On July 14, 2026, Acumatica acquired ISV partner Vertrax. Why does this matter? While most cloud ERP vendors talk about extensibility, Acumatica has a history of cultivating industry partners to the point of acquisition. 

A classic example? The 2020 acquisition of JAAS Systems, which built an entire manufacturing vertical on the Acumatica platform (I wrote about the run-up to this acquisition in 2019). 

Yes, Agentic AI now takes a big place in software roadmaps, but as I see it, that makes the buildout of vertical functionality more important - not less. Traditional ERP was too often siloed from the industry software businesses rely on. 

AI doesn't work well on siloed data, with old school integration bolt ons. A "single source of truth" is a far better way to run a business than swapping in and out of systems. 

But wait, it doesn't stop there - AI needs platforms, not point solutions. I wrote in The components of AI buyer trust, AI buyers care about platforms and ecosystems, not just standalone vendors. I think it's time for a gut check on where this is all headed, don't you? I did just that with Jon Pollock, Acumatica Chief Product Officer, and Roman Bukary, SVP of Partner Strategy and Programs.

Pollock joined Acumatica eight months ago; Bukary joined even more recently. When we spoke, they were fresh back from Ascent, Acumatica's new partner conference, centered on education for the AI era. I've tracked Acumatica since diginomica started; how would these two hold up to my pesky questions? Start with Pollock: how does his product purview overlap with partners? He responded: 

I've got the product team that includes design, product managers, as well as product operations, pricing, and packaging. My team is at the tip of the spear with regards to defining our AI strategy, defining what things we build, how to partner with our ecosystem. We're knee-deep in understanding usage around AI with our customer base.

Pollock says Acumatica's ecosystem approach to agentic AI is taking hold: 

We have been investing heavily, as you probably know, in building out our AI building blocks inside of Acumatica. We announced AI Studio several years ago, and we've been building out low-code/no-code capabilities for our partners and ISVs, to build AI automations and a number of agentic capabilities within that. So that investment we started a couple of years ago is starting to really pay off. 

The bolt-on approach to AI ERP isn't going to work: 

We essentially said: AI is not a bolt-on for the side of ERP. It needs to be built from the foundation up, from the data models up, because we all know that AI, without direct access to the business-level data, is useless. and so we built it from the ground up in terms of how we how we think about those building blocks.

Acumatica is committed to vertical ERP; more than 80 percent of new customers sign up for an industry edition. But that begs the question: how can partners meaningfully pursue the right verticals with Acumatica already doing so? Pollock says it's about building the "highways" and infrastructure for partners to extend: 

Acumatica is building what I would call the major freeways around agentic capabilities. Think about it by vertical. There are uber-workflows that cross construction, manufacturing, and distribution. If you think about breaking those specifically into the core workflows within construction, there's a half a dozen core workflows. From there, there are dozens of other kinds of minor workflows that go in there. The idea is we build the highways, the major arteries in terms of the capabilities, and we expect that our partner ecosystem VARS - actually our customers as well as ISVs - will build the last mile. 

Last mile customizations will be critical... Partners are already integral to Acumatica's go-to-market; they will play a critical role here. 

[They can] customize specifically to a unique use case within a manufacturing or construction company, and so we're building the highways. They're building the side streets all the way down to the driveway. 

But hold up - most cloud ERP platforms now feature some kind of extensibility. The era of vanilla SaaS ERP is long gone. Isn't really about enabling partners to build out micro-verticals, not just one-offs? 

100%. And to your point, we're doing the uber-verticals, but the micros are are going to be serviced largely by our partners. 

One of the things that's real clear in our mind is our partner model allows us to extend into those micro-verticals where we just don't have the bandwidth - or the resources - to go to that level. 

Ah, but there's another pitfall ahead: the dreaded roadmap duplication trap. How can partners avoid building out areas that are then covered by Acumatica - or other partners? Pollock:

We're creating almost a clearinghouse where we can see what they're building, and then they can present what they're building back to the ecosystem and say, 'Hey, here's some capabilities that we built, that may or may not be applicable' - and so creating a marketplace or a clearinghouse where we can actually present all of those relevant [solutions].

Pollock says the Vertrax sale shows this verticalized approach in action: 

It's vertical-specific capabilities built on top of ERP. In this case, it's the distribution edition. So they took the distribution edition, and then built all of their deep vertical fuel distribution capabilities on top of it, and they bolted on a bunch of proprietary tech on top of it with smart drops - and the other capabilities they have to make it unique.

How did this official tie-up come about? 

They came to us and said, 'Hey, we think there's a deeper partnership here with with Acumatica in its resources and its reach and all of the other stuff that goes with it. We'd love to have a conversation about how we could take it to the next level, and you know those conversations happened months ago, and it all culminated in acquisition. 

