Workday Q&A: COO Talks Oracle, SAP

footprint, like a compensation system or a recruiting system, as a win against Workday. Most of the wins against Workday are preservations of existing SAP, PeopleSoft, or Oracle footprints. We call that a "loss to do nothing," but each quarter, about 20% of our wins are resurrections of previous losses to do nothing.

IW: What's a loss to do nothing?

Stankey: That's when a customer simply defers the decision to replace their core system of record and just augments it with a small edge application. They're just deferring the inevitable for another couple of years. Eventually they will come back around, and they will attempt to fix the larger mess. When they do that, we win.

[Want SAP's take? Read SAP Reports Cloud Computing Tailwind.]

IW: Two years ago, all Oracle and SAP could offer was edge apps, but hasn't that changed with all their acquisitions and Oracle's launch of Fusion Applications?

Stankey: They claim to have complete footprints in the cloud, but when we try to find reference customers running the complete footprint at scale, we can't find them. You have to start with the core employee record, which is the basis for any accounting or HR system for hiring, onboarding, pay increases, transfers -- all the nuts and bolts of any HR organization and operational efficiency. Recruiting and performance management are edge applications. If you try to find customers that have done the core plus edge applications, which is what our product represents, you don't find them among our competition. You'll typically find the edge or, if you find the core, you'll find it in a small company.

IW: What's the "mess" you mentioned and why does it exist?

Stankey: Companies that have grown via acquisitions typically have multiple core systems of record. They may have two PeopleSofts, three Oracles, or four SAPs running their core applications. They'll have just as many performance-management systems and compensation-management systems. They may have an onboarding system from a different vendor, or SAP or Oracle may have acquired those applications and tried to stitch them into their legacy systems.

When companies try to figure out how many employees they have, how many contractors they have, the cost of those workers, and the performance and capabilities of those workers -- so they can do human capital optimization -- they're left with that mess to try to divine the answers.

The fix for the mess, historically, has been data warehouses. You dump all of that data into a warehouse, but it's immediately stale, and it isn't normalized across countries, businesses, or job types. It also hasn't solved the problem. That's the problem that we're attacking.

IW: Can the cloud really solve the problem created by accumulating multiple systems through mergers and acquisitions?

Stankey: Systems tend to become obsolete is in terms of technology and architecture. We're now on our third major version of technology used to deliver the Workday user experience. When Workday was founded, a mere eight years ago, HTML4 was the dominant user interface technology, and it was implemented using Ajax. The market then migrated to Flash for a richer user experience, and we made that move about six years ago. HTML5 then became the technology direction, and we just completed our migration to that.

In each case we migrated our customers, and the only infrastructure requirement on their end was having reasonably up-to-date browsers. We break our architecture up into services, and as technology changes, we're able to advance with it and insulate our customers from technology change.

We update our customers twice a year with new versions, and as things change in countries, whether it be data-privacy laws that have to be enforced in the software, or changes in business processes around works councils in Germany or France, or whether there is some other data requirement placed upon us, we're able to evolve the architecture and deliver it to our customers. Since every customer of Workday uses the same version of the software, we do it once for the benefit of many. We're able to move faster [than on-premises deployments], and our customers move faster.

IW: We've talked to Workday customers, like Brown University, that have spent a year or more deploying the technology. And we've talked to on-premises customers, using SAP Rapid Deployment Solutions, for example, that have taken a similar amount of time to deploy. Is cloud agility overblown?

Stankey: They're using templates and methodologies that say, "Let me make my product fit your industry," but if you want to change the way it works, it's still a customization. You can go live fast, initially, but when you want to roll out and become more sophisticated, the cost- and time-penalties hit you. We're trying to go in fast initially and preserve the agility and the cost advantage. Anybody can build concrete forms and pour in the concrete fast, but when you need to change it, you have to chip out the old stuff, and it's a mess. That isn't the Workday story.

We're building products that are easier to implement. If you listen to everybody else talk about it, you'd think the cloud is simply a location and a pricing model: It runs at somebody else's site, and you can buy it on a subscription basis. But if that's all there is to the cloud, Workday wouldn't exist and Salesforce.com wouldn't exist, because that problem -- location and subscription -- was solved 25 years ago with application service providers. "Your mess for less" doesn't solve the fundamental business problem.

Integrating your private cloud with public clouds can provide agility, security, and control. But getting the minutia right is daunting. Get the new Hybrid Cloud: Details Matter issue of Network Computing Tech Digest today. (Free registration required.)