Most of the world relies on Arm technology every day, yet many don’t know the company by name. What started as a joint venture between Apple and a company called VLSI in 1990 to create low-power computer chip architecture, Arm Holdings now sells designs via a license, receiving royalties from products that get shipped using the technology. About 70 percent of the population of the world today is using a product that has an Arm CPU in it.
In conversation with host Jack McCullough, CFO Jason Child shares his journey from his early days at Amazon and Groupon to navigating complex IPOs and why curiosity—paired with AI tools—is the most essential asset for the next generation of financial leaders. Listen by clicking below. The Q&A, lightly trimmed and edited for clarity, follows.
—
Listen to the podcast here
Welcome back, rock stars. We have a fantastic guest, and I have been looking forward to this one for a little bit. Jason Child is the CFO of Arm Holdings. Arm is really interesting because it is one of those companies that is not necessarily a household name, but it is one of the most important companies in the world. I say that with no sense of exaggeration. It is one of the most important semiconductor companies on the planet, even though most consumers have not heard of it. Rather than manufacturing the chips themselves, Arm designs the underlying chip architecture and licenses that intellectual property to companies like Apple, Qualcomm, Nvidia and several others. Jason, welcome to the show.
Thank you for having me.
How did ChatGPT do in describing your company? I just read the words that it gave me.
As usual, pretty good, high-level. Probably not enough detail to do too much with, but I agree.
Lucky we have the CFO here, and I want to let you know what people might be interested in that I did not touch upon.
Maybe to put some numbers behind it, we have been around for about 35 years. We were created as a joint venture, actually with Apple and a company called VLSI, back in 1990, when the goal Apple wanted to create the first battery-powered computer. They needed an extremely low-power chip architecture to be able to do that. That is how the company got started. Fast forward now. We sell designs via a license, and then we get royalties from products that get shipped using our technology. To date, we have received royalties on over 350 billion chips that have been shipped over 35 years.
If you put that in context, it works out to be about 70 percent of the population of the world today is using a product that has an Arm CPU in it. The CPU is the brain of the computer. That is what allocates all the workloads in a chip, even at the front of a GPU. It is what drives now agentic AI. We started as a mobile company where about 100 percent of mobile phones use the Arm CPU. It is now extended into TVs, PCs and cars. Almost 80 percent of every car on the road uses our chips.
Of course, now what everyone is talking about is the cloud, AI and all the hyperscalers are using Arm. We also just announced for the first time in history that we’re actually now building a chip. We’re now selling silicon, and that’s just something we announced about two months ago. It is an exciting time, and the company, as you said, is integral in this digital evolution and transformation that we have seen over the past couple of decades.
It’s just an absolutely fascinating story. Thank you for that. With hindsight, you should have just done the intro directly rather than me trying to do it. We are at a good spot. I want to return to that. I always like for the audience to get to know my guests’ background a little bit. Where did you grow up, Jason?
I grew up in Seattle, Washington.
Are you from a big family?
I’m the oldest of four. I guess nowadays that’s big.
By modern standards, that’s a pretty good-sized family. I always like to ask, what was your first job? I do not mean after college, but maybe when you were in high school or early college.
My first job was scooping ice cream at Baskin-Robbins. I leveraged that into selling tires and batteries at Sears Automotive. That’s actually the job I did through college. Fun fact, I made more money in my last year working 25 hours a week selling tires than I made in my first year at Arthur Anderson as an accountant.
That is really something, the potential of a career in sales right there. What is interesting, the most recent guest we had, the CFO of Wayfair, grew up on the other side of the country in Maine, but her first job was also scooping ice cream. I have noticed that a lot of CFOs grew up with that background, that their first job was restaurant-related in some form or another. Where did you go to school?
I went to the University of Washington.
Did you study accounting?
Yes.
What was it that was appealing? Clearly, it was a great career decision for you. What was it that was appealing to the seventeen-year-old Jason about a career in finance and accounting?
