Observations of The Cannabis POS Landscape, Including Payments Nonsense

We recently started getting into cannabis (or more commonly termed MRB for marijuana-related business) as it turns out that data + data science + card network data works pretty much everywhere.

The obvious caveat to those familiar with the space, however, is that US MRBs don’t take credit card payments on the major network rails.

At least not legally.

As we dove further into the industry we thought it would be helpful to articulate our learnings and observations in the event that it helped others navigate the landscape better.

The US cannabis industry is small but growing quickly. There are about 7,000 US retail dispensaries and industry CAGR is 28%. Total sales for 2022 are estimated at $33B. 

Generally speaking, cannabis falls into two camps: medical and recreational. There’s not a ton of difference between the two except that the former requires a prescription/medical card while the latter does not. Product is basically the same, although permitting/licensure of the type of retail establishment does affect the speed and ability to open a location. 

To that point, because there are so many changing laws you’re going to want a lawyer if you’re wading into cannabis.

There’s a consistent supply problem in the frontier states, meaning geographies where cannabis is relatively new: dispensaries can’t get enough product to satiate customer demand. This occurs because retailers line up their operating licenses, much like the land boomer days of the 1800’s, only to find that plants and operating wrinkles aren’t yet worked out of the system. For the uninitiated, this makes it appear that demand is through the roof, fueling more build outs and investment.

But when licenses convert to stable operations and the supply chain frees up, like is seen in the mature MRB markets of California and Michigan, demand becomes a clear focus for retailers. 

If you step back and look at the market you see that, outside of these peculiarities, it’s pretty much the same as any other retail vertical.

There are a few multi-location retailers – termed MSO for multi-state operators – but none of them has more than a couple hundred locations. All in, the MSOs account for about 1,000 locations, or 15% of the industry, making cannabis more similar to retail or restaurant than grocery, where a handful of players command 50% of the market.

Cannabis point of sale is just as fractured, too.

In our research the POS players break into these categories based upon raised capital.

Large

• Dutchie (which acquired LeafLogix and Greenbits, with LeafLogix as their main platform that also does e-commerce and is forcing all Leaf Logix customers to Dutchie November 1, 2022)
• Flowhub
• Treez

Medium

• Biotrack
• Blaze
• Cova
• MJ Freeway

Small

• GrowFlow
• Indica Online
• Meadow

E-commerce only

• Buddi (Canada only)
• Jane (but might be building a POS)
• Weedmaps (tried POS but exited the space)

There’s clearly a lot of competition, and of note is how much venture capital has been poured into the POS ecosystem despite a TAM of roughly 7,000 locations… But it’s not exactly as straightforward as it sounds.

Many institutional investors have “sin clauses” from their LPs stating that they won’t invest in industries that their LPs don’t deem above board. Cannabis, it turns out, falls very much into this sin category, meaning that few investors are able to wade into the market, even if they wanted to.

That’s made meaningful funding pretty hard to come by, but that hasn’t kept a few cannabis companies from finding the small pockets of institutional money that do exist.

Dutchie, for instance, has raised $603M according to Crunchbase. They count Tiger Global, a Toast investor, as a backer, undoubtedly operating under the 2012-2021 thesis of “Buy market share with cheap capital and build vertical SaaS.” Dutchie even employs Toast’s former CFO as executive chairman.

Dutchie hasn’t been coy about their desire to own everything, either:

“I always think about it like this: in the restaurant industry, there are dozens of permutations on how a commerce transaction can be completed. Someone can use a QR code on a table in a restaurant; they could use an app with the store’s name on it; they could go to a store’s website; they can order on DoorDash; they can order on Uber. They can eat in the store, or the food can be delivered by the restaurant or by a third party, or they can pick it up. If you’re the core system of record, you want it to be such that any way the transaction happens, it’s going through some version of your pipe. In some cases our brand may be more evident, and in some cases less. In some cases it’s higher basis points, and in some cases it’s lower. Either way, you want to be that platform.”