Of course, there's a crucial twist now: the intelligence layer. That's how partner-built apps turn into the outcomes businesses now expect. But here's the catch: AI can't do this on its own. It needs those end-to-end workflows. 

Then you layer the intelligence layer on top of it to accelerate those workflows and the outcomes and the insights, and it just goes to the next level. Without that foundation, AI is relatively incapable of transforming an end-to-end business, and so that's why we start with our foundation, build out the end-to-end, and then you do the AI overlay, in terms of how do we accelerate that process.

I have another caveat: I judge the health of partner ecosystems by the morale of its smallest partners. Smaller partners tend to burn for innovation, but they also expose any friction in the app building and marketplace functions. But before I could press that point, Bukary pre-empted me: 

Number three is the fact that you can start small as a partner and build. So we have value-added resellers whose entire area of focus is garage door installers. What the heck does that even mean? It means that the business that typically does garage door installing is relatively small. A lot of it is highly seasonal. A lot of it has peaks and valleys. 

A lot of it is relatively dynamic, because if there's a burst of weather that comes through Utah or Colorado, garage door installers are a lot more busy. It is a manufacturing operation in that they do assembly. It's distribution in that they need to stock their trucks. And if that particular partner extends and builds further, that may become an area of expansion for them or for us. 

Reducing friction for small partners? Big win. Best way to prove it? Get them on stage in front of customers. Nothing shows off the health of a partner ecosystem more than that - yet few vendors really seize the chance to put their smallest (and sometimes most innovative) partners in front of customers. What will Acumatica do at their annual user Summit? We'll find out in a few months. 

For those types of partners, the Vetrax acquisition is an encouraging sign post. Pollock believes Acumatica is in good shape there: "The great thing about Roman's org is he he's got you know a dozen vertexes that are in various different stages of development."

I believe Acumatica's overall take on embedding AI is the right one - including the emphasis on partner education. As I've said before, if an ERP vendor doesn't encourage its partners to disrupt, then the AI disruption will come from the outside. 

Before we wrapped, I asked Pollock and Bukary some not-easy questions about AI disruptions. Why? Because as much as I think Acumatica's contextualized, platform approach to AI is one that will serve their customers best, AI vendors of various flavors are going to come in with their own products and agendas. They will claim, rightly or wrongly, that their models are trained on a customer's verticals. They will claim that they can also pull in a customer's proprietary data safely. I would question whether you can replace vertical (human) domain expertise with an LLM version, but the sales pitch will be made. They will also claim a great AI user experience - and in some cases, they will provide that. 

These are not easy issues for ERP vendors to contend with. A proper/bold response requires shifts across product, go-to-market and pretty much everything else. The timetable to show customers you are on top of this is not forgiving, but cutting corners to get there is a can of worms until itself. These aren't easy conversations, but Pollock says Acumatica is having them: in this AI upheaval, where are the moats? 

We talk about this quite a bit internally around: what are our defendable moats, and how do we see both the large language models, as well as some of these smaller startup [AI-first vendors], and how we're uniquely kind of different from them. You talked about industry depth - that's critical. 

I do think that there is a moat around vertical expertise and industry depth. They might say, "Hey, you know, use a forward-deployed engineer that understands that, and then they can they can leverage a lot of the large language model to build the capabilities.' I truly believe that we have an advantage. I think it's a multi-year advantage, but we have to get traction very, very quickly and build real value into the workflows and into the insights, and all the things that we can provide to our customer base. 

So what does that multi-year AI advantage look like? Well, that's an article unto itself, but here's part of Pollock's take: 

Everything we're building over the next year and a half is all focused on: how do we give the users insights into their business that will help them make better decisions, move the business forward, and so on. Every time we interact with the user and give them a great nugget of advice that's industry-specific, we score some trust points in that scale. We do that over and over again, building that trust over time, so that advice piece is really critical. 

Is part of Acumatica's moat its big head start on consumption-based pricing? After all, Acumatica has been the only cloud ERP with a consumption-based pricing model for years now. Acumatica's team certainly thinks so. As we shift into outcome-based pricing discussions, it will be fascinating to see how Acumatica also evolves. 

Part of a good ecosystem play is also AI alignment, safety and protection against rogue behaviors. Those are not easy questions for vendors to answer. But we should assume that even if all partners in an ecosystem build with good intentions, they could occasionally cross an unintended line - perhaps with customer data exposure, or over-authorizing an agent. A big part of this is the educational component, and the transparency involved in app building across partners. It sounds like Ascent is off to a good start here.

Which leads us to what's next: the Acumatica 2026 R2 release event is right around the corner now; I'm looking forward to hearing more on Acumatica's MCP plans... Expect updates on these pages.