There’s been a ton of great coverage out of Workday’s 2015 Tech Summit, a preliminary list of which I’ve included below and to which I’ll add as others bring their coverage to my attention. But having attended this well-run event and after reading the coverage below and more, I was struck by how little attention was paid to what I considered to be two of the bigger takeaways from this event. Of course, every well-run vendor influencer day leaves me gasping from information overload, and my true analyst/media/influencer colleagues do a better job than I ever could of capturing all the relevant details, which is why I don’t do many such event reports. But now that the dust has settled on this particular influencer event, I thought I might add a little color commentary.
For me the first big underreported takeaway is the fact that not one single line remains of the original code in Workday’s platform. Let me repeat, not one single original line of code remains in Workday’s platform. That’s because the original design provided for changing out whole services, whole blocks of code, for new or rewritten services, adding entirely new services and/or removing no longer needed ones, replacing custom code with commercially available code, to include both open source and acquired code, and so much more while the software is in flight. That’s right, while in full production, with new customers being added all the time, the folks “in the basement,” which is how Workday’s chief architect referred to his own work location, have been updating, refreshing, replacing, rewriting, re-everythinging their applications platform (not to mention the applications themselves, but that’s an entirely different story as discussed below) without any disruption.
So, to the question I always get from financial analysts: “won’t Workday have to replatform soon, just because of all the technological and other changes that have occurred in the past ten years, and won’t that replatforming be a costly, error-prone, disruptive, black hole?” the unequivocal answer is NO. To my very technical colleagues who attended the Tech Summit and may have yawned about this because they know it’s no biggie, let me just say that I nearly fell off my chair and needed a 2nd confirmation (that’s confirmation, not resuscitation). I don’t presume to know how this is done, but I’m blown away that it has been done without anyone being inconvenienced. So much for the FUD around replatforming. And although Workday again said that they have no immediate plans for unleashing their platform in the way that Salesforce does — per Workday, and I agree (not that my opinion on this matters), this is an entirely different business for which the time is not yet right — it’s nice to know that, if and when they decide to go in that direction, they’ll be doing so with an always fresh, always evolving platform.
For me the second big but underreported takeaway is the high degree of object model and framework reuse which, along with the underlying architecture’s definitional development environment, are creating a real competitive advantage in better time, cost and quality to market. Do you remember the punditry from the learning community when Workday announced they would be building their own learning application? Of course that’s a major undertaking, but when you look at the amount of reuse across the objects they already had, those they were building for their student applications, and those needed for learning, the amount of reuse suggested that this major undertaking would be cut down to size. We saw the same thing happen when Workday announced they would be building a recruiting/staffing application, and with the same results of rapid time-to-market, no cast of a thousand developers, plenty of global and compliance coverage, and so on. And every time they look at an application type which lends itself to a framework, so to a generalized but systemic architectural solution, from work books to grid computing to the latest spreadsheet motif, it becomes clear that they’re building additional platform capabilities that can be unleashed across the entire applications suite. I’ve always cringed when hearing enterprise software vendors pounding their chests over their huge development organizations because I know from my own programming days just how hard it is to create anything elegant and lean in such organizations. Thus, I’m naturally biased toward finding ways to reduce the amount of code needed to birth and maintain applications.
Now some will say, and perhaps correctly, that what I’ve described here is the norm among enterprise software vendors, but it’s not been the norm in my experience. Do I need to get out more?
Partial coverage of Workday’s Tech Summit*
“Workday – 2015 Tech Summit Update” from Saugatuck’s Bill McNee.
“Has Workday ceded the cloud platform to Salesforce and Microsoft?” from Phil Wainewright of Diginomica was actually written before the Tech Summit at which the whole issue of platforms was discussed quite thoroughly.
“Progress Report – Workday Tech Summit – Good Progress, More Insights, Less Concerns” from Holger Mueller, which includes link to his Storify of the Tech Summit twitterstream.
“Workday, North by Northeast” from Vinnie Mirchandani.
“Workday HCM Growth Forges Expanded Partnerships with HRO Providers” from Nelson-Hall’s Gary Bragar.
“Workday Tech Summit 2015 – fighting talk as CEO thanks Oracle” from Dennis Howlett of Diginomica
“Workday Works Wonders on Platform for HCM” from Mark Smith of Ventana
* If I’ve left out your terrific coverage, please drop me a note at naomibloom at mindspring.com dot com.
Disclosure: Workday has been a client but my attendance at their Tech Summit was on my own time with Workday covering the bulk of my travel expenses. I was compensated by Workday for being a panelist on their annual Predict & Prepare video broadcast earlier in the week of the Tech Summit, a participation which is totally free of Workday content review or control.
(Cross-posted @ In Full Bloom)