Yesterday, Salesforce.com pre-announced a cloud based database. On its face this sounds exciting. Or at least you'd think that from the drooling Tweets I saw. As Marc Benioff, CEO Salesforce bounced around on stage extolling the 'amazingness' of Database.com I just could not get excited about what to me seems like a breaking down of the Force.com platform into bite sized pieces for individual consumption. If that's what you want to do. For me, the more pressing problem comes in figuring out what does a 'record' or 'transaction' mean in Mr Benioff's cloudy world? Especially when the hook sounds like a generous free for three users plus up to 100,000 records and 50,000 transactions per month. More to the point, when does a user hit the trigger point for pay to use?
Bob Warfield has a partial answer:
Who will be the first to put up a service on Amazon AWS that delivers exactly the same function using MySQL and for a lot less money? You see, Salesforce’s initial pricing on the thing is their Achille’s heel. I won’t even delve into their by-the-transaction and by-the-record pricing. $10 a month to autheticate the user is a deal killer. How can I afford to give up that much of my monthly SaaS billing just to authenticate? The answer is I won’t, but Salesforce won’t care, because they want bigger fish who will. I suspect their newfound Freemium interest for Chatter is just their discovery that they can’t get a per seat price for everything, or at least certainly not one as expensive as they’ve tried in the past.
I was more blunt about it: It won't fly. Period. Such trifles aside, Phil Wainewright reckons Salesforce.com just steamrollered many of the situational app vendors:
Of course they have established customer bases who will remain loyal, and which they will continue to serve, for many years to come. But the impact of Database.com on their ability to attract new prospects must be highly damaging.
He backs up the contention with a comment from Matt Robinson, CEO of Rollbase:
“Regarding the impact of this on Rollbase and others like LongJump, we will likely feel it in the form of less prospects for our respective hosted offerings,” he told me. Indeed, the momentum of Force.com has already pushed Rollbase towards providing its platform as an installable package for use by ISVs or within individual enterprises.
OK - so Phil caught one CEO with his eyeballs caught in the Salesforce headlights. That does not make a wholly convincing case.
Despite Phil's reporting from a visit to the oracle (sic) I think he's way off base. Bob Warfield offers the open source card as evidence that Database.com is not the killer Phil believes. I'll offer different fare: the millions of Microsoft and SAP developers who are not going to abandon what they know (and make plenty of money upon) any time soon. While I am willing to bet that as outlined below, many of those will take a sniff at Database.com, few will bite in the short to medium term unless they see significant development advantage in going the Database.com route as tied to Salesforce. Right now, the cost model alone won't cut it.
It took an early evening shared beer with Parker Harris, Salesforce.com co-founder to get a bead on what this means to the the company.
When you strip away all the hype and attempts at nuancing this announcement it's all a bit prosaic but with a clear intent. Mr Harris says that internally to Salesforce.com, this is about its own people finding a way to connect to the outside world of developers: "We've got this Force.com platform and that's amazing but what we really need is to make a connection to the vast number of developers who don't know who we are or what we do." That makes sense given growth over the last few years at SAP's Community Network. But that's very much a first step.
Database.com doesn't let you get access to 'The Database' underpinning Salesforce.com. That would be a step into territory almost certain to break the multi tenancy upon which the company has built so much of its business model. Instead, it allows access to a meta layer that leads you to think you are interacting with the database. The difference is subtle but important in that Salesforce.com will allow you to create tables with rows and columns - sort of - that are in fact objects which inherit stuff like Salesforce.com's security and inherent search. Is that a database or not? More to the point is it a new database or simply a rehashing of something that's 11 years old? Geeks will decide the answer to that question. So is it a big deal or what?
I suggested to Mr Harris that from what I saw, Database.com will make an ideal proving ground for situational applications of the kind Phil describes but only in the context of Salesforce. By that I mean the kind of app that has a temporary use case and needs lashing up over days, doesn't need to be bullet proof in the first instance and which can be easily embedded into Salesforce for testing and limited runtime use. He got that: "I'm sure that's where we're likely to see a lot of early adoption. Make a Facebook app, deploy, run but you know that can quickly become the place where people start to build out business critical apps." That's essentially Phil's argument.
Salesforce.com has been careful to make sure that developers don't see this as a way to build high volume transaction based systems. Database.com won't be capable of doing that. At least not for the next few iterations. Even so, the fact Database.com is being touted as open (build in whatever you like, run on whatever device you like) doesn't get past the fact this is another form of vendor lock on an as yet untested (and unavailable) database platform.
Later, I spoke with Anshu Sharma who is the operational brains behind the marketing of this initiative. Anshu is no slouch. I asked him about the community aspects of Database.com. He said: "We don't think there should be any difference whether a person is asking a support style question, posting a query or offering a code sample. We should be able to use Chatter as a way of figuring out which is which and allowing our people access to the community to deal with whatever is occurring in the stream. We've got the scale to do that but we want to scale much more. Community should help us all." An excellent point and one that is slightly differentiated from SAP's approach but along similar lines.
So...is this a big deal? That's hard to tell. A clue comes from Mr Harris's last words with me: "Last year when we launched Chatter, plenty of people kinda looked at it and shrugged. Adoption has been phenomenal [I can attest to that] and it's going in directions we didn't think at that time. Who's to say about Database.com? We use Dreamforce to throw out these ideas and then see what happens."
That seemed a fitting point at which to end our conversation. And then what do you know? Salesforce goes and acquires Heroku. That gives it an instant leg up in the community stakes with a claimed 1 million developers using the Ruby platform and 105,000 Heroku built applications. Or put another way, if you're not sure you can build a community then why not buy into one and especially one that ticks all the marketing buzz word compliant cool companies like Twitter, Groupon and 37 Signals? Add in the enterprise pull of companies like Best Buy as Heroku users and you can see where this goes. Larry Dignan certainly see it:
The game plan here is pretty clear. Salesforce.com wants to ramp its platform-as-a-service efforts. Salesforce.com expects that it will combine Heroku with VMforce, an enterprise Java platform, to offer a broad platform.
Fast forward a year to Dreamforce 2011. Salesforce.com lines up developers who do a kind of Demo Jam where they showcase some of the stuff they've built and the direction they are taking Salesforce.com. If that happens then everyone else better watch out.
Despite my stated reservations, Salesforce.com has an uncanny knack of building estraordinary momentum behind what it does, orchestrated by Marc Benioff, the PT Barnum of the enterprise apps world. That just got ramped a notch. Provided Salesforce.com finds the right way to set up valuable two way communication with developers, preserves the Heroku community AND delivers what enterprise developers really want, I see no reason to think Database.com will be any different. And if you think that's nutty then check this: during a conversation about Mr Benioff I said he's on the cusp of becoming an adjective as in: "doing a Benioff" Anyone disagree?