, but didn’t want to visit brick-and-mortar dispensaries to do so.

“We discovered that there are people out there seeking exclusivity, privacy, and luxury around their cannabis,” says ClubM partner and co-founder Chris Husong. “These are totally new concepts in the industry. When we looked closer and saw that women prefer delivery because of safety and optics concerns, we were able to tailor our offerings for them more precisely. Analytics have been huge for us.”

Within a grey market like state-but-not-yet-federally-legal marijuana, the task of protecting customer data takes on a whole new dimension.

“We see a lot of people who are reluctant to share data with us just because of safety concerns,” says Cy Scott, CEO of Headset, the Seattle-based business intelligence startup providing both industry trends and data analytics to commercial cannabis retailers while maintaining the privacy of customers.

“That’s been a challenge, but we’re changing that conversation and working through it. Once people see the benefit of giving us the data and getting back the insights that we give them, their hesitation is usually short-lived.”

“Data-wise we definitely don’t want to touch any personally identifiable information,” says Scott Vickers, Headset’s CTO. “But understanding what people are purchasing, and what type of products are getting introduced is fascinating. It’s so sophisticated now.

“Fast forward five years from now I’ll laugh at saying that, because I’m sure it’s going to be even more sophisticated, especially compared to when we first started Leafly, when you’d go to a dispensary and there would be a few strains, and a few jars, and maybe a brownie in a ziplock bag,” he says with a chuckle.

Brian Wansolich, the CDO of Headset, agrees with Scott. “In our opinion, looking at [the complexity of] data for cannabis is more interesting than looking at data for consumer consumption of snack foods.” 

“The novelty of the industry will definitely appeal to the data scientists out there. They’re going to want to sink their teeth into this whole industry, just because it’s uncharted.”

Marijuana tech startups have as many reasons to be optimistic as they do to be ambitious, but few imagined they’d be using their data science skills in support of political activism. Opinions about marijuana are shifting and changing, but there’s still an ingrained societal skepticism about their industry that can and must be replaced with data.

For the first time in modern history, a majority of Americans want some form of marijuana legalization. Pew’s latest poll pegs 53% of Americans in favor of legal cannabis (Gallup is slightly higher at 58%).

Data offers a necessary dimension to the debate over marijuana legalization, and getting people to think about marijuana not as a scary drug, but as a new and legitimate product.

“It’s rare to have the opportunity to step into a [potentially] $20 billion+ industry early enough that you can simultaneously shape the customer experience, the legal environment, societal acceptance, and the industry at a global level,” says Scott Dunlap, CMO of Eaze.

All of the founders we spoke with discovered that, what at first had seemed only like just a golden business opportunity-turned-medical alternative, could impact the standard of living for millions of people.

“I think it’s hard to not have at least some kind of streak inside you that wants to see the movement evolve,” Brian Wansolich of Headset says. “I think everybody who is in the industry now, that you would regard as a player, is an activist in that sense. You take that on automatically I feel. You want to see it through because it means the success of your business but it also… I think it evolves us as people anyway.

“We’re not looking for an Instagram kind of exit, and we’re not building a company to get big really fast and get acquired,” Brian continues. “There are almost zero startups that have been acquired in this industry right now. The way we think about it is much more of an Amazon situation where it’s a long road of steady fails and not a lot of profitability because you’re building the infrastructure of this new, big, beautiful thing called cannabis.”

One of the benefits of being contrarian entrepreneurs entering this industry is that marijuana has always been contrary to the established order. In many ways, it’s a match made in heaven. 

Legal marijuana gives tech founders an opportunity to be on the right side of history. Signs are pointing to cannabis being either the next big medical breakthrough or, at the very least, the next great American industry.

At first blush, this thinking feels like the sort of hyperbole that’s all too common in the Valley, where so many claim they have unlimited potential to change the world, using tech as a force for what they see as “good.”

“The biggest gift data has to give to the cannabis industry will be educating people about cannabis, how and why people use it, and its safety,” says Scott. “After all, it’s hard to argue with the data.”

Photographs are courtesy of Jeff Turner, Don Goofy, and torbakhopper, and are made available under an Attribution 2.0 Generic license; product photos are courtesy of Eaze and Headset