Tune in or read on to learn how our co-founders Chris Cochran (Chief Executive Officer) and Erik Carlin (Chief Product Officer) met and started ProsperOps, and what they’ve learned on the journey to over $1 billion in AWS compute spend under management, or watch the full video:
Jon Myer: Hi everybody and welcome to the Jon Myer podcast. Chris, Erik, thank you for joining me. Unfortunately, Chris Kuehl is unable to make it today. How did you meet?
Erik Carlin: That’s a great question. So the three co-founders all worked together at a company called Rackspace, and we had worked on OpenStack back in the day. This is like circa 2010 and 2015. We had worked on OpenStack trying to create the open source public cloud, and the company decided to sort of pivot a little bit and diversify, and we ended up moving into managed hyper-scale public clouds.
Chris Cochran, Chris Kuehl, and myself were the three co-founders inside of Rackspace to start that business. So we had the good fortune of getting to work together, starting something new, seeing each other’s skills, and learning to work together. When the opportunity arose years later to start a company, we’d already been through the trenches together through starting something new. There are no two other people I’d rather be working together with than the two Chris’s.
How did you guys come up with this idea?
Chris Cochran: Oh, well, it’s absolutely Chris and Erik. They’re brilliant technologists. And we spent a lot of time back at Rackspace serving customers. Rackspace’s modus operandi was using humans, arming them with great technology and tools, and putting them to work to help customers achieve some outcome.
That’s a great strategy and it did a lot of great work for customers using AWS or many of the different clouds. But ultimately some problems actually can be better solved with algorithms. And the idea that we had when we started ProsperOps was how do you use technology, how do you use algorithms to achieve outcomes that humans couldn’t do as well, but maybe more importantly, unlock them to go do things where they were just best in class and were amazing?
“The idea that we had when we started was, ‘How do you use technology and algorithms to achieve outcomes that humans couldn’t do as well, but maybe more importantly, unlock them to go do things where they are best in class and amazing?”– Chris Cochran, CEO on the origin of ProsperOps
And I think Erik and Chris understood and saw that. That was the company that we started back in 2018, and I have been happy to be in the back seat along for the ride since that time.
Erik, can you tell me the exact moment that you guys were thinking about (and how you thought of) the idea of ProsperOps, and the [problem] that you’re trying to solve?
Erik Carlin: Yeah, it has to do with the idea of a “RoboAdvisor” in the personal investing space. Companies have emerged like Betterment (who are actually a publicly referenceable customer of ours), and Wealthfront, etc., that instead of using human advisors to manage money for customers, they implement algorithms and use computer science to effectively automate the investing strategy.
The impetus for the idea of ProsperOps was “Can you take that same concept of robo-advising and apply it to cloud spend?” How do you have customers answer some questions, configure settings, and rather than the bulk of the work be on the shoulders of the FinOps practitioner, how do you transition that to algorithms and automation and allow that to actually do the work to achieve the savings outcomes?
In fact, in the early days, one of our early tagline ideas was “The RoboAdvisor for the Cloud”. What we found in our testing is that half the people don’t know what a RoboAdvisor is, so we had to actually abandon that tagline and move on to something else. So these are the things you try and learn and iterate on.
Chris, let me ask you: did you think that this would be part of your personal cloud story or FinOps journey?
Chris Cochran: For me individually, I think at the time we started, it was 2018, and FinOps as a discipline was just getting started. What was interesting to me personally was largely getting to work with Erik and Chris on solving some hard problems. We had been in the AWS ecosystem long enough to know that people needed help, and that as good as AWS is (and it’s amazing) it was a lot even for really good people to be able to get everything done. And so when Chris and Erik understood that there were ways technology could go and accelerate humans and unlock them to do more things, I just saw it as a very interesting problem.
And over time, I think as we were growing up, the FinOps discipline was maturing. It’s been a fun ride! What I really like is it’s great to play in a space where there’s a community that’s formed, where people see FinOps as sort of their vocation and a discipline that they want to become experts in, and the company’s been able to grow during that entire pattern. That’s been cool to watch individually, but then also with the community at large.
