Symonds said he refuses to produce customized manufacturing processes for each customer. He must keep his manufacturing application as one core system, but he can add modules as needed to meet certain customer's needs. If a customer needs a manufacturing step or process that his system doesn't do, Plex will create it. Then it's available for any future customer to use as well.
This method of creating manufacturing software has a continuous feedback loop. The more customers, the more inputs on what's needed in manufacturing, the more manufacturing modules and the more powerful the system becomes. By showing the poka-yoke features of his system, Symonds entices new customers to sign up.
There is nothing new in software applications being available on-demand or American companies adopting Japanese manufacturing approaches. But what is new is the ability of forward-thinking software companies to standardize these practices, capture them in one system instead of multiple versions, and make them available to a broad set of customers over the Internet. That may still mean it's not suitable for some manufacturers. Each manufacturing process is different and typically requires a highly individualized system. With Plex, a customer's manufacturing process has to fit its generic, one-size-fits-all approach, with only a few exceptions.
Nevertheless, the Plex approach points to IT as a service, not just software as a service. It is how IT will be delivered in the future. Manufacturing is one branch of ERP, but one day I expect a future SAP or Oracle will have assembled a cross-Plex/NetSuite/Salesforce set of applications into a complete ERP set. You'll no longer read about those multi-million dollar ERP installations gone wrong because ERP will run in a cloud data center, installed by the supplier who knows it inside and out. But this change is still a ways off, and will not come easily.
Older companies will resist ERP as a service, knowing the pain and massive configuration effort their packaged systems required and disbelieving any one suite could be a match on all fronts to their company. But younger companies, seeking the reduced investment allowed by on-demand services, will more quickly adapt themselves to it--or find ways to get an online system to adapt to them.
"Many of today's ERP brands will not survive. The chasm is too deep and wide for them to get a true and sustainable SaaS business model," Symonds concludes.
Hmmm. That sounds a little like how the mainframe was about to be displaced by the PC. Still, the new ERP and other forms of SaaS will incorporate advanced thinking and quality assurance inside the system more quickly than traditional packages can. Then they will deliver IT as a service, not as an installation problem.
The ability of online software to quickly capture new thinking--along with the advantages of being able to avoid the upfront license cost, the need to install a massive package, and the need to make periodic upgrades--will push more and more business users toward adoption of SaaS.
Treb Ryan, CEO of OpSource, a supplier of managed hosting and infrastructure as a service, was also present at the release of "Vision from the Top." His four-year-old, who can't read, can find the applications he wants to use by clicking on icons on a screen, he noted. "Somewhere between now and when he's CIO of a company, the transition to the cloud will be complete," he predicted.
Charles Babcock is an editor-at-large for InformationWeek.
In the new, all-digital issue of InformationWeek Government: More than half of federal agencies will use cloud computing within 12 months, our new survey finds. Security, ROI, and management challenges await them. Download it now. (Free registration required.)