Cloud ERP benefits are about data and automation now. Here's what Acumatica customers have learned - and how AI fits in

Jon Reed Profile picture for user jreed March 11, 2024
Too often, cloud ERP customers fall short of the true benefits they can achieve. At Acumatica Summit 2024, I had the chance to delve into the results of dashboarding and workflow automation. Acumatica has now issued its Principles of Innovation, so how does AI fit in?

Sean Barnett of Spohn Associates at Acumatica Summit 2024
(Sean Barnett, Spohn Associates at Acumatica Summit)

In the month since Acumatica Summit 2024, Acumatica had some self-assigned homework: finish up the principles of innovation.

When it comes to building AI trust with customers, transparency matters. In my Summit 2024 review, Customers sound off on sustainability, AI, and why ERP user licensing matters, I wrote:

Acumatica already has a customer bill of rights. Will Acumatica issue something similar for AI development? Yes, though with a twist. On day one, Acumatica CPO Ali Jani laid out a three part framework for Acumatica's soon-to-be-released principles of innovation - for any new technology, not just AI:

  • "responsible," with guardrails to protect customer data
  • "practical," to address real-world needs
  • "valuable," driven by customer input

Via a blog post by Jani, the official version of Acumatica's Principles of Innovation is now released. And yet, in my talks with customers, AI wasn't the top thing on their minds. How will all this shake out?

I'll take a shot at explaining that, but first, let's look at some key areas where Acumatica customers are deriving value (note: my last piece covered the impact of Acumatica's consumption-based pricing model, so I'll skip licensing here).

I don't want to minimize the impact of moving to a cloud/mobile-friendly cloud ERP platform. But my views on extracting true value from cloud ERP are about pushing beyond a user-friendly system of record into more strategic benefits. Top on that list? Better decision making, and better results with your own end customers (see: Extracting value from cloud ERP in a customer-first world - what should we be pushing for?).

Want cloud ERP results? Bear down on workflow automation

I would argue that without process automation, whatever AI you apply later will be limited. Get your automations in place - then you have scenarios that AI can recommend, enhance (e.g. document recognition) or trigger. What do Acumatica customers have to say?

During out sit down, Sean Barnett, CFO at Spohn Associates, Inc., went deeper into the automation benefits noted in Acumatica's own Spohn case study. I asked Barnett about the Acumatica automations they've achieved to date, including AIA (progressive) invoicing. Has this helped Spohn's construction projects? He responded:

There are built-in functions where I don't have to worry about emailing the project managers to approve their invoices; they just get it. They get an email every morning at nine o'clock: 'Here's all the invoices you have to approve.' We send out automatic emails to our subcontractors for their certificate of insurances, and getting those updated when they're getting ready to expire. We do a few other [automatic] emails, such as under-billings on our projects. Just to remind the project managers, 'You need to bill this - let's go.'

Sometimes, automation is a part of a broader push for process efficiency - and better customer turnaround times. That was the case with the Sunshine Lighting, when I interviewed three key team members: Schneur Perl, Yisroel Hartman and Jacob Hartman. As they told me:

How do we make a difference for our customers? Shipping times. We've cut down picking times... We caught down our picking times from seven and a half hours to about three and a half hours on average. That obviously impacts same day customer shipping.

So all these operational things have affected the customer. Then on the order entry side, being able to supply feeds to our customers integrations - giving our customers live information - and more accurate information - is definitely impacting. Giving them shipping data as send notifications. All this information, we didn't have any of these capabilities before [Acumatica]. It'd be in the dark. They'd have to call in every single time they wanted the notifications. Now slowly we're plugging in those holes. They get live information.

Dashboarding - and learning to trust a "single source of truth"

Acumatica customers have a range of options for reporting and dashboards. This year, I heard from more customers getting results from Acumatica's own dashboarding, including the heavily-used "side panels," which embeds reporting data alongside transactional screens. There are healthy debates on the limits of dashboards. For dashboards to be more than prettified numbers, users must act on what they see. It's not easy to get well-designed, role-based dashboards in place throughout a company.

Nor do users trust the data they see in a new system overnight. That process of data trust plays out over time. While a "single source of truth" may be an exaggeration for some company's systems, it  poses a useful contrast to spreadsheet chaos.

Despite these dashboarding caveats, I'm hearing good things. Start with this feedback from Cassie Moore, CFO at Alpha Insulation & Waterproofing. When I asked Moore about their biggest Acumatica result, she said:

Visibility into data and giving us actionable insights.