Erik, let me ask you the same questions. Back in 2018, when you started ProsperOps, did you envision that this would be part of your FinOps journey, that it was ultimately going to play a huge part as it does now?
Erik Carlin: Like Chris said, when we started, FinOps wasn’t even a concept really. There was this idea of cost management. We actually named the company ProsperOps because we had this idea of FinOps and we started calling it that even though it wasn’t nearly the term that is used today, and DevOps, this idea of automation and ProsperOps, we saw being the bridge between DevOps and FinOps. So that’s where the name came from. But when you start a company, you have an idea, but you have no idea if it’s going to work, and if people are going to buy your product. You do it because you see it in your mind’s eye and you believe it’s valuable for customers.
Conceptually, we wouldn’t have started the company if we didn’t know it was going to be successful and had a chance. But when you actually hit milestones and realize certain goals, it’s marvelous because as you reach milestones along the way, it’s hard to see the next milestone. You don’t know how you’re going to get there, and then you hit it and you’re amazed. It’s just been a great journey and I’m very thankful.
Erik, let me ask you the billion dollar question. What is the differentiating factor of ProsperOps versus the others that are out there?
Erik Carlin: Well, I think there are two answers to that question. I think one is, you’ve got traditional cost optimization tools, which are visibility-based. We refer to these as the “v1 tools” (part of cloud management platforms), and they’re very helpful in terms of giving you tens, dozens, even hundreds of reports to help give you visibility into your spending, but they don’t actually automate or execute anything on your behalf to achieve savings outcomes.
They might give you some recommendations, but they don’t actually do work and deliver a savings outcome. What makes us different from a lot of other cost management tools is that when you actually subscribe to ProsperOps, we are taking that work off your plate, and our software is executing on your behalf to deliver savings outcomes.
“Traditional cost optimization tools give you visibility into your spending, but they don’t actually automate or execute anything on your behalf to achieve savings outcomes. ProsperOps takes that work off your plate and executes on your behalf to deliver savings outcomes.”– Erik Carlin, CPO of ProsperOps
FinOps is such a large domain. There’s so much work to do. We can take a piece of that off your plate and use algorithms to achieve outcomes better than any human could achieve, freeing you up to focus on other parts of the FinOps problem space. That’s one of the big things that makes this different.
For other competitors in the automation space, unfortunately, there’s a lot of what I call “AI washing” where, because AI is such a hot topic, everyone’s talking about “AI, AI, AI…”
As we win customers away from our competitors we see the results of what their platforms have done. We can see lots of signs of a lack of automation. Computers do things in a way that’s incredibly repeatable. And so, unfortunately, there are a lot of promises being made about AI and automation that behind the scenes are just not reality.
The fact that we are a true AI automation platform is another differentiator.
Chris, did you ever think back in 2018 that you would end up where you are now, relatively, so quickly?
Chris Cochran: No (chuckles). Look, I’ve spent a large part of my career in large companies, but I’ve been working my way “down” into smaller and smaller entities. But when you finally get the chance to start, everything in front of you is a blank page. It’s exciting, but it’s also pretty scary. Along the way you just start one step at a time.
Not to be too cliche, but Erik and I were just together earlier today, and I made a comment when we started off and got to our first million in AR [annual revenue], “There’s no way we’re going to get to five. I can’t see it.” And then you get to five and you go, “There’s no way we’re going to get to ten.” And then you do it and just keep going.
And I think about it as iterative along the way, but it’s been a joy. It’s got its ups and downs as you can imagine, but no, you start and you ask, “Do I know that there’s a problem to be solved causing customers pain? And am I teaming up with folks that I think are the best in the world?” I think Erik and Chris are some of the best folks in the world at what they do. I felt like I was handed a deck full of face cards, but obviously every little bit, you can’t see it until you hit it and then you go to the next milestone.
Erik, same question for you. Back in 2018, did you envision that you would be where you’re at in 2023 at $1 billion in compute spend under management?
