Mary Hayes Weier of InformationWeek poses some great questions about Oracle’s desired switch (according to a recent announcement) to subscription-based pricing, in her article “Oracle Mulls Shift To Subscription-Based Pricing”, she asks: “How is Oracle going to do subscription-based pricing, and is that going to require Oracle to invest heavily in a multi-tenant architecture?

Let’s tackle the first part of the question. It is, indeed, a huge shift, moving to a subscription based pricing model. Oracle probably has a twofold objective – to protect the current customer base (probably a bit easier because these customers have already spent millions on software/consultants and are seemingly ‘locked’ in) as well as acquire new midmarket customers that haven’t yet had a taste of Oracle’s software. Can Oracle potentially repackage what customers would have paid into annual license and ongoing maintenance fees into a monthly subscription fee spread over the life of the contract, and protect current revenues as well as get new customers?

I don’t think so. When it comes to midmarket, vendors more grounded in the SaaS world (or the multi-tenancy experts) will surely come with lower priced options and Oracle will have a tough time competing. The answer would have to be lower revenue per customer. Which means that just to produce the same revenue numbers, Oracle will need to acquire more customers. More customers implies more salespeople, more infrastructure, more support, more (you fill in the blanks). In summary, that means incurring higher costs but getting the same revenue. Safra Catz will not like that.

Which brings me to the second part of Mary’s question will Oracle need to invest heavily in a multi-tenant architecture to solve this? Absolutely. Multi-tenancy will be necessary to deliver cloud services and shift to subscription-based pricing, while protecting the bottom line at the same time. Yes, building and delivering that kind of architecture isn’t Oracle’s forte. But their industry and financial stature gives them so many interesting choices that it shouldn’t be that hard for them either. They could go and buy a company (and a platform) tomorrow. They could sell more database licenses to and then use as the platform to build and deliver multi-tenant versions of Oracle apps (Peter Coffee points out in his blog entry What the Cloud Entails? how increasingly complex multi-tenant apps are being built in a faster timeframe on Or they could hire the best cloud talent and build the multi-tenancy infrastructure themselves.

Did I miss any other options that Safra has? I’d love to hear your comments. Oh, and of course, here’s a (free) whitepaper (from yours truly) that may have a useful point or two about the pitfalls and benefits of the subscription-based pricing world:

Whitepaper: The Dos and Don?ts of the transition to Cloud Computing