Clicky

The Cost of Cloud
by Daniel Newman | June 21, 2021

The Six Five team talk about the study that Andreessen Horowitz did called The Cost of Cloud.

Watch the clip below:

If you are interested in watching the full episode you can check it out here.

Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this podcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we do not ask that you treat us as such.

Transcript:

Patrick Moorhead: I want to bounce to the next topic. And this was a study that Andreessen Horowitz did called The Cost of Cloud. And let me be really clear. I’m not a cloud denier. Back in 2005, when I was with AMD, I was the primary supplier for Google Cloud. I absolutely know the capabilities. Daniel and I both do a lot of business in the public cloud, and we do a lot of business in the private cloud and on-prem companies. So it was funny. This was the first article from somebody who I think has made a ton of money off the cloud saying, you really need to think about the cost of the cloud.

And what they did is they looked at some of the top companies out here, the Palentir, Slacks, Snowflake and Datadog, and they showed on average, 50% of their cogs are with IS providers. So, and then they talked about what Dropbox did, and how they repatriated that, and how they made $50 million of savings by repatriating a hundred million dollars. They went on to say that, not that this was a bad thing. They gave the public cloud, right, all its due credit and respect. And then they came out with some recommendations on, hey, how do you manage that? It was like, start in the cloud, but don’t end in the cloud. And they said as a KPI, incentivize the right behaviors, optimize. Think about repatriation, incrementally repatriate your workloads.

I thought it was a very intelligent write up, and I think it was what we all knew, is that if you have a fully utilized on-prem infrastructure is less expensive than going into the public cloud. It’s not as quick. It’s not as agile. It doesn’t have as many access to services, but by the way, it gets really interesting when you see the hybrid solutions from Dell Tech, HPE, and oh, by the way, even AWS and Azure and Google Cloud with Anthos. So super smart type of analysis. The first of its kind I’ve ever seen from somebody who made a ton of money in the public cloud, and SAS companies just kind of laying it out there. Daniel, what do you think?

Daniel Newman: Yeah, it was a really interesting article, and there was some really good data to extract, Pat. You mentioned a few of the companies, Palantir, Slack, Snowflake, Datadog, and Asana. 38 to 63% of their cost of revenue was cloud spent. 63% at Asana. I mean, it’s remarkable. Now, I think what we all knew and have all come to the conclusion on, is that when you’re starting a company, the cloud makes a ton of sense. You swipe a credit card, you spin up a few workloads. The question mark becomes.

Patrick Moorhead: Capbacks.

Daniel Newman: Economy’s a scale, right? The economy’s a scale start to become a question, and back to Economics 101, you maybe have heard the term elasticity. But the elasticity is as the economics of your business grow, as you’re getting more revenue in, how elastic is the cloud? Now we use elastic in terms of its access to resources. It’s very elastic in that way. But when it comes to cost, not always so much. And that basically, Pat, is why…

Forget startups. Big companies have largely acknowledged hybrid and multi architectures will be the way forward. You know, we’ve seen now the public cloud workloads rise from somewhere in the twenties to maybe even the low thirties percentage wise of all workloads at the Enterprise level, but that’s still, well over half are still deployed on-prem. Why? Well, it’s a couple of reasons. One is, not everything’s easily moved to the cloud, so we hear a lot about making things cloud native, and also Pat, not everything makes the most economic sense to go into the cloud. And as we build hybrid architectures, hybrid control planes, you talked about a bunch of them, got things like IBM satellites. You’ve got HPE GreenLake. And of course, AWS Outposts has your arc, it comes from both ends.

But in the end, Pat, what you’re actually seeing is a gravity being created where public is gravitating towards hybrid and prem. And then the prem based are gravitating towards public. But the recognition is basically, we’re going to end up meeting and settling somewhere in the middle. In the end, you don’t buy technology to solve technology problems. You buy technology to solve business problems. What Andreesen Horowitz does very well, Pat, is understands how to make money. There’s very few firms in the world that can take money and turn it into more money than these venture capital companies. And what they’re basically saying here is for the companies that are investing in, for companies that are making these kinds of investments, that it isn’t a one way street, everything isn’t going to go public. And if you’re doing the business correctly, you’re going to analyze this, determine it, and make decisions to ideally go hybrid. And that’s where the future is going to lie. And I don’t see that changing anytime soon, especially as data continues to grow exponentially.

Patrick Moorhead: Yeah, Daniel, I like to use the analogy of a candle being burnt on both ends. You have the cloud native folks who started obviously in the cloud, and then slowly moved over to support hybrid. And then you had the classical on-prem folks moving to a hybrid as a service type of opportunity. And I was thinking about the Dell APEX launch and thinking, okay, they’re going to give the option to put hardware inside of Equinix. If you could do that, and Dell would invest in CapEx and put hardware there before people need it, and then put a multi-tenant solution in there, you essentially have the public cloud, and everything you would need. So I think this is great for the market. I think this is great for enterprises. Gives them multiple opportunities and multiple opportunities for people to make money.

And I was wondering why would Andreessen Horowitz have done this? And I think you probably hit it. This was kind of a message to its own investment companies that says, start in the cloud, but as you hit a certain point, you should think about repatriating some of your workloads. And I was wondering, gosh, are they going to make a play for some of the on-prem folks who are trading well below some of these cloud titans? And then the final thing was all kind of like, okay, I get this. AWS 31% operating income, 31%. Compare that to HPE’s, or Dell Tech’s ISG bank. And AWS is getting paid for the value that they’re providing, but interesting.

Daniel Newman: Well they all have better than Google Cloud, but. All right, I couldn’t help myself. No, that was fair game. But that’s-

Patrick Moorhead: Daniel-

Daniel Newman: … investment in growth. That’s investment in growth. Won’t be too harsh on it, but just saying.

Patrick Moorhead: Yeah, Google Cloud, reach out to Daniel Newman. Brief that guy.

About the Author

Daniel Newman is the Principal Analyst of Futurum Research and the CEO of Broadsuite Media Group. Living his life at the intersection of people and technology, Daniel works with the world’s largest technology brands exploring Digital Transformation and how it is influencing the enterprise. Read Full Bio