Erik Carlin: It’s hard to believe now much less back in 2018. As Chris said, you have an idea. I think I knew that this was going to be valuable because we saw the pain and we saw the results. What we were doing back at Rackspace was incredible. We had some of the best people in the world, we had some of the best technology at the time, third-party things that we had built, etc.
But at the end of the day, we felt like there was a way to rethink it and achieve something, a step function change that was different from anything that existed. And I just knew at the time that if we could get it right, it was going to be extremely valuable.
But you just don’t know what that can look like, so the idea of a billion dollars of compute under management is still hard for me to get my head around even today.
Chris, let me ask you the question, and this is to give an understanding to the audience: “$1 billion in AWS compute under management.” Really, what does that mean? From the sound of it, it sounds like an astronomical amount that I don’t think I could ever see or visualize.
Chris Cochran: The way we think about it is, on one end we provide service to our customers who are trying to figure out ways to get their business moving forward, faster, doing more, etc. I think about the billion dollars as more of a proxy for the opportunity that’s out there. And there’s more work to do.
AWS is climbing towards a hundred billion and we know there are more needs out there. For us, I think it’s great! We’re going to celebrate it, but at the end of the day, there’s more work to be done and there’s more people that need help out there, and we want to get to as many of those folks as possible.
Erik, you were talking earlier about not seeing $1 million in savings for customers. Now you’re at $1 billion and the achievement there. What does that mean to you?
Erik Carlin: Again, it’s even hard to wrap your head around. A billion is one with nine zeros behind it. You don’t achieve that unless you have customers that trust you, and that you’ve been able to consistently deliver value to that remain customers. And so I’m just thankful to our customers for trusting us, and to some of our early customers, for giving us a chance. I can still remember our very first customer that said, “Let’s do it.” And I was almost like, “Really? Somebody actually wants to pay us money?” (laughs).
And so you have faith. You do all this work. You build it. You believe it’s going to be valuable, and when people actually subscribe, and you can deliver value, and continue to deliver value? I’m just thankful to our customers for trusting us with their cloud spend so that we can help optimize it for them.
Chris, let me ask you the question: managing $1 billion, you’ve collectively saved hundreds of AWS organizations over $550 million. What is ProsperOps doing with that $1 billion to generate massive savings for customers?
Chris Cochran: Well, there are lots of different types of ways to optimize the cloud, and to be candid, the companies that are just crushing it are dividing and conquering and doing “all the things.”
Many companies are doing usage optimization, and finding ways to use the right things. We’re helping them deliver rate optimization so that for whatever they use, they pay the best price. And many companies are engaging the cloud providers directly and generating enterprise discount programs or private pricing agreements that provide additional savings. That’s a lot of work to do.
The companies that are most successful in using the cloud are doing all three concurrently, and our job is to focus today on that middle one-rate optimization. We use technology to basically take it off their plate, allowing them to focus on usage optimization (or any of the contract work they have to do directly with AWS), and we’ll handle the rate optimization piece.
And no matter what the customer does, no matter if they’re spinning things up or down, our service is just passively watching behind the scenes. And as things change, we’re adapting their commitment portfolio to generate the best amount of discounts that are possible at that moment.
If something changes five minutes later, we’ll adapt and generate savings in that new environment!
That’s really how we think about it. When the customers are doing great is when we are doing our part to free them up to solve some other parts of the optimization problem-set that they might not get to otherwise.
“It’s not that humans aren’t smart, it’s just that the cloud is so complex and so dynamic, it really necessitates a system that is watching in real time.”– Erik Carlin, CPO of ProsperOps
Nice! Erik, let’s talk a little bit about ProsperOps and your capabilities. What is the capability of ProsperOps? You handle Reserved Instances (RIs) and Savings Plans (SPs). What are your capabilities to help customers out?
Erik Carlin: Yeah, I mean you pretty much just nailed it there.