I had maybe an unusual upbringing. My dad was a rock star in the ‘60s, and my mom initially met him as a groupie. Unfortunately, that marriage did not last. My mom then got remarried when I was 11 years old to a partner from PwC. As I went through college, I was going to be an engineer. I graduated in ‘91. At that time, it’s hard to remember, but that was a recessionary time. I switched my major. I actually went into business.
I switched from finance to accounting in the last quarter, because I had read that the best place to get a job is in the Big Six. Even though I did not love accounting in school, I thought it was okay. It was a little boring. It was a great place to get a job. I actually then ended up lucky enough to get a job at Arthur Anderson. At that point, it was the largest accounting and professional services firm in the world. Actually, once I started working, I realized accounting was more interesting in actual practice than it was maybe in textbooks.
It’s boring as heck as a collegiate study. When you actually get into it, it can be interesting. I tell young people that they do not really believe me.
I tell them it is a language of business. I do not know how many MBAs I fired over the years. That basically said, I wish someone had made me do an undergrad in accounting because there are too many classes. You cannot do it as a masters degree program. There are just too many classes to take. I wish someone had forced me to do it when I was younger to have that base knowledge. That is something I do tell folks as well.
I’m intrigued. I somehow did not know that your dad was a rock star. Is he someone I might have heard of?
He played in a bunch of bands in the Northwest in the ‘60s. There was Paul Revere & the Raiders. There was The Kingsmen. A band called Don and the Goodtimes, which was from guys from both of those bands. He recorded a few albums with them and was basically in the whole circuit. Everyone on my dad’s side of the family is an artist. They’re on the creative side. I was the black sheep, and they always felt sorry for me and said, “Even though you’re going to do business, you still might be happy.” I said, “I’ll try.”
There you go. The Kingsmen, was that Louie Louie [song]?
Yes.
That is the best piece of trivia we’ve had on all of our guests so far. Anyway, I would like to chat a little about your career journey. You mentioned that you started with Arthur Anderson, which was one of the original monster firms. One of the great places to work, I never worked there, but my auditors were there in my first job out of public accounting. World-class professionalism, very impressive company. They had an ending that was a little unfortunate. What did you learn from public accounting that maybe is still relevant to you from a leadership perspective and a skills perspective, even at this point in your career?
I found that public accounting was an incredible training ground. One of the things I always tell folks that I advise or have a chance to mentor is getting a chance to work with just really intelligent people in an organization that really invests in people and invests in development and invests in giving folks on-the-job training, education capability, maybe even in the classroom, whatever. Anderson, at the time, was pretty unique in that they had the highest percentage of revenue spent on development.
It was something like seven percent of revenue. They had this St. Charles campus outside of Chicago. I must have gone at least once a year for the seven or eight years I was there. Just that focus on development and growth, and ultimately curiosity to want to be able to grow, was super helpful for me and probably one of the most important aspects of my career.
When I was looking at your career, you worked for a series of groundbreaking companies. It’s fair to say, obviously, Amazon changed business and retail as we know it. We met when you were the CFO of Groupon, and you’ve worked for Splunk and a lot of others. Were you just naturally attracted to that company coming out of public accounting, or is it a conscious decision to follow that path?
I would say I generally, as probably most CFOs, have a natural orientation towards lower risk and am more risk-averse in general. What was interesting is I had been at Anderson for seven and a half years. I spent three or four years as an auditor. I moved into the consulting side, which eventually became Accenture. I actually had really gotten excited about Amazon as a product because I was in charge of like the book of the month for our consulting group book of the month, where you pick a book, and we would all go read it and learn from it and whatnot. That’s how I got into Amazon.
