Long post warning: grab a cuppa, turn off the phone
Last evening/night's media and analyst call with Marc Benioff, CEO Salesforce.com and Larry Ellison, CEO Oracle closed out a week which saw the two companies become new 'best friends forever.'
In one important sense it could not have been a better demonstration of how Silicon Valley eats itself and just how easy it is for competitors to re-cast history in a manner reminiscent of the Ministry of Truth in George Orwell's 1984.
However, the call left many questions unanswered.
In this piece, four of the diginomica team co-founders offer individual perspectives on these important events.
Dennis Howlett - unanswered questions
The quality of question asked in the Q&A was, quite frankly, subpar. None of the important outstanding questions were asked. Here was my laundry list:
Is this a nine year or a 12 year deal? The press release said nine years, but Benioff both Tweeted and repeated (without correction) that this is a 12 year deal. Either way, that's an extraordinary length of time over which to commit to a single technology architecture. There is one explanation that isn't as outlandish as it sounds. At some point, possibly as early as in the next two/three years when Ellison reaches age 70, Benioff becomes Ellison's successor in a friendly takeover by Oracle. There are other reasons to draw this conclusion but this provides the strongest evidence because it speaks to stability for the companies in a situation where Ellison's succession has importance to everyone in the market.
Where does this leave Siebel OnDemand and Oracle CRM? Oracle will use Salesforce.com but not standardize on it. This creates uncertainty for customers because one of the most important questions a customer should ask is: 'Do you use your own software?' My guess is that Oracle needs Salesforce.com for deals more than Salesforce.com needs Oracle. Despite insistence that competition continues, this could easily end up as 2+2=5 play. Once we start to hear about 'joint customers' then we'll know the answer.
How much re-engineering will be involved and how disruptive will this prove to be for Salesforce.com and Force.com platform developers? The Oracle 12c database is new with many very different features that will help Salesforce.com grow without having to worry about an 'old' architecture. Continuing with the developer meme, I wonder what this means for the 'bring your own language' mantra that Salesforce.com developer leads have preached. Now, Java reigns supreme in the Salesforce.com reality distortion field. That will not sit comfortably with those who prefer other languages and tools.
Salesforce.com was always the beacon for disruptive technology players. It was the alternative many customers wanted to their monolithic suppliers like SAP and Oracle. How does this change the landscape? Can Salesforce.com remain that bolthole or will we see it slowly sucked into the Oracle orbit? Today, both companies are trying their best to pitch this as good for customers while maintaining a healthy distance. I'm not so sure that customers will see it that way. Appirio recommends caution.
Stuart Lauchlan - a long week
We have always been at war with Eurasia!
I also find myself with a lot of questions still lingering.
This probably doesn't reflect well on me but I found myself almost disappointed with the bonhomie between the two CEOs. I know from a mature, adult, sensible, customer-friendly PoV, it's so much better to have Larry and Marc on speaking terms, but a little bit of me suspects we're going to miss the sparring now that the 'roach motel' has been upgraded to the kind of place where Larry would stay.
How long will he stay? Dennis reckons BFFs. As he notes, the original statement of intent on Tuesday talked about a 9 year technology pact. By last night Benioff was talking about committing to 12c for 12 years. A moveable feast? Or did someone realise that 12 years on 12c was a nicer bit of messaging? Either way, we're in for keeps - for now at least. I'd love to know BTW who blinked first? Who made THAT phone call to the other and set all this in motion.
I still want more detail though. I'm frankly confused by the seeming conflicted statements about who's using what and where and when? For example, fifty Exadata boxes are going into the SFDC data centre. Fine. Does that mean hundreds of Dell servers are coming out? Maybe the analyst call wasn't the place to discuss such unpleasantness, but it would be good to get some clarity on this.
Dennis picked up on the Siebel question. I echo that and add in the idea that SFDC is going to use Oracle Fusion HCM but that this doesn't mean they're not going to use Workday? It's not a zero sum game, I'm told. Fair enough - so SFDC will have 2 HCM systems internally? One of which will come from a provider that the new BFF loathes with a vengeance? One solution to the Siebel question is the suggestion that companies acquired by Oracle will use SFDC while 'native' Oracle-ites will be using Fusion Sales. Feasible, but does that map onto HCM? Call me a tad bewildered.
Finally - for this week at least - I'm intrigued by comments from Constellation Research's Ray Wang who's been arguing that customers feel "betrayed" by Benioff's new leaning towards Oracle. There's certainly some tricky conversations to be had I suspect.
