In November 2019, Plex Systems, which made its name via its aggressive forays into cloud ERP for manufacturers, made a sneaky big announcement:
Plex Systems today announced its manufacturing operations capabilities are now available as a best-of-breed shop floor-specific offering called the Plex Manufacturing Execution Suite (Plex MES).
On some level, this news really surprised me (Plex has had cloud MES for a long time - the news is offering cloud MES as a separate component).
Why was I surprised? Because I believe that to achieve the connected manufacturing visions of the future, cloud-based MES and ERP need better integration, not decoupling. But on the other hand, the resurgence of the best-of-breed enterprise has forced me to rethink: maybe one vendor doesn't have to provide all the core software on a data platform.
Yesterday, I had the chance to dig at these questions via a phone catchup with Plex CEO Bill Berutti. Question number one: why standalone cloud MES? What was the perceived market opportunity?
Legacy MES - an impediment to Industry 4.0
Start with this: legacy MES is everywhere, hindering companies from stepping into so-called Industry 4.0 pursuits - all of which rely on connected data. As Berutti told me:
I think we all know that MES, from a legacy standpoint, there's a lot of old software out there that does its job - but is rooted in 20-30 year old software. And that has all the drawbacks I don't need to tell you about.
Yes, that includes a significant amount of paper-based workflows:
Surprisingly enough, in a lot of companies, large and small, there's still a lot of paper; there's a lot of homegrown stuff; there's a lot of pieces that are nip and tucked together. There's still a lot of manual activity along with legacy technology when you really get into the manufacturing operations on the shop floor... There's a big opportunity for getting into the modern MES age.
Another trend driving cloud MES: Industry 4.0 is getting real. Berutti:
There's a lot more traction occurring around this idea of Industry 4.0... We can throw out lots of buzzwords, but the realness of that stuff is happening right before our eyes. Whether it's for competitive reasons, or just "Hey, I've got to get myself more efficient" reasons, people are beginning to invest in shop floor technology at an accelerated pace.
Industry 4.0 can seem like a visionary aspiration, but it has a practical edge as well:
Faster inventory turns, higher quality, greater amounts of throughput and overall higher levels of efficiency. Whether it's "Hey, I'm going to start turning on smart machines and I'm going to put sensors in my plant," or "I can't hire enough workers to support the increased business volume I have." Whatever the reason is, the move towards the investment in Industry 4.0 is becoming less hype, and more reality.
But you're not getting anywhere in Industry 4.0 if your most important shop floor equipment is a data silo. So, the push for advanced analytics is part of this also:
We know that's happening in every single industry, but in manufacturing, the desire to really understand how to harness the operational information that's coming off of the shop floor, to again drive those same benefits of efficiencies, quality, throughput, customer benefits, et cetera. It is becoming more and more in the forefront of manufacturers' minds.
So that in and of itself requires them to rethink, "Okay, do I have the right physical infrastructure in place? Do I have the right software infrastructure in place to be able to capitalize on that data, and capture it in an efficient means?"
Then there is the surge of cloud adoption in manufacturing, an adoption trend that has fueled Plex's year-over-year gains ("another double-digit year of growth," says Berutti). We can cut Berutti some slack for taking this as a point of pride. Years ago, it wasn't always easy pushing for cloud manufacturing adoption. "We're the only single instance, multi-tenant software company that serves manufacturers exclusively," he asserts.
But that doesn't totally answer my questions. After all, Plex has been offering cloud MES as part of its ERP manufacturing for years. The next big piece of the puzzle ties into Plex's moves upmarket:
We grew the company on the backs initially of top floor to shop floor, full ERP all the way down to the shop floor, MES connected manufacturing. That's really powerful in small and midsized companies that want one throat to choke, and really only want one vendor.
Upmarket, needs are different:
As we moved up market, we really got pulled there by a lot of our larger, multi-billion-dollar global customers, they looked at Plex and said, "Hey, we love all of Plex, but listen, we're not sure it's worth the effort of replacing our financial consolidation system.
We're not sure that we would prefer to use you guys over Salesforce for CRM or Workday for HCM. Those are pretty good solutions in their best-in-class domain, but you guys are best-in-class on the shop floor. You guys are best-in-class in running the plants. And so we want to combine those."