This was in 1998. One of my former clients had gone there and just said, “They’re looking for someone in this role. You should come and talk to this company.” I met with the CFO and met with the whole team. We only had 300 people in Seattle at the time. It was a very small company at that time. Basically thought, “They offered me a job.” I remember everyone telling me, “Why would you leave the world’s largest professional services firm? You could be a partner, and you could do great.” I do not know, I just felt something about this opportunity.
The first week, they sent me to Germany, because that was not part of the job, but why not? Literally in the first month, I realized this job was so much more exciting and interesting than anything I had ever done. I just got addicted to being in this high-growth environment that was trying to build an industry that did not exist. Apparently, that meant I now went from risk-averse to a little more, a little more risk on. I spent almost, I guess, eleven and a half years at Amazon. When I joined, we were a sub-billion-dollar company. By the time I left, we were about a $50 billion company.
Really got a front row seat at building the FP&A team. At one point, I was a controller. I got to do some of the accounting stuff back when pre-SOX, when things were a little more complicated, and that growth and that just the thrill of trying to create an industry, which is both painful and exciting. I do not know, I somehow got used to that. That’s what then launched me into my next opportunity, which was at Groupon, which was at the time, the fastest growing company in history.
It went from zero to a $1 billion in two years. The year I joined, which was year after year two, it went from $1 to $4 billion. In its third year, it went from that to $7 billion in the other year and a half later. That was the, from my perspective, having been in Amazon and Groupon, I guess I just got excited about how the world of, at one point it was internet, then it became e-commerce, then of course it became mobile, and so on. It just became an area that really fascinated me.
You picked some good ones. In fact, one of them, I think you spoke at the MIT event. You were with Jawbone, is that the name of the company?
Yes.
You gave all of the organizers this speaker, and I know Jawbone. I do not recall how exactly they ended, but I got to tell you, I still use that. It was so far ahead of the pack.
They [Jawbone] invented the first wireless headset, which we all used back in the late 2000s and 2010s. They invented the first wireless speaker, and then they invented the first Fitbit-type device. They actually invented that category. They created three categories. Unfortunately, others monetized it way better than they did. It was another company that fit that mold of these guys, they have some of the best designers and some of the best engineers in the world. In that case, great ideas, great concepts, financially had not been managed well by the time I got there, it did not take me more than a couple of months to realize that they had severe challenges. They’re no longer in business anymore.
I am always sorry to hear that. I often work out in my basement. That’s when I blame my 1970s rock and roll on it. Often, I think of you when I put it on. No, that’s great. One question I want to ask about your earlier career, because it’s relevant to a lot of the members and audiences, is, you worked for phenomenal companies along the way, and you must have had some great mentors. I’m wondering if, when you look back, there are a couple that stick out as, “They really made an impact on my career, and maybe they do not even know it.”
There are so many mentors. It’s funny because of the notion of mentoring. There’s formal mentoring and there’s informal mentoring. I would say informally, I have probably had mentors throughout my entire journey that I would sometimes informally ask or somehow create a relationship. I certainly had those at Anderson. I would say at Amazon, I had some formal mentors. One of them was Jeff Wilke, who was the CEO of retail for many years there.
He actually approached me and offered to be my mentor. That was actually really cool. That was probably the only real formalized mentorship I had. I worked very closely with Andy Jassy, who’s now, of course, the CEO, working closely with a guy named Jason Kilar. He’s the guy who left to go start Hulu. Actually, Joy Covey, the first CFO, was fantastic and was an early mentor. Over the past 10 or 12 years, Mary Meeker was on our board at Groupon, and I had known her from my days at Amazon.
When it comes to big-picture internet thinking, there is no one better. I would say John Connors, former Microsoft CFO who was on our board at Splunk, has been. He was actually part of the reason I was connected to Arm. He was a key person in helping identify the opportunity for both sides. Honestly, there are so many mentors along the way, but those are some of the big ones I can think of.
I want to go back to Arm, because it’s inherently fascinating. At the risk of asking a really dumb question, can you tell us about how the opportunity came about and what made you decide to take it? I could probably guess pretty easily why you decided to take it.