In the world of AWS, there is this world of usage optimization, which might be things like right-sizing. It’s basically using less “stuff”, and then the world of rate optimization, which is making sure you’re paying the best price possible for the things you are using. We’re focused on the rate optimization space of RIs and Savings Plans. And that’s a complex world, particularly when you’ve got a very dynamic environment as the cloud is increasingly becoming more and more dynamic.
This is what makes it difficult for a human with a report and a recommendation tool to actually achieve a world-class outcome. It’s not that humans aren’t smart, it’s just that the cloud is so complex and so dynamic, it really necessitates a system that is watching in real-time.
So we ingest real-time telemetry from our customers. We’re continually checkpointing that and we’re basically building a picture of what their environment looks like. And then we look at the RI and Savings Plan portfolio and we say, “Can this be adapted at this exact moment in time to optimize cost?” And if so, it’s adapted. And we do that over and over and over again, 24 hours a day, 365 days a year.
If an auto-scaling group changes in the middle of the night at 2 am our platform will adapt while our customers can happily and soundly sleep at night.
So they can focus on what matters to them.
Erik, talk to me really quickly about ESR (Effective Savings Rate), because I know you’re really passionate about it. There’s a website link that I’ll put in the description below, but I want to educate our audience on ESR.
Erik Carlin: When we started ProsperOps, we said we were going to convince people that our platform can do a better job. We asked ourselves, “How do we make this a quantitative discussion with data versus a qualitative discussion?”
And at the time, there was no metric that basically allowed you to measure the effectiveness of your rate optimization efforts. So a lot of the clouds will use metrics like utilization and coverage as a way to describe how much of your spend is covered? Or how much of your commitments are actually generating discounts? But really at the end of the day, what people care about is, “How much did I save?”
Nothing existed, and we said we have to have a way, an objective way that’s not gameable, that allows you to measure an apples-to-apples comparison of which option will generate the highest effective savings rate, and a lot of times sales discussions with vendors can be subjective. We measure a net of our savings share charge and take a percentage of savings.
And the wonderful thing about FinOps is that it’s quantifiable. And our sales process is largely us running a savings analysis for customers, overlaying our algorithms on top of their data, and presenting them with data that says, “If you subscribed, you’re doing a great job today, but ProsperOps can achieve something where the net of our charge has a higher effective savings rate (ESR).”
You can think of it as the ROI of cloud savings. Just like in the investing world, you might measure two different options and say which one delivers the higher ROI, you can think of ESR as the ROI of cloud savings.
Chris, I’m going to let you jump in and comment on ESR, because the value that I see of it doesn’t come from a ProsperOps perspective; it comes from education and knowledge. You’re saying “Hey, this is how FinOps metrics should be valued.” What are your thoughts?
Chris Cochran: Yeah, look, it’s just math. So the definition, the equation, it’s not ours, it’s just, as Erik said, “It’s not gameable numbers out of the AWS system of record where we’re just calculating an ROI.”
And earlier you had asked Erik about differentiation. I think one of the things that we hang our hat on is that we focus on our customer’s ESR as the core measure for what we’re doing because at the end of the day, that’s what they care about.
If we can do things that consistently raise their ESR (net of our charge), and we can compare that to the ESR that they were generating while they were doing it on their own, then you can see that there are actually quantifiable dollars.
This means the customer not only gets their time back, but the incremental dollars that are saved are dollars that they can reinvest in other projects that they otherwise wouldn’t be able to fund.
Everything we do is oriented around “How do we get that ESR as high as possible, and how do we keep moving it up to the right over time?”
Chris, how do you educate customers on the ESR? Say I’m a new customer coming in. I never even thought of ESR or its value. Is that done within ProsperOps? Do you do a little education of “Here’s what we’re able to do and achieve, and here’s your ESR,” and the value of understanding it?
Chris Cochran: Yep, we do it a couple of different ways. Number one, other players and vendors in the space have adopted ESR, which we think is outstanding. The FinOps Foundation has been involved with it, and we want to continue to put it out there. Once again, it’s not ours, it’s just math, but it ultimately tells you as a customer all the things you need to know.