If you've driven a revolution and fired up the angry masses against the old order and then suddenly reach detente with the establishment leader, there's going to be a few armed rebels up in the mountains who just don't want to come down! The provisional wing of the cloud computing movement?
While I haven't had Ray's experience of hearing from "betrayed" customers, I did quiz some CIOs and ITDMs at Cloud World Forum in London on Thursday, a couple of whom said they were "disappointed' by the move. (It's that British understatement thing going on perhaps?) This is going to need some careful messaging going forward I predict.
For what it's worth, it seems to me a good move for the wider enterprise customer base of both Oracle and SFDC (and beyond?) and for the cause of further legitimising cloud computing as a mainstream delivery channel. I just want a bit more clarity - and doubtless that will emerge over the weeks to come. Larry's coming to Dreamforce in November. Will Marc be back at OpenWorld in September? I can't now imagine he won't be. We're going to get used to this double-act.
One last (slightly frivolous) observation. I've long been fascinated by the question of the succession planning at Oracle for that dark day when Larry decides to dedicate more time to sailing. Changing CEO when that CEO is such a figurehead and a founder of the company can be tricky. See Microsoft for how it's done and Apple for how it isn't. Larry's the hardest act to follow. Did we just get a glimpse into a possible future this week I wonder...?
Phil Wainewright - three observations
1: Whither Siebel?
Whereas the partnerships with both NetSuite and Salesforce.com give a boost to Oracle's Fusion HCM and Financials products, it's a different story for the CRM product line, including Siebel and CRM OnDemand, as both Dennis and Stuart have mentioned. Enterprises that have remained loyal to the Oracle CRM portfolio must be wondering what the new partnership with Salesforce.com means for them.
On yesterday's call, Larry Ellison made it clear that Oracle will stop short of actively selling Salesforce: "I'm not sure that's where they need our help. I'm sure they're better at that than we are," he said. But the mere fact of giving Benioff's sales teams free rein to sell the cloud CRM package to the Oracle customer base can't be good news for Oracle's own bag-carriers. That in turn suggests Oracle's CRM products won't be at the front of the line for further investment.
It's uncharacteristic of Ellison to concede defeat without having a comeback strategy. Pragmatism has dictated conceding the CRM market to Salesforce.com for the short-term gain of competing better against SAP, Workday and others. In the long-term, I'm with Dennis: acquiring or merging with Salesforce.com is on the cards, because it would regain the ground conceded this week. Therefore, customers (and rival vendors) who want to avoid being locked into the red corner will want to look at alternative allies in sales and marketing automation, including SAP, Adobe and (insert whoever ends up buying) Marketo.
2: Internet of Toothbrushes
A little light relief now. It's unfair to make fun of someone whose habitual eloquence has been partially derailed by jetlag, but I guess I will anyway. I couldn't help wondering where Marc Benioff was going when he started talking about smart toothbrushes that record his daily toothbrushing habits and reports them to his dentist. This hot on the heels of revealing that he was wearing two computers on his wrists and a brief reference to reports this week that researchers have developed electronic pills that send body metrics from your stomach over wi-fi. Here's the gist of his line of comments on the topic from my notes - make of it what you will:
"Wearables .. ingestibles ... we are in a new world," said Benioff. "Everything's connected. This new world of services, all these heterogenous devices, we have this beautiful capability now ... you're going to be able to build very sophisticated, complex yet very easy-to-use apps that work in this new world ... The value that's going to be created is just going to be epic."
Internet of Things (which as we now know includes the Internet of Toothbrushes) will be a big theme (epic, even) at this year's Oracle OpenWorld event, promised Ellison.
3: Exadata in the datacenter
As I left the Cloud World Forum show in London yesterday afternoon, I noticed a poster declaring "9 out of 10 SaaS providers are powered by Oracle." Whatever the truth of the statement, what we do know now is that at least one SaaS provider is going to be powered by Exadata. But does that mean others are going to make the same choice? Even more importantly for Oracle's hardware revenues, are enterprises going to do likewise?
I have to say I didn't see any Exadata boxes on my tour of the new cloud-friendly Equinix datacenter in Amsterdam last week. The conventional wisdom in the cloud arena is that you scale out across cheap commodity hardware rather than scaling up on huge monolithic systems. While it's true that traditional server vendors like HP and Dell are getting squeezed by the advent of the cloud, the main reason is the increasing popularity of custom-built systems from OEMs such as Quanta rather than because companies are turning to Exadata servers, or indeed IBM mainframes.
