It's been almost two years since I wrote about Cuba Flowers and his mission to fuse DevOps with cloud ERP in Totally cannabis, and aggressively DevOps - inside Privateer Holdings' cloud ERP pursuits. Two years is a lifetime in the smoking hot cannabis industry.
So, in advance of the Acumatica Summit event in January, I got Flowers on the phone to find out what's new. Turns out a heck of a lot has changed, including his fundamental cloud ERP strategy.
First things first: Flowers' job title has changed also. He's moved from CIO of Privateer Holdings to CIO of Tilray, which was a Privateer subsidiary. In April 2018, Flowers and a handful of other executives were moved to Tilray to focus on the pending IPO and the intense growth of the cannabis market.
A growth surge puts new expectations on Tilray - and Acumatica
Now, Tilray, or TLRY on Nasdaq, is a separate entity, with Privateer as the majority shareholder. Tilray's cannabis market epicenter is currently Canada, but Tilray is incorporated in Delaware. Flowers assures me that's the short, less complicated version. Flowers is now CIO of Tilray Inc, and to say the IPO went well is an understatement. As Flowers told me:
The IPO went really well, incredibly well, actually. We were very successful in that, and now the industry has changed a lot, too. Canadian legalization had not happened in 2017 when we last talked. Now we're in a new era of Canadian adult use, in addition to all the other changes that have gone on.
Tilray's focus remains, as they put it, "pioneering the future of medical cannabis globally," but the Canadian demand has exceeded all forecasts.
There was an expectation that once adult use became legal that medical would take a large hit, and that didn't happen. That surprised everyone pretty much. That's part of why we have a supply shortage. It honestly doesn't surprise me terribly much. I've always held a theory that having cannabis more accessible and visible out in the market will encourage bonafide patients to try that as a medicine.
Readers are probably jaded about "high growth" hype, so let me give you a stat: ten months ago, Tilray had 250 employees. They will pass 1,000 soon. Yes, it puts new expectations from Tilray onto Acumatica. And, get ready for this: Flowers isn't as much of a DevOps guy anymore.
Yes, Flowers still believes in agile. He still buys into Acumatica's frequent SaaS updates. But his plan in 2017 to bring Acumatica into his private cloud is on the shelf. Flowers doesn't have time to be an IT shop:
Remember my vision back then was to bring a lot of modern software practices to the industry? Well, it turns out, we've been growing so fast. I guess we do work in somewhat of a traditional waterfall model after all, but we're using really modern tools like Acumatica.
Now it's all about ease-of-integration, with off-the-shelf cloud components:
We have done almost no custom software development at all. A little bit on the medical side with e-commerce, but we just can't move fast enough if we do things custom.
Best-of-breed SaaS is the priority:
We really look for best-of-breed, out of the box solutions, which is why we went with Magento Enterprise to replace our e-commerce portal on the medical side. We do end up moving very fast, mostly with buy-before-build, so we find ourselves as more of an integration team a lot of times than a software development team.
One thing that struck me: Flowers points out that this growth surge places new challenges on Acumatica.
What's interesting is that Acumatica has a bunch of customers that are growing up, and we're one of them. Like I said, our headcount is going to crack a thousand in 2019. Acumatica's challenge is going to be, even though they're targeted at mid-market, how do they effectively support their very successful customers that grow out of the mid-market?
And how is that going?
They're committed to doing that. I feel like they're a partner, and we're a partner with them on that journey.
Facing cloud ERP integration challenges
Now Tilray has integration needs - a whole new world of challenges. Other cloud solutions don't always have out-of-the-box connectors to Acumatica.
That's the challenge of using a smaller market share product like Acumatica. You don't get a lot of those connectors out of the box typically. As an example, third party HR systems will probably already have these connections to Oracle and Dynamics and Sage. Often those things don't exist for Acumatica.
So how is Acumatica addressing this issue?
Again, I'm really happy. The reason I've been happy with Acumatica this whole time is they're a great partner. If they don't have something, they will work really hard to provide me that connector. They'll make calls; they'll start talking to those companies. They'll see if they can develop it and resell it. If none of that pans out, then they'll very likely work with me on a co-development, or splitting the cost of an integration.
Testing the cloud integration argument
If your IT department is consumed with integration chores, relying on best-of-breed solutions to focus on growth falls apart. Companies got burned by this exact dilemma with the on-premise best-of-breed movement of the late 1990s/early 2000s.
Over time, I've been persuaded by customers that cloud-based integration is much more manageable - though not without risk. Several components must be in place:
1. Use of standard/out-of-the box SaaS functionality wherever possible. Configuration that doesn't create unique setups/conflicts with the maintenance of APIs and standard connectors. No custom code. Add-on apps and extensions are used with discretion, and must adhere to open platform standards, not proprietary code.
