The short answer is 'no.' Nothing happened that changed anything of any substance from what we heard in June and it doesn't matter anyway. The longer form answer takes a bit of working through and that does matter from a buyer's standpoint.
Understanding Valley Fever
Regular readers will know this but the casual or buying reader won't necessarily be aware that technology companies say a lot of things that on their face sound polemic but which aren't. Equally, they hint at big things but in a muted terms. It's all part of the language surrounding the endless fashion parade that the technology industry has become. Most of us get sucked into it from time to time. I call it Valley Fever. Against that backdrop this is what we are really seeing:
Salesforce.com and Oracle
In answer to Doug Henschen's direct question about Salesforce.com's applications landscape, Marc Benioff, CEO Salesforce.com said the company had made the decision to stick with Oracle financials for its internal use but that Workday HCM is their HR solution of choice.
Neither of those statements has the slightest bearing on Workday other than the fact that for the time being at least, Workday cannot look at Salesforce.com as a potential financials customer. And neither should it. Workday financials still has a ways to go before becoming a serious large scale, global enterprise competitor. Everyone who's looked at it and poked around in the detail knows that. So no news there.
Going back to June, the real story was about the long term agreement made between Salesforce.com and Oracle under circumstances where Salesforce.com needed to renew its database license agreement. There had been talk that Salesforce.com was attempting to find an alternative to Oracle but that it was unsuccessful. That fits with what was announced as a logical extension of the status quo. As our own Phil Wainewright said back in June:
Salesforce.com approaches multi-tenancy differently from most other SaaS providers. Its architecture consists of very large instances that run on highly scaled-up systems, whereas most SaaS providers replicate much smaller instances distributed across scaled-out commodity systems. Putting those scaled-up instances onto Exadata boxes merely takes Salesforce.com’s architecture to its logical conclusion.
From past conversations with Salesforce.com engineers, it is clear that although the company uses Oracle DB, a casual Oracle database administrator would struggle to recognize it as such. Salesforce.com has made countless modifications in order to operate at massive scale. Paradoxically, this makes moving off Oracle extremely challenging.
Therefore, while the cognoscenti might argue at length on this topic the fact is pragmatic decisions were made. And for as long as customers are not impacted - who cares other than Valley Fever infected people?
Workday, Salesforce.com and Oracle go to market
Another topic of discussion that Marc Benioff squished flat talks to the go to market piece. He was clear on this point in June and yesterday. Salesforce.com and Workday do not go to market together as a matter of policy.
There may be plenty of occasions where that happens in circumstances where a potential customer is looking to make wholesale change but it isn't policy. Neither do Oracle and Salesforce.com go to market together. Why would they when they are direct competitors?
In similar vein, if Salesforce.com needs Workday or Oracle to close out a deal you can bet it will choose whichever partner is best profiled for the occasion. That is sensible business. So again, no impact on customers and nobody cares.
Workday, Salesforce.com and Oracle - the integration piece
Again we see similarities. In June, per Henschen:
"When customers choose cloud applications they expect rapid low-cost implementations; they also expect application integrations to work right out of the box," said Ellison at the time. "That's why Marc and I believe it's important that our two companies work together."
Yesterday? From my reporting Benioff said:
“This is primarily an engineering release. These phenomenal solutions are better when they work together. These are the kinds of partnership we want."
Now...that was tempered by a veiled reference to SAP when he also said:
"Increasingly, customers will have an aversion to these kinds of [black box] relationships.”
Mark Hurd, Oracle president in fielding questions on the earnings call said:
"...it's significantly easier obviously to have very integrated and very seamless integration when you own all the products as we do for those customers that have mixed environments, especially the large installed base of one of our competitors. The big benefit, of course, in working with us is that it's all based on our technology. I think it gets significantly harder when you're talking about almost random companies who are trying to make the experience seamless when their systems are so very, very different, the definitions, the underlying platforms are very, very different."
Any problem here? No. Both Hurd and Benioff are stating the technically obvious. However, that's not to say that Salesforce.com and Workday will have an easy time of it building pre-packaged integrations that customers really can use. The thriving business in integrations by the large systems integrators didn't just vanish overnight. If anything, the announcement creates more good news. Here's why.
Workday and Salesforce.com - the analytics piece
I wrote at length on this topic yesterday and here is where I see a mutual benefit that may not be so obvious in an Oracle-Salesforce.com scenario. As an example:
The promise of tying together financial, sales, marketing and HR data provides (near) end to end real-time analytics possibilities for joint Salesforce.com and Workday customers. This is something that has never been possible inside ANY landscape without huge expense.