As I wrote yesterday in my analysis of the Salesforce.com and Oracle deal (scroll down past the humor to "Acceptance" for the serious takeaways), Salesforce.com has opted for a scale-up architecture that's not typical of what most SaaS providers do:
"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."
I'm waiting to see more detail of the Oracle 12C announcement to find out exactly how the latest iteration of the Oracle database scales out across virtualized databases and will also look with interest at how that's implemented in the Exadata boxes. Factors such as power consumption and cooling are a big deal in a cloud datacenter and if the combination of 12C and Exadata is delivering significant savings there then it may prove more widely attractive.
But on the other hand, the deal that Salesforce has undoubtedly been able cut to become such a powerful reference customer for Oracle is probably at a bargain-basement price that won't be offered to others. In the long run (which is quite a few years before this 9- no make that 12-year deal expires) commodity platforms tend to win out over proprietary scale-up alternatives. So I'll need a lot of convincing before I'm prepared to say we're going to see others following Salesforce.com down the Exadata path.
Jon Reed - who came out ahead?
I’m struggling with this one. And not just because I’m going to miss the ‘false cloud’ (Benioff) versus ‘what cloud?’ (Ellison) fisticuffs of keynotes past.
The flurry of Oracle announcements this week wrecked my narrative of how the enterprise software market was going to play out. I anticipated more of an ‘alliance-of-upstart-cloud-players versus old ERP guard’ face off, say in 2015.
Here’s where my analysis was off:
- I underestimated Salesforce’s dependency and/or close relationship with Oracle from a database angle.
- I overestimated the natural alliance I thought SaaS players like Workday and Salesforce.com would have against the incumbents.
- The possibility of Benioff as a successor to Ellison seemed ridiculous to me. Now the topic has graduated to legit fodder for speculative bar talk, and who knows where it could go from there.
Here’s how I see the week’s events from a vendor angle:
It was a very good week for Oracle. It was also a validation of the imperative of cloud data security. On the press call yesterday, Benioff acknowledged that Oracle gave Salesforce.com a range of options when it comes to securing cloud data for customers going forward. We’ll learn more about this as 12c and future releases are unfurled.
It was a good week for NetSuite.
I’m not sure if it was a good week for Salesforce.com, but it was a necessary one. If you take the long view, these announcements are a bigtime validation of cloud business models, so in that sense, Salesforce.com wins.
It wasn’t a good or bad week for SAP, because SAP already knew what it was up against. Contrary to popular opinion, SAP’s biggest enemy is not Oracle, or Workday, but itself. It’s the urgency of internal re-invention versus entitlement culture. I don’t see how this week changes what SAP needs to accomplish.
I don’t see how you can argue this was a terrific week for Workday, but: its stock went up 3 percent the day after the Oracle-Salesforce.com announcement, so I get that annoying feeling I am missing something. Ellison and Benioff’s comments in yesterday’s press conference about the potential of ‘one click integration’ between Oracle HCM and Salesforce.com could force a revealing customer vote between deep cloud integration (between Oracle/Salesforce.com) versus the arguably superior Workday user experience. I don’t see Workday’s CRM play as clearly as I thought I did a week ago either, but Workday has the momentum and time to make its own fate.
From a customer vantage point, I’m in grumpy skeptic mode - for now. Customers will certainly appreciate the performance benefits and painlessness of enhancements these announcements imply (though new license costs will surely go under the microscope).
But do large enterprises get more competitive choice from these new alliances? The SME and mid-markets already have a healthy range of ERP and cloud options. With the large enterprise market, you're gonna have to sell me.
Look, I’ll fess up: I was rooting for a cloud ERP upstart to challenge the incumbent vendors with an entirely different licensing model and customer experience to consider. There was no one better at railing against the incumbents from the keynote stage than Benioff. Workday may prove to be that challenger, with or without a strong Salesforce.com alliance. Infor is making a push, though there are legacy apps to modernize there also.
Perhaps what is happening is more nuanced, that the upstarts are stabilizing with long term alliances and the incumbents are getting faster and smarter. If so, customers may indeed be the winners. For now, I’m declaring wait-and-see mode. Hopefully more disruptions - if you will allow me to use that word without gagging on your breakfast - are coming.
Disclosure: at time of writing, Oracle, Salesforce.com, SAP and Workday are premium partners of diginomica.