You may not know that the story is probably different from what you think. I had been at Splunk for about three and a half years. Now Splunk, of course, has been sold to Cisco. This was disclosed when Cisco approached Splunk. We had an interim CEO, who was a former board member, running the company at the time. I went through a three-month deep diligence process reviewing all the financials that helped figure out what the price could be.
At that time, the board, for whatever reason, just decided they did not want to sell and hired a new CEO. I honestly thought that was not the right approach. We were trading at $110 and this other company offered something like $195. It was pretty large. They eventually sold, and ended up selling for less than that. In hindsight, it probably would have been the right decision at the time. As a result, with a new CEO coming in, when you’re a CFO, you get hired typically by the CEO. I was hired by the CEO, Doug Merritt, who’s a phenomenal CEO.
For a variety of reasons, he ended up moving on. When that happens as a CFO, you’re just naturally in a “Where am I?” Most CEOs want to bring in their CFO. We brought in a CEO, and he was great. I do not know; there was no issue there, but I just felt like we did not pick each other. That was always one thing that at least made me think about things. I would say the biggest thing was that I was actually right at that time. This was June of 2022. I got diagnosed with prostate cancer, which, unfortunately, as you probably know, one in seven men is going to get diagnosed with it at some point. It’s shocking but not too shocking.
Unfortunately, it’s one in seven. I spent a month or two trying to figure out my situation and what I was going to need to do. I got that, and so I started going down that path, and I was going to take medical leave and have surgery. I got a call from a head-hunter saying, “Arm really wants to talk to you.” I just said, “Look, this is not going to work. I have some health issues.” They said, “They still want to talk to you.” I discussed with Rene Haas, who had just been appointed CEO a couple of months earlier. It was one of those instant love connections.
We just completely saw the world the same way. I just flat out said, “Look, I’m about to have surgery, and so this is an awkward discussion.” He was like, “I know a bunch of people who have prostate cancer, you’ll be fine.” I said, “That’s it?” He is like, “Yes, that’s it.” I said, “Okay.” I literally interviewed the rest of the team, and I think I had an interview with Masa from SoftBank, who is the chairman, within 4 or 5 days. I had an offer in like seven or eight days. I thought, for a $40, $50 billion company, that’s not a usual process. It showed the level of conviction that the CEO and the team had in myself.
I, of course, had a lot of conviction. The company is a pretty incredible company. I thought, “This is a real once-in-a-lifetime opportunity.” Anyway, then I went and took a couple of months to deal with my health issues, and I’m happy to say everything’s fine now. It’s been almost four years now. This is my fifth CFO job. What that basically means is that over four CFO jobs, there were probably a bunch of things that I felt like I could have done better, or something did not work the way I had hoped it would. Luckily, now, almost four years in at Arm, it’s probably the best job I’ve had.
It’s the job where all the pieces that I’ve figured out over time fit with the CEO, the product, where it is with the market needs, and does it have the right product at the right time? Timing is everything. Culture of the company. Team dynamic, investor base, chairman board, all that stuff. There are so many ingredients that go into whether or not it’s going to be a really successful CFO role. This has been, every job I’ve had until now had some great pieces and then had some pieces that just weren’t quite right. This one is where it’s all come together. It has been a pretty incredible experience thus far.
Congratulations on surviving cancer as well. I did not previously know that about you. I’m glad that you beat it. You use the phrase once in a lifetime. It seems to me, as I look at your career, you’ve had three once-in-a-lifetime types of things. With Amazon, Groupon and now Arm. It was interesting to me that the timing of when you joined Arm, you were hired several months after Rene. It seems like they get Rene, they get Jason. Clearly, they’re building the leadership team to take the company public. That indeed is what happened. What’s it like to take a company like that through an IPO? Again, not famous with the general public, but a truly critical company to the overall economic infrastructure and therefore highly visible in the markets?
