I hear that vendors are bundling cloud services with their other software licensing deals, and I have some thoughts about why. Azure credits are being bundled into Microsoft software license deals. Oracle customers can buy cloud credits as a way of avoiding problems that stem from database software licensing true-ups. There are a couple of ways of looking at such practices. One is that these credits are a great way of getting customers hooked on your cloud. Oops, I meant to say a great way of helping customers learn the value of your cloud. The less positive perspective is that the largely unused credits inflate the cloud services’ revenue without customers actually using the cloud. Naturally, the reality is more complex. I suspect that these are the primary reasons for bundling cloud services into license deals.
This bundling of cloud service credits does seem to be a common tactic for cloud providers with other business lines. Oracle, Microsoft, and VMware have enterprise licensing agreements with most large organizations. The enterprise customers want to pay as little as possible to get the products they use. There is no problem for the enterprise if the bundle covers more products than they use; most agreements do cover fixed bundles of products. Within each vendor organization, the revenue from the enterprise bundle is split among the product groups that supply the bundle. Naturally, there is an internal fight over who gets how much of which bundle. Most product groups want to be in as many bundles as possible in order to get the most revenue. There is a natural tendency to stuff the bundles with products that customers are not yet using. If this doesn’t increase the license cost for the customer, then there is no negative impact on the customer.
Having more products in the bundle lowers the customer’s cost to adopt additional products. As an example, a customer that hasn’t yet deployed a DR solution might have DR software included in their bundle. If so, the software license cost is not an impediment to deploying a DR solution. This is the positive reason for including cloud services in bundles: it lowers the barrier to entry. If a new project comes along that suits deployment into the cloud, then why not use the cloud service you already paid for? It is essentially free, since it was included in the bundle. The cloud vendor hopes that over time, you will come to love its cloud and choose to put more into that cloud. Eventually, the cloud credit component of the license bundle becomes valuable to the customer. At the next renewal, it becomes an important part of the deal.
The less positive side is the overinflation of cloud adoption figures. If 10% of the income for every enterprise bundle is allocated to the cloud services, then lots of business are paying lots of money for cloud. Now, we can say our cloud service is doing $X billion. The fact that most of the cloud services that are being paid for are not being used means that such figures are meaningless. They don’t reflect the value being delivered to customers by the cloud service and are really just an accounting trick. This is part of the battle for mindshare and credibility among cloud providers. For a customer, it is far easier to justify deploying an application onto the third largest cloud than the twenty-fifth largest.
Notice that the vendors that are able to do this are ones with established non-cloud businesses. They are leveraging existing relationships to make their cloud business work. There is nothing wrong with this leverage: it is good business sense. Notice also that these businesses are cloud followers rather than initial leaders of cloud business like AWS and Google. Funneling this money into the cloud business is crucial. Building all those data centers and services is expensive. Using this revenue to build out the cloud business is probably more appealing than using debt.
Stuffing cloud service credits into existing license bundles is a fact of life. It is only an option for businesses with pre–cloud product bundles that they can stuff. While the bundling can be beneficial to existing customers, it can also make a cloud look larger than it currently is. Personally, I’m in favor of having cloud choice, so I’m glad that these organizations are trying to build competitors of the AWS juggernaut. If it takes an accounting trick to scrape together the $10 billion that it costs to build a competitive infrastructure, I’m fine with it.