OVUM COMMENTS

Ovum meets Marc Benioff, CEO of Salesforce.com

Salesforce.com is one of the very few successful ASPs. It offers CRM functionality over the Internet, priced at $65 to $125 per user per month. It made its first sales in 2000, and expects to reach $90-100 million in its current financial year. This week, we met founder and CEO Marc Benioff. The first question was obvious ...

Why is Salesforce.com one of the few successful ASPs?

First and foremost we're functionally competitive product against other CRM products. So this is nobody's stepchild. I think that this product would be successful if it were a conventional software product. But I don't think we would have the revenue growth.

There are other products which have the same features and functions. But you've got to install it, you've got to buy the Oracle or the DB2, the computers, the application servers, the security, the load balancers, the SSL accelerators, and so on. Today, deploying software is worse than it's ever been.
The power of ASP is that we've made all the choices for you. We don't sell you tyres, seats, radio, engine, a hood, taillights - we sell you a car.

Why has no-one else succeeded?

Two reasons. The first reason is that this is really complicated. You're dealing with a radically new technology model. You have to be ready to deliver a very complex back-end architecture. We have an architecture just like Amazon.com. It had to be written from the ground up to be multi-tenanted [ie support multiple customers on the same platform]. And we effectively have a $10 million data centre. You also have to deliver the reliability, and the performance. We'll deliver 80 million pages this month.

The second thing is that the business model is non-trivial. When I was at Oracle, we had one slide with a growth curve. But it doesn't look like that any more. Larry Ellison recently gave a whole dissertation to the financial community on how to value Oracle based on its maintenance fees. But how do you value a company if their new licence revenue is falling? It's not a good sign.

We recognise our revenues only on delivery - like Ovum in fact. And then we have a big balance of deferred revenue. But you don't meet too many software companies that have the same revenue recognition model that you or I have.

This quarter [to end April], we will generate about $19 million of revenue. We'll generate about $200-300,000 of profit - not very interesting. But we'll generate over $2 million of cash. Why? Because we have $49 million of deferred revenue, and most deals are billed a year or two in advance. This is critical - because a lot of our expenses are incurred before we get any revenue.

This fooled a lot of the ASP entrepreneurs and their venture capital investors. $1-2 billion of investment has evaporated on this model.

Who would be your nightmare competitor?

I think Microsoft. But fortunately, with their CRM product, they have 'built Sales Logix' [ie not an ASP product]. But we think Siebel is working on an ASP offering.

What about PeopleSoft?

They have best potential for executing against SAP than any other company in the world. They have the best product, happy customers, good brand - but they don't seem to be able to execute. Their financial results this week validate that. PeopleSoft's CRM product is pretty good - but no-one's buying it. They should have bought Siebel.

What about the future?

We see Salesforce.com's future as a platform, rather than just an application. It's open. Using the APIs in Salesforce.com, you can address each application service discretely to bring data out. So you can use our platform to build your own applications.

Presumably Web Services will be important in this?

Let's look at an example - Segway. They make these little scooters which are self-balancing with gyroscopes on. And you can only buy them on Amazon.com. Now, Amazon is also an application which runs on the Net with an XML wrapper. When you put your order into Amazon, it sends it to us using an XML RPC. We provide the CRM, and we integrate with Oracle Financials run by Segway, also using an XML RPC. So clearly it's a best-of-breed system.

Let's take a different example - Microsoft's Office 2003. You can modify Word and Excel on the fly and make Web Services calls and bring back the data. That's the power of Web Services, and Microsoft really gets that. So if you have a store on Amazon.com, it will be easy to build an Office 2003 spreadsheet that shows you the performance of your store.

Now compare Office 2003, fully rewritten, fully integrated and open, with monolithic applications like SAP, Oracle Applications and PeopleSoft. They are not modern architectures, they haven't been rewritten, and they are not open. There aren't independent developers who build on these packages. They just have users at their conferences, who go from room to room like rats in a maze. And then they get presented with new functions.

Web Services gives you the ability to build applications using standard APIs more than ever before. This is the next generation of systems architecture that I think CIOs are going to demand.