When we start our sales process, we obviously want to make sure that we’re grounded in data. We start and talk about the ESR concept, and then we say, “Let’s move it out of the conceptual into the real.” And the way we do that is we’ll generate a savings analysis on your behalf and it’ll take literally a couple of hours and we can review it with you. And as part of that, you can learn the concept and see how it translates into actual numbers to really drive it home.
And that’s all part of the journey that we go on with our customers. And our customers are awesome. They’re all generally sophisticated FinOps practitioners. So they come to the discussion with a lot of knowledge. And what we try to do is show them, if they’re not familiar with ESR, how they can take their existing heuristics, tweak them a bit, and gain a whole new understanding of what performance could be and should be so that we could get them motivated about saying, let’s go get it. Let’s go make it happen. And then we could focus on other items that take our company forward.
Erik, Chris mentioned that the FinOps foundation has adopted the ESR. How’s it make you feel when others are adopting something that you’ve put a lot of thought into and educating people on ESR and what it means?
Erik Carlin: So they are in the process of embracing effective savings rate. In fact, just a small plug for anyone who’s going to be at FinOps X, which is the FinOps Foundation conference, I’m actually giving a talk on effective savings rate. There’s a working group that is in the process of embracing ESR and would love for anybody who’s interested to join.
Chris said this before, it’s just math. ProsperOps created effective savings rate, but it’s just math. And in my view, every organization, whether you’re a ProsperOps customer or not should know your effective savings rate. Just like anybody who’s investing money, you should know what your return is. Otherwise, you have no way to quantify the effectiveness of your efforts. And so it’s an incredibly beneficial metric for the industry at large.
It’s a completely cloud-agnostic metric. And so while we today are supporting AWS, there’s nothing unique to AWS about this metric. You can measure ESR on Google, you can measure ESR on Azure. And so we think this is a metric that is useful in the sense of assessing your relative savings performance, which for us is important as an automated service in making the case. But apart from that, it’s just useful for the industry at large and for folks to understand what their effective savings rate is.
And the cool thing too is it’s benchmarkable, and so part of us donating this to the FinOps foundation is also submitting all of our benchmarking data so you can know if you’re achieving a 30% effective savings rate. Well, is that an A plus? Is that a B minus? Is that a D? How do you know?
Effective savings rate allows you to not only understand your savings rate but then say, “Okay, where do I sit relative to peers in the industry? And how do I continue to push that higher if there’s more to be had?”
Erik, can you share some of the features that you’re working on or that you just released?
Erik Carlin: Well, I can’t reveal what we’re working on, but I can tell you some of the recent features that we’ve launched. To give you a couple of examples of some recent launches, one feature we recently launched is called Advanced Cyclical Optimization Support. I mentioned earlier that the cloud is just more and more dynamic. And generally what we find is that customers who have cyclical usage (apart from an automated solution that can continually evaluate and adapt) end up covering at the trough of usage, and the actual optimal savings point is higher than the trough. It takes a lot of complex math and forecasting, etc., to figure out that point and continually maintain it.
This is an interesting feature that we launched completely free for our existing customers, at no extra charge, but it increases their effective savings rate because it finds the more optimal coverage point.
And that’s just an example of us working to continually drive effective savings rate higher and higher, and creates that gap that exists between “DIY approaches” versus what a sophisticated algorithmic automated solution can achieve.
One other feature I’ll mention is something we call Dynamic Prepay Amortization Tracking, or DPAT for short. Many of our customers are prepaying their commitments, where if you prepay, Amazon gives you an incremental discount. This is also another way to drive effective savings rates higher. But in a very dynamic world of spend where services are changing, and commitments are moving around, it becomes very difficult for finance and accounting teams to properly amortize their prepayments. And so this became a problem that no other solution on the market is really solving. And so our customer said, “This is something that would be valuable if we had a solution to.”