This wasn't always the case. Prior to Acumatica, shadow spreadsheets were a coping mechanism for a legacy system. Decision-making was hampered. Moore explained: 

Our previous ERP wasn't reliable. The downtime was extreme, especially during month-end close. And then some of the data that went in didn't necessarily get processed correctly or accurately, to the point that we also had shadow spreadsheets, keeping track of some of these things because we couldn't rely on the data. And then the visibility wasn't there. For the guys in the branch, we didn't have security in a way that we could parse it by branch. So they didn't really get to see a lot of financial information.

Now that's changed:

Now they can see pretty much whatever they want to. I was pleasantly surprised that they really liked that. So instead of it being, 'Here's additional responsibility for you to check your own financials,' they like having ownership.

When I asked Moore for an example of "actionable insights" from this newfound visibility, she gave me a good one:

One of the biggest ones is our total inventory value as a company went down 30%. The way that we were able to accomplish that is the insight into what slow-moving inventory, what buyouts do we need for the projects, like: 'How long is your inventory turnover on average,' and just drawing attention to that and giving those guys visibility into what's causing this to happen. They were able to course-correct. This year, year over year, our inventory value is down 30% for it.

Over at Spohn Associates, Barnett shared what happens when users start to trust an ERP data platform:

The POs and subcontracts - because it's all in the system now. Before, they were doing it in Word or Excel, and had no tracking to go with it. At least now, it's all in the system. Anybody, if you want to see where a PO or subcontract is, you don't have to go searching through the folders and trying to find it. It's in Acumatica - anybody can find it right away.

Plus now, I can actually track things. And we can assign invoices to the PO, so I can look at it and see things like, 'We're over the PO on this invoice.' Whereas before, I'd have to guess and hope, and look at the budgets.

The wrap - how does AI change the cloud ERP benefits discussion?

 I've argued that most customers don't achieve the advanced benefits of cloud ERP - because they don't happen out of the box; you have to push for them. But does this wave of AI and automation change that? Consider this on expense management from Spohn Associates' Sean Barnett:

All of our people, if they're out in the field buying something, and they bring up the [Acumatica mobile app], it captures everything. Then they just have to update the job number, make sure the expense is right, everything has come across right, and just hit save Everyone in our office uses that.

Barnett has even done this on an escalator:

The other day, I took a picture walking down the escalator. By the time I'm at the bottom of the escalator, I was done. It's fast, and it picked up that it was from Denny's; it picked up the right amount. All that stuff. It's already had an impact. I just had to make sure it's coded correctly for me.

Now that this type of functionality ships out of the box, I'll modify my stance. Most workflow automations still involve customers making a concerted change to their processes (and activating new features, customizing a workflow for their needs, etc.). But the more automation that ships by default, and the more "intelligent" it is, the better. We can argue how much of this type of example is automation and how much is "AI," but that's not really the point.

The point is that Acumatica plans to embed this type of functionality in its product going forward. Ali Jani is right: customers won't really care whether it's called AI or not. If it helps them, they'll use it.

So why the new "Principles of Innovation"? These types of guidelines are important, because they set a tone for internal development consistency. Customers have big questions about AI; they are aware that AI is dependent on data - including their data. They want to know how that data will be used.

The principles Jani laid out only take you so far. Every time you roll out AI features, you either build on that expectation of trust, or you violate it. And one violation can tear down a lot of trust. The big consumer AI companies set a move-fast-and-break-things tone when it comes to releasing new AI models. Perhaps they can withstand the PR hits, but enterprise software vendors can't.

Acumatica's Principles of Innovation contain a number of items on data privacy, including:

  • We do implement extensive security and privacy measures for all new features.
  • We do deliver AI results that are aligned with the user access policy, without leaks between users.

I was especially interested in the items pertaining to explainability, such as:

  • We do document and maintain guardrail parameters and provide rationale as to how each parameter is established.

Explainability is a notorious weak point of deep learning systems in general and generative AI in particular. If Acumatica can improve upon that with their own customers, that should be welcomed. But now Acumatica has to prove they mean business. As the Sunshine Lighting team told me:

Acumatica is always busy improving. Knowing that and not being stuck -  because today's technology is going fast, very fast. If you're part of it, you've always got to check in to make sure that it's evolving in the right direction.

A vendor can push ahead, but it's on customers to maintain their own innovation discipline - and, at times, press the vendor with hard/informed questions. That's the type of relationship an enterprise software vendor wants to have with customers. If Acumatica can add impactful AI while maintaining that level of trust, then I like their chances.

End note: Acumatica's partners are an important part of this mix too, including the need for validated partner AI solutions. I got into the AI aspects for Acumatica partners in my last piece. Josh Greenbaum raised some key points on Acumatica's partner ecosystem and where it needs to go next in his post, Acumatica’s Summit 2024: Refreshingly Grounded in Customer Reality.


Sunshine Lighting team at Acumatica Summit 2024
(Sunshine Lighting team at Acumatica Summit 2024)
A grey colored placeholder image