Sramana Mitra: Now, I’m going to switch my question to Pure Storage and the strategy for a company like Pure Storage. They are trying to move up the stack. Based on what we have discussed for the last half hour, we understand why they acquired Portworx.
If you put the Chief Strategy Officer hat on for Pure Storage, what else should they be acquiring? You were talking about Elasticsearch in this context as an application that shows up in this process. What else should they be bringing in to climb up the stacks?
Murli Thirumale: It’s funny that you mention the Chief Strategy Officer for Pure Storage. Matt Kixmoeller is the Chief Strategy Officer. I have no pretensions about being as brainy as him. He was involved with Pure Storage from the beginning. I think he was employee 11. He helped build that set of disruption and is now leading the second generation of disruption. Kix is a large reason why we are at Pure Storage.
I think the strategy conceptually is simple. It’s hard to implement and execute, because it’s trying to marry hardware and software in the modern stack. The model is simple. Customers are going through digital transformation.
The digital transformation stack involves the traditional IT stack at the bottom. It’s cloudified from a network compute and storage end at the bottom. Pure Storage has already been on that journey for a while trying to provide more of a paper use subscription model for their solutions.
They have taken their arrays and the flash blade and all of that and offered it to a solution they call Pure One as a subscription model. I think that is what people in the storage business are doing. However, when you get further up the stack, then you get into these areas of data services.
What are data services? Data services is essentially taking the stored data and then serving it up to applications in various forms. A huge part of data services is databases. The modern stacks like PostgreSQL, Cassandra, Redis, and MongoDB are open source models. Others include Kafka and RabbitMQ, which are message queues.
Recently, there is also the replacement for Duke, and Big Data is streaming data which is with technologies like Spark. What Pure Storage is doing is bridging the gap from storing data to be able to provide data services with technology like Portworx.
What we do is we swim in our everyday world which is that world of data services. All of our customers are deploying those containerized apps using these data services. Portworx is enabling Pure Storage to say, “Hey, you are using us as your premiere storage platform, but now you want to manage data.”
Now, the stepping stone from storage to data management is the first step and that happens to software like Portworx. What we are going to do next is to provide those in the form of data services because the world is not just moving to a software world; it’s also moving into a services world.
Everything is consumed as a service. All of the data too is going to be consumed as a service. Step one is an awesome cloudified storage that is highly reliable and high performance. Step two is providing data management, which has high availability, backup, and DR following the lifecycle of the application.
The third step is providing data services. Pure Storage is going down that journey and it turns out that the data management and data services steps have much more to do with software than having the underlying hardware. They are complementary to each other.
It’s a very smart strategy because there are other elements that this strategy provides. As you move up the stack, you move into higher-margin software. Our margins are wonderful. Pure Storage, as a hardware-oriented company, has managed itself pretty well surprisingly. They have a high margin for that class of companies.
The third thing is that it’s a subscription model. Most of the consumption of data services and even the consumption of software is very much on an annual subscription basis versus the CapEx perpetual license model.
It’s no surprise that the Wall Street and market rewards subscription services at a much higher multiple on the same revenue. Typically, SaaS companies would be 10x revenue. If you look at mixed software SaaS companies, it would be 6x revenue. Hardware can range from 1x to 3x in revenue.
I think this model is not just to transform themselves in the marketplace and offer a complete digital transformation like an enterprise experience, but also transform themselves as a company to being a much more mixed mode of subscription software and hardware. This is something that every company in the IT landscape is trying to do.