Tim Barash, former Toast CFO and Dutchie Executive Chairman

But we can only see a few ways this math pencils:

1. CAGR on cannabis swells materially and Dutchie organically grows into a large entity
2. Dutchie pulls a Toast and starts to build all third party applications, with hefty integration fees for non-Dutchie software integrations
3. Dutchie rakes customers on payments fees (also from the Toast playbook)

Let’s table that third thought until we’ve finished discussing the rest of the POS market. 

The early software providers (like MJ Freeway and Biotrack) were built with seed-to-sale capabilities, meaning that the software facilitated compliance and commerce for growers, manufacturers, distributors, and retailers. This matters because certain states require reporting of full custody for compliance purposes, and only a few softwares have been certified.

Over time these software providers experienced scope creep, attempting to get further into POS. This challenge was met by specialized POS software providers that didn’t dare try to build the entire software supply chain, knowing that it’s incredibly hard to do one thing well enough to be relevant. 

There is still a fierce battle between best-of-breeds and all-in-ones. 

Because payments revenue isn’t a given yet (we promise: we’ll get to it), there are virtually no ISOs distributing POS software. Accordingly, it’s not uncommon for even single-location retailers to pay several thousand dollars a month for their POS software. 

That’s not to say there aren’t cannabis POS companies trying to position some version of the “Free POS” we saw in retail and restaurant with payments over the past decade (though it’s more like “cheaper POS” than anything).

Dutchie, for example, monetizes their e-commerce tools, taking a percentage of the sale. They appear to be using these revenues to offset the upfront costs of their POS offering, and they’re clearly going for an all-in-one bundled solution, just like Toast, hoping to expand ARPU with lending, compliance, insurance, and other features on top of the POS.

Just like the restaurant space there are also third party delivery companies for cannabis – like Lantern – but it’s not clear if they’re fully sharing customer data.

Interestingly some of the POS companies are running their own e-commerce platforms that function a bit like third party delivery companies in the restaurant industry: highest bidder gets the customer.

It appears that companies like Dutchie and I Heart Jane partner with retailers, then list the retailers in marketplace searches for products. The retailer that pays the most seems to get to the top, just like third party delivery companies in the restaurant industry. We’ve had a hard time pinning this down with screenshots, but did hear multiple points of anecdotal evidence.

The major difference we see in cannabis, however, is the commoditization of the product. With legalization came professional, pharma-grade manufacturing of an assortment of cannabis products. Manufacturers are logically leaning into reliable, consistent product categories – like edibles and tinctures – and away from the raw flower, which comes with the transportation headaches of shelf life, packaging, and consistency.

Unlike the restaurant industry, where each independent restaurant is essentially a manufacturing facility offering significant product/experience differentiation versus larger chains, cannabis is going the way of retail, with consistent SKUs and a race to lower prices. 

Given high complexity, cost, and compliance, it certainly feels like larger and more sophisticated operators are going to eat market share over time. If this plays out, all-in-one POS companies are going to struggle as larger retailers demand best-of-breed solutions.

In fairness, there are limits on cannabis retailing licenses in certain markets, so MSOs can only grow and consolidate so fast. But competition pulls prices downward over time, and independent dispensaries will have a hard time competing against larger outfits.

And that’s not even taking into account big box retailers which are infinitely more capitalized than pure-play cannabis MSOs. Circle K, for instance, is going to start offering cannabis in Florida (although Florida authorities have not formally approved this), and we expect many more big box retailers – like Walmart – to get into the space over the next decade.

But this won’t happen until the payment wrinkles are ironed out. 

Payments in US cannabis is the wild west.

And not in the good way.

In our research everybody and their brother purports to offer approved payment processing on Visa, Mastercard, and American Express rails – the cards most of the American public already has in their wallets – but it’s all a sham.

Creative processors find loopholes and a partner bank that will allow an MRB to deposit money, but eventually the racket gets shut down and the MRB can lose tens if not hundreds of thousands of dollars when the federal government comes in to seize the cash (and naturally the payments processor is making hundreds of basis points on each transaction the entire time, regardless of the outcome).