I’ve taken two companies public, and they’re both at very different ends of the spectrum, I would say. The first one was Groupon, and Mary Meeker was on the board with John Doerr, Marc Andreessen and Howard Schultz. We had all these well-known characters that were all brilliant in their own right. They will all tell you, “Groupon, I had never seen a company that had so many things just all converge from a challenge perspective, that are so far beyond,” maybe even just any of the actions the company took, just so many things that came together that made it really tough.
In particular, when I joined the company, it was literally its second anniversary, and it had just hit a billion-dollar run rate. The next year, we went public almost exactly a year after I joined. In that period, we went from one to four billion, we went from one country to 48 countries, and we basically went from about 10 million to 150 million customers all in one year. After we went public, we had some issues where the auditors had been picked by the PCAOB to do the reviews. After we did earnings, we had to delay the filing of our first 10-K, which then meant you had to wait a longer timeframe.
With a company growing that fast, that young, you have to go in and reevaluate things like refunds. If you remember Groupon at the time, it grew so quickly, our average customer was only with us for three or four months. We were having to make projections on customer behavior on a customer cohort on an average of three and a half months, which basically means you got a fairly low confidence interval on exactly what it’s going to be. The more time you have to try to project, in this case, we basically filed our 10-K, the full 90 days right after the end of the quarter.
We saw that the return behavior was increasing. We had to basically adjust our finances. That created this downward spiral of a lack of trust in the market because you had to revise. It was not a restatement, but it was a revision to our financials, all that stuff. Having gone through all that, it was basically just a really tricky, very early stage. In fact, you probably know that the jobs that came partially out of that IPO, there were all sorts of quiet period violations. The company just had so much focus on it that it was just impossible to try to control the process.
Actually, came up with the confidential filing and the JOBS Act just right after that, in part because of how crazy that IPO was. I always thought that if I’m ever going to do another IPO, it’s going to be very different from that one. When I heard about Arm, I thought, “Here’s a company that actually was public before.” Went public in, I think, ‘98 went from a billion to roughly $25 billion valuation. In 2016, Masa bought the company for $32 billion. Does a lot of experimentation, does a bunch of things as a company does when it’s taken private.
Now they’ve tweaked the business model, and they’re getting ready to take it public again. My thought was, “This seems like the exact right setup I’d like to see relative to what I saw on Groupon.” We have incredible customer cohort information on the predictability of the business. We can, in fact, our business is extremely predictable. We basically sell licenses. Most of those licenses are renewals. We retain 99.5 percent of our customers.
The products that we sign contracts for and the royalties we receive are signed up right now. We just said this at our Arm Everywhere event two months ago. Our royalties for the next five years are already 85 percent under contract. All I really have to do is predict what I think the whole industry is going to do. We get paid on almost every chip in the industry. As a result, I was like, “This is a great setup as a CFO, because I definitely learned that the public markets really love a predictable business that does not disappoint and has low volatility.”
That was certainly a key ingredient. I also felt like Rene, he was a first-time CEO at about 60 years old. A very mature, seasoned leader who honestly had not been CEO before, just because he did not really care to be, but he certainly had the talent and capability. It just was not a focus. He had a really good makeup of someone who is mature, experienced and just really had the right mindset. We had Masa, who, as my boss likes to joke, is the only true long-only investor. He basically bought the company in 2016 and has not sold a share since 2016.
Even today, people say, “He owns 87 percent of the company. When the stock goes up, he must be willing to sell. It is like he believes his company is worth way more than even what our public company value is. It was a really good setup as a CFO who is trying to always control all these somewhat dynamic variables that are non-controllable. It was a really good setup to at least minimize some of the things that can create a lot of volatility.
What a fascinating story that is. One of the questions I would like to ask, Arm is making its biggest strategic move, I guess ever. As a CFO, you’re a very strategic, forward-looking CFO. What does that mean for you?