Of course the realité is highly dependent upon the extent to which Workday and Salesforce.com can make easy integration a job well done. There was no announced timeline and given past performance it could easily be a year before we hear anything substantial.
Nevertheless, during the analysts' technical event last week, Workday's Stan Swete, CTO, talked at length about the performant nature of changes the company has made in its architecture. While it all got a bit muddled as a discussion topic, the company said it can now process 200 million journal line items and report back in less than five seconds. It's not a huge number when working at enterprise scale but it's still a good starting point when considering real time operations. Later, the company hinted it will blow past that benchmark pretty quickly. No doubt we will hear more as new releases come into focus.
Salesforce.com systems including Chatter are generating billions of pieces of data. While there is no problem processing those at the transaction level, analysis has very different technical requirements. It is that part of the equation that Workday is concentrating upon. Workday is working out how to merge meaningful unstructured data with structured data from the various systems of record to offer valuable insights across the whole business landscape. A tantalizing prospect.
The tools Oracle has to offer, while powerful are long in the tooth and still require considerable IT involvement. Both Workday and Salesforce.com have tried to keep the IT workload to a minimum. I see no reason why they should not continue on that quest although it gets increasingly harder to sustain as data types proliferate. Hence the good news for SIs.
It is still good news for customers in the long haul but we don't know the end unit cost. It certainly cannot run the millions of dollars that Oracle and SAP expect from their technologies.
One warning. Nothing was said to indicate how the price points for these types of analytics will hash out. I anticipate seeing highly significant value to the point where data value extracted is worth way more than the applications themselves. But neither workday nor Salesforce.com should look at this as an opportunity to plunder the customer base. That would be a big mistake.
Much ado about nothing? Not quite and not in the directions most observers might secretly wish. Forget Valley Fever antics. This is about two very high growth companies seeking to take as much off the legacy vendor table as possible in the shortest amount of time. Let's not forget that despite stratospheric stock valuations and impressive growth, both Salesforce.com and Workday combined are comparatively small when stacked up against an SAP or Oracle.
Analytics matters more than at any time in the past and this is where I see the greatest potential. But that only works if the integration piece is well executed AND Workday continues to find new ways to consume any data in real-time and at scale. Having seen what Workday's technical teams have done with a relatively clean sheet of development paper, I have little doubt they will come back and amaze us.
The differentiating value here is probably more about the way both companies openly care about their customers in ways that are not visible inside the legacy customer base. Workday's Aneel Bhusri said on a number of occasions that 'this is what customers want and expect of us,' a sentiment echoed by Benioff. Their joint continued existence depends upon this service quality. In the recent past, we have seen Workday quick to rough up partners who do not perform. That alone should provide continuing comfort.
While we like to think of customers making rational, objective choices, those same customers are people who want to feel that their partnership with a vendor means just that. On what we see evidenced in the market place, that's what Workday and Salesforce.com are delivering together. I'm not so convinced the same can be said of Oracle and Salesforce.com combined.
Bonus points: Holger Mueller has a technical assessment of the integration challenges, pointing to the potential issues for buyers, sellers and developers.
Workday hosted an analyst briefing on the features in Workday 20, the latest release of the company's software. Much of the discussion centered on what they call Big Data Analytics. Here are some points I picked up for consideration.
- As expected, they talked about ingesting data from third party sources with the obvious example of sales data as a way of measuring performance in talent assessment scenarios.
- There was no obvious or clear message around master data management (MDM.) This is important because the semantic meaning of terms used in different systems often means that mixed data becomes 'dirty.' The common examples are 'customer' and 'contact' inside sales and accounting systems. Instead, Workday talked about custom fields, another way to populate a report grid for analysis purposes. We will need to see how customer experience triangulates against Workday's intent.
- Workday is not selling analytics without some other Workday product commitment such as HR or financials. I think this is a missed opportunity for broadening the engagement to Force.com developers. As far as I can tell, it would not necessarily burden Workday development but it would open up the market.
- Workday is seeking to make analytics projects cost effective by taking the security and permissions topics off the table. Who gets to see what will therefore be configured inside the Workday system according to the preference policies a customer chooses. This is a net good.
- Workday is emphasising native support for iOS devices such as iPad and iPhone. There is a generic Android client in the Google Play Store but no immediate plans to natively support (say) Samsung as a device maker in the same way it supports Apple devices. Instead there is an HTML 5 client.
Image credits: Stuart Lauchlan and Gapingvoid
Disclosure: Oracle, SAP and Workday are all premier partners.