So we built a capability in our platform that allows our customer’s accounting and finance teams to simply log in at the end of the month. We track all of that and we show them how their prepayment is amortized over time. And we have multiple customers today that use that to close their monthly books.
That’s just a taste of some of the recent features that we’ve launched for our customers.
I want to wrap things up and talk about your channel partner program. It sounds like there’s a massive opportunity for cloud partners. What can technology partners, resellers, MSPs, and consultants do with ProsperOps?
Chris Cochran: Yeah, we serve many AWS distributors, resellers, managed service providers, and even consultants. And we love to partner. We find in general that a lot of those service providers end up solving a very broad set of problems for our customers. And we’re always very complimentary because we slide right into that automated rate optimization spot that compliments all the other things that they do for customers.
But one of the things that we learned back in our prior life was that when you’re a service provider and you’re delivering your value to end customers, in many cases humans can be the hardest thing to scale in the equation, even if you’re really good at hiring, even if you’re really good at training, even really if you’re arming them with tools. They’re always hard to scale and you’re always trying to deliver a very consistent level of quality.
And one of the benefits of an automated service like ours being used by MSP, etc., is that they get this automated delivery. It’s infinitely scalable and delivers a very consistent optimization outcome to customers. And it allows our partners to focus on other things that they do very well. So we love to work with folks and we can add a lot of value together for our customers.
Erik, I’m thinking startups always have an interesting story to tell. Do you have one that you could share with us?
Erik Carlin: Yeah, there’s one that always comes to mind, and I think you’re right, startups all tend to have just a story. And for me, it was our very first customer and I was just amazed that someone was going to pay us. But a little bit later, we landed our first big customer and in the early days of startups, you’re building the plane in real-time as you’re flying. And so our first big customer hit some edge cases and pushed us in ways that we weren’t completely ready for. And I remember that it happened when I was visiting my brother who lives in Central Virginia and didn’t have great internet at the time. I remember being up, working late and he didn’t have good internet (laughs) so I had to go to McDonald’s in Lynchburg, Virginia working at McDonald’s on their Wi-Fi until they closed at 11 pm, and they’re mopping the floors and I’m still working. And finally, they kicked me out at midnight.
I went back to my brother’s house to his basement and tethered my laptop to my phone. I had one bar, and I remember working until, I think it was four o’clock in the morning, my brother has five cats that all sleep in the basement. So cats are walking on my shoulders and walking on my laptop, and I’m working in the basement with these five cats around me, tethered to my phone in Central Virginia with one bar, all so we could get this customer online and start saving them money.
Every startup’s going to have some moment like that where it just sticks in your mind as part of the journey.
True customer obsession, and the journey to it. So Erik, I want to end it with you, and I want to know what’s next for ProsperOps? Where are you guys going? What events are happening?
Erik Carlin: The way that I think about what’s next for ProsperOps is really taking a step back and thinking about the problem that exists today. So if you look at the three big public clouds, AWS is about an 80 billion run rate. Azure is about 60 [billion], Google’s about 30 [billion]. So altogether that’s about $170 billion a year growing to half a trillion to a trillion dollars. And it’s estimated that about 30% of all cloud spend is waste. That is, there’s inefficiency that exists somewhere in customers’ use of the cloud, and of course that’s not intentional. But in the course of all of the agility and all of the power that exists, there is inefficiency. And traditional solutions are not effective at removing that waste; if it was, the waste wouldn’t exist. And so to really remove that waste requires automation and complex algorithms.
And our mission is really one of public cloud waste removal. We’ve attacked a specific problem set today that we think is the biggest lever that you can pull right now in removing waste, but there’s a lot more to do in this space. If I think about what’s next for ProsperOps, it’s really about continuing to innovate in ways that our platform will remove more and more inefficiency, and more and more waste so that they can keep more money and use that to invest in further projects or whatever they want to do with those funds.
Chris and Erik, thank you so much for joining the show.
Chris Cochran: Jon, it’s a pleasure to be here. It’s good to see you again.
Erik Carlin: Still looking forward to getting some barbecue in Austin!