Which brings us to the new best-of-breed enterprise. As I wrote last year:
The new best-of-breed is bolstered by APIs, (micro)services, and a better approach to integration. This puts a new pressure on vendors: adoption is product by product, not "cloud suite lock in."
Cloud MES - do customers want a best-of-breed enterprise?
In theory, this has huge appeal. But enterprise CXOs have long memories - memories which include getting burned by the best-of-breed promises of the late '90s/early 2000s. Can this new wave of best-of-breed avoid leaving customers with brutal integration chores? It better. Data integration used to be about transactional efficiency. Now it's much higher stakes. Unless we want to have a new breed of "disconnected" cloud manufacturers.
I can see the appeal of cloud MES for best-of-breed Plex customers: if they can make a cloud MES move that plays well with their Salesforce/Workday/etc. SaaS investments, now you can push into the impactful shop floor pursuits of predictive maintenance and beyond - without ripping out cloud software that's working, or overhauling your legacy ERP.
Start with the brass tacks issue: SaaS-to-SaaS integration. No, cloud integration isn't some magical cure-all, but: is it significantly easier than the headaches of stitching together on-premises systems? Berutti:
If customers are just going to use us for the shop floor or the plant, we need to make sure that we're intentional about the APIs being available, well-documented, restful, and support all of the functions necessary just to be an MES solution, because this ought to get brain-dead easy. We ought to have proven integrations to third party enterprise software that they're going to want to use it with.
That means stepping up the API game:
We launched, in our announcement, a whole bunch of new APIs, and a bunch of hardened APIs that are available and intended to allow us to be used just in those use cases that we're talking about on the shop floor.
But even with robust APIs, that doesn't mean integration challenges are resolved. Berutti acknowledges the point:
I agree it's not a panacea, but we've made a lot progress.
Version-independent, restful APIs that allow you to leverage integrations over time are important. And also, things that are happening with Dell Boomi and Salesforce and modern integration-as-a-service providers also are helping to make this easier. And we have efforts underway with, for instance, with Dell Boomi and others. So the idea that there are ways to make this problem much easier is a huge part of the answer here.
There is another point we shouldn't overlook. If this best-of-breed cloud approach is going to work for customers, part of that is accepting more responsibility:
A lot of our customers are doing this stuff themselves. They know how to maintain and leverage these integrations even without our help.
The plant tours I've done reinforced a potent lesson: there are big/expensive/breakable machines on the shop floor that cause chaos and big losses when they are down. The ability to "smarten" those machines via sensors and MES upgrades has powerful appeal to customers. It's hard to imagine any type of industrial IoT projects that don't start with pulling data from those big machines. You can see why a customer might choose to start with cloud MES - and leave ERP well enough alone. As long as they are confident future cloud and ERP integrations will remain valid. No surprises here - Berutti agrees.
[Some customers] might be happy with a Workday, with Salesforce. But in financials they might say, "You know what? It just ain't worth it. I'll just leave Oracle in place to do that. Or I'll use Hyperion." That's not what value is, Jon, is my point... The value isn't in better financial consolidation. The value is in streamlined operations and quality improvements in the plant - and how do we help you get that.
The good thing about this best-of-breed SaaS model? It isn't a hypothetical. The market will decide how viable it is. There is plenty of work defining how a data platform across best-of-breed products would deliver. Analytics vendors are certainly salivating at that prospect.
And yes, some customers will insist on the "one throat to choke" appeal of getting all of their ERP/CRM/HCM products from the same vendor (or two). In general, however, I'm inspired by the vision of customers taking more ownership over their software landscapes - whether it's build, or buy, or some combination of the two. Yes, there is some IT pain and investment. But the advantage is picking products with the best chance of user adoption - and business benefits.
Manufacturers won't follow a single template here. But there's little question manufacturers are being pushed out of their comfort zone. The opportunity? Make shop floor data an asset - or fall behind as competitors get moving. Let's see how this plays out.
Updated, 7:30am ET December 9 with section headers, and a few small fixes for reading clarity.