2. Selection of vendors that adhere to open standards, with well-designed APIs across the product line. Big bonus if they have a (micro) services-based architecture or are moving towards "decomposing" their monolithic apps into consumable services or workflows.
3. Emphasis on well-maintained connectors between vendors that assume responsibility for ongoing integration - rather than pointing you towards their API documentation and wishing you good luck.
Flowers' Acumatica experience appears to meet these criteria. I asked Flowers about point number three. Would he avoid vendors that just pointed him to a set of APIs without assuming some responsibility?
Definitely. Because we are not running an in-house software shop, so where I can either buy the connection or the integration off the shelf, I will. My second step would be to try to put it on the vendor themselves to integrate with Acumatica, and probably my third choice would be to have Acumatica write the integration to that product.
Only if all of those paths fail would I look at actually outsourcing the development of the integration. But again, I probably would not do it in-house.
The wrap - growing in a volatile market
We talked in detail about the cannabis market. Flowers described the situation as "volatile," but also "very explosive" from a growth standpoint. The Canadian legalization on October 17 defied expectations:
Honestly, the demand outstripped virtually everybody's projections, to the point where provinces are reporting they're only getting about 25 percent of what they want to order across the board. There's a huge supply shortage. All of our online stores and retail stores ran out.
We're all trying to grow just as absolutely fast as we can to scale up production, so that we get that market share, we get that shelf space, and that's a great example why we need to go so quickly.
If Acumatica didn't provide the right functionality for the cannabis industry, that would pose a big IT dilemma for Flowers. He told me Acumatica already has most of the functionality they need. They are working with Acumatica on the agriculture side:
We brought Acumatica's leadership out to our High Park farms in Ontario, which is our biggest growth site, because we were having some interesting challenges on the agriculture side. This would be a limitation of any ERP that is more geared towards discrete manufacturing. That's pretty much what we are, as opposed to agriculture. We do both.
They came out and toured our facility to watch how our grow teams actually work, how they track labor, how we report on that, how we want to run standard costing from seed to sale. We were working on them with on some data projects and configuration projects to help smooth that out on the upstream - what we call the upstream side - which is the grow.
Flowers told me they can already track their entire "seed to sale" process inside Acumatica (though for Tilray, it's really clone-to-sale, not seeds). But he sees an even bigger opportunity for Acumatica adding/integrating agriculture and discrete manufacturing functionality in the future:
I'm sure they are looking at the cultivation handling in ERP, because we have been quite vocal about that. That's a problem they're going to want to solve.
But solving problems is a doable thing - with the right partner.
I think even more important is just that they created a modern, flexible, nimble program relative to other ERPs. Because of that and because of their size, they can sit down at a table with me and listen to my needs and get that implemented in their roadmap - not a three-year plan like legacy ERP vendors would do.
Tilray has ambitious plans beyond Canada. They are now in 13 countries, and are watching the U.S. market with interest. Potential mergers/acquisitions are also possible, which would bring Flowers a whole new integration challenge. For the U.S market, Flowers says:
You can't overstate the importance of the US market. My guess, and this is my personal opinion. I would not be surprised to see federal prohibition come down in the next two to three years, regardless of what happens in the White House. [meaning: legalization across the board in the U.S.].
Sounds like we better buckle up - interesting times ahead. This model of a customer and vendor growing together - and pushing each other - is, in my view, the winning one. I look forward to the next update.
Update, January 6, 2019 - what about the "one throat to choke" objection? Over the holidays, I thought about the one objection I hear the most about best-of-breed. Some CIOs and tech managers worry about losing the ability to push problems to one vendor. I believe the insecurity about this can create an unhealthy lock-in. But to be fair, I do think it's the question that must be reckoned with:
So I put the question to Flowers: "Do you ever worry about not having as much of a "one throat to choke" approach so to speak, as would be the case if you got all the components from the same vendor? How do you see that tradeoff?" His response pointed to a middle ground. And: the advantages of avoiding your own integration code.
As far as holding a single vendor accountable, yes, we like to minimize the number of vendors in play, but a company can only reduce to a certain point, since few companies make products covering ERP, PLM, QMS, LIMS and HCM and everything else we need. I feel comfortable working with multiple vendors, but I prefer that the vendor write the integration so that we can hold them accountable for it.
End note: I don’t partake in cannabis or alcohol, but I don’t believe in the criminalization of such users. Though medical cannabis is promising for chronic pain, I do not personally endorse the consumption of drugs for recreational use. I derive inspiration from the punk straight edge movement, though I have no plans to get a tattoo.