What it means is our business is going to get much more complicated as a CFO because one of the reasons I had not worked in the semiconductor business but because Arm really looks more like a software company, which is why it made sense that I would come from Splunk, and some of the things they wanted me to deal with were more to deal with. I helped Splunk go through the transformation from on-premise or enterprise software to SaaS. That’s when you have to do things like ARR or ACV and try to figure out how to reconcile some of the lumpy RevRec with metrics that can help you understand what the fundamental health of the business is.
Those are some of the things that were interesting and needed to be solved for Arm. By now, moving into a business where you still have that software component, but now we’re also selling chips, then you get into the whole world of having to deal with all the supply chain aspects and the working capital pieces. These are things I have dealt with in part at Amazon and a part of Jawbone, where I had to go out and get asset-backed loans and ABLs to fund working capital, or had to go to China and meet with contract manufacturers and go figure out how to work on defect rates or return provisions because of an issue that occurred at the factory.
I have had exposure to some of those issues, but those are much more complicated issues than dealing with software, which is mostly around RevRec and net retention. How do I make sure the customers love the product? That is going to be a challenge for us. Luckily, we have got a number of folks on the team who have worked in the Silicon industry. They understand a lot of the supply chain pieces, but it is fun, but there is a lot of complexity, certainly from a CFO perspective on that.
It is interesting because of the chip. Change is profound. I always say the CFO is the most trusted person within the company. You have got a narrative with the investor community and the financial community generally. How do you take that approach about this is why we’re changing, because it is not like you were doing poorly before this change.
We spent a lot of time on our big announcement of that. It was called Arm Everywhere. We did this in San Francisco in late March. We certainly did the big unveil, where the CEO shows the chip and talks about why now is the right time. Certainly, the timing was perfect. The rise of agentic AI, which really just took off with the open cloud moment back in February. Agents are very much a CPU-oriented workload, whereas training is more of a GPU or XP-oriented workload. What we were launching was the CPU.
The fact that this happened to be just two months after this shift to CPU occurred meant our timing was perfect. We’ve been working on this for a while. You cannot predict that stuff. That was certainly something that helped us. The thing I had to do then in my financial presentation was explain from a financial perspective why this is the right thing to do. Basically, I had to put together some slides that showed that when we sell IP, we have two different flavors of the intellectual property that we sell.
We sell just IP blocks, which would be, here’s a CPU, here’s a GPU, here’s an NPU, here’s a memory configuration, and we’ll charge a certain rate for that. Typically, it works out to be about 5 percent as a percentage of the value of whatever the end product is. Since this is a server chip are usually anywhere from $1,000 to $2,000 or $3,000. I basically just said, on a per $1,000 basis, if we’re selling just basic IP blocks, we get about 5 percent, or call it $50.
If we build what’s called a compute subsystem, which is a fully optimized, almost a compute die, or like a chiplet, where it’s basically a CPU, memory and any of the fabric or mesh that a chip needs, and it ends up being probably about 80 percent of the design of the chip. If we sell that, which we just started selling when we went public a couple of years ago, we get about a ten percent rev share or $100. If we sell the full chip, we actually get about a 50 percent margin, and we get about $500. It is about 5X to 10X more profit per unit than you get if you’re just selling the royalty.
The question is, I get that, so then help me understand how you’re going to manage the fact that you used to sell just the IP to your customers and now you are producing chips. Our perspective was that now is the perfect time because the market for this product is so massive. There is really no way any one vendor can serve all the demand.
Truly, there are so many different pockets of demand that what we’re going to sell is only going to work for certain use cases. Amazon builds its own IP on arms, so does NVIDIA, so does Microsoft, Google, Alibaba, ByteDance, every one of them. A bunch of those folks also want to buy a chip from us because they have different workloads in different use cases. Anyway, that was basically the needle we had to thread. It seems to have gone pretty well, and certain stock prices have had a positive reaction since then.