That’s because 18 states have legalized cannabis and do allow MRBs to bank their deposits with state financial institutions, but at the federal level cannabis is still illegal. Quoting from the Cato institute’s article:

To be clear, the existing [federal banking] policy does not ban financial services for the cannabis industry. Rather there are just too many legal risks and compliance costs, so most financial institutions stay away… [financial institutions] are required to file reports to Uncle Sam detailing a customer’s suspicious or illegal activities. That can prove costly. A bank can be subject to large fines if it incorrectly reports on its transactions, or if a future bank regulator accuses it of not following the reporting guidance properly. The reporting can be extensive, often covering every single action a customer takes, as it is based on the premise that the illegal activity is happening underground… One small credit union in Oregon that serves marijuana businesses filed 13,500 reports over the past two years for approximately 500 cannabis clients. 

The long and short of it is that the major, aforementioned credit card networks are in no hurry to allow US cannabis transactions over their rails (although they do in Canada, where cannabis is already legal at the federal level). 

And even if the US federal government relaxes legislation on financial institutions by 1) revising the Bank Secrecy Act, 2) creating a cannabis exemption in the SAFE Banking Act, or 3) outright legalizing cannabis at the federal level, there’s nothing forcing the major card networks from accepting cannabis payments. (Interestingly, cannabis chargebacks rarely occur in Canada, so there would appear to be little technical reason for the card networks to avoid opening their rails).

As a workaround we see MRBs using cards as cashless ATMs, but Visa issued a strong warning against these activities in December of 2021. And P2P ACH payments – like Venmo or Cash App – don’t seem to have much penetration.

This puts at risk the large amount of money that companies like Dutchie and Flowhub have raised. 

Our research shows that these two companies relied heavily on a payments revenue stream in their business models, and that could be many, many years out. If it does come to pass, one could logically assume that these companies would need to recoup substantial amounts of money, likely making hundreds of basis points on each transaction with exorbitant payment fees. 

For reference, US payment processors make about 20 basis points on average, even though greedy companies like Toast make 55 basis points (bps), going to 70 points (and likely higher in the short run). 

Dutchie and Flowhub could make Toast jealous, pushing their margins to 200 bps or more.

Because think about it: if Dutchie raised $350M at $3.75B, they’re going to need to triple that valuation to $11B in 5 years (25% IRR). To do that at median SaaS multiples they’ll need to earn $1.1B of recurring revenue in 5 years.

Using Toast’s $10K ARPU per rooftop, Dutchie would need to be in 110,000 retailers.

Uh, well, the US has 7,000 dispensaries today. Even if we assume that count grows at 28% CAGR (which seems super unlikely given the impending stagflationary recession) we come to 25,000 locations.

And Dutchie isn’t going to own all of them.

So Dutchie would have to make substantially more than $10,000 per rooftop. The easiest way to do that?

Rake it on payments. 

Assuming Dutchie owns half the US cannabis market (12,500 locations, which feels like a stretch), they would need to earn $88,000 per store per year to reach $1.1B in revenue – 9x what Toast pulls in.

Sure, the average US dispensary might do $2.5M in revenue, a bit more than twice the average Toast merchant today, but their average check is also considerably higher at around $70, meaning there aren’t materially more transactions where Dutchie could charge a per-transaction fee. 

In essence, it’s a wash.

So Dutchie would legitimately need to ~4x Toast’s payments margin of 55 bps to 200 bps. 

Jealous yet, Toast?

And if Dutchie can’t get that payment margin, expect them to behave like Toast and start choking off third party integrations while they build their own, inferior versions of said tools. This post is now five years old but aging amazingly.

For an example of the polar opposite tech look at Treez, which is solely focused on a best-of-breed POS and maintains open integrations. 

As we wrap up our thoughts, it’s an inescapable observation that the US cannabis market feels dramatically overcapitalized and highly unstable.

Dutchie has laid off at least 8% of its staff.

MJ Freeway, acquired by publicly traded software company Akerna, has gone through layoffs and corporate restructuring leading to stock price reduction.

Flowhub has gone through multiple rounds of layoffs.

Cova eyed an acquisition but it fell through.

Our take?

The cannabis POS market will consolidate. Expect purported features to slim down and fewer all-in-one providers as larger retailers enter the market and demand best-of-breeds. Payments will undeniably be a part of the cannabis ecosystem, but as larger retailers start gobbling market share expect payment margins to look more like they do at McDonald’s than they do at your average Toast merchant. 

Not that appetizing.