It’s also still early. We have not shipped the product yet. It does not ship until later this year. I have to keep telling everyone, look, we’re not shipping until the end of this year. Everyone, please do not run too far ahead of us. There is a lot of demand, but there are also a lot of supply chain constraints. No one could get a memory. TSMC is sold out. There are all sorts of challenges. There is a lot of execution that still needs to occur.
Many things are beyond your control, even for the best managers of companies. I want to ask you, what does success look like for, say, five years from now? We live in such volatile times that it’s a very difficult question to ask. What will fundamentally change about the company in the next five years?
The objective answer, as CFOs, we hate to give long-term targets. It’s the bane of our existence because then you are stuck and committed to the future. When we had this event, because this was such an important shift, we had to be a little more aggressive. We actually gave a five-year target of where we expect to be. We basically said, “Look, we went public at about a $2.6 billion revenue company. We’re now just shy of five billion. About doubled in roughly two and a half years. Over the next five years, we’re going to grow to a $25 billion company. We’re going to grow 5X.
The objective answer to your question is, where is our Arm going to be in five years? It’s going to be five times bigger, and it’s going to develop, it’s going to be five times bigger in revenue, and it’s going to be, I think, it’s about six or seven times more in earnings per share. That’s the objective measure. How do you do that? You have to continue to deepen our penetration across all the various aspects of cloud AI as well as edge AI. Luckily, we’re one of the only companies on the planet that virtually every smartphone, tablet, PC, whatever, they’re all highly dependent upon Arm.
As you start to see the AI move from mostly a cloud-based product, but starting to go into self-driving cars, robots or smartphones, and or smart devices, Meta, Apple, Glasses, all the different products out there, they’re all using Arm. We certainly have an opportunity, and of course, our internal targets are to do much better than the 5X. I do expect that we are in a pretty unique spot to see, as we watch AI proliferate, there are all sorts of pretty aggressive aspects of, when are we going to hit AGI or ASI?
When humanoid robots are going to be constructing data centers and fabricating chips on the moon, all that stuff is on the moon. All those things are certainly possibilities, but the good news is our designs, because of how efficient they are, generally are at the heart of every one of these efforts. It’s exciting. That certainly is exciting.
I’m about to ask you a question, and the answer that you give is the one I’m going to start giving to everybody, because I think you’re uniquely positioned to answer it. There’s this ongoing debate around AI investment and whether it’s going to deliver real long-term returns. I just hear it all the time, and you’re in the throes of it. What do you think? Is it for real?
Yes. If you step back and say, “Return obviously is going to be profit or cashflow.” The next step before that is that it needs to be like revenue, or we’ve seen revenue from AI. Clearly, if you just look over the past few months, we hear about OpenAI, which, of course, our chairman is a big investor in, and we also see everything coming from Anthropic and even Gemini and all the different products out there. The revenue growth has been through the roof for all the model companies. What we’ve seen now is, it’s clear that the agentic AI revenue model is significant.
In fact, unfortunately, it’s so significant that it’s really caused a lot of questions if you’re a big SaaS investor, because we do not know what the terminal value of those companies is anymore. What is the terminal cashflow and profitability, and to what extent are they going to be disrupted? The question of revenue and ROI has progressed a lot just in the last three months. Now, I think the next question, though, is that we’re still constrained by computers. Computers are there, there is power, there are data centers and chips, and all these things are constrained right now.
There are always talks about bubbles, whatever, if there’s a bubble now, it’s going to be different than any other bubble we’ve seen. Everyone famously talks about the 2000 dot-com bubble, where we had a huge infrastructure build-out, but it was not really being used. Literally took 15 to 20 years until a lot of that infrastructure got used, and what’s so surprising right now is all the infrastructure that’s being deployed is literally getting utilized immediately. A lot of the, like, when you use an AI model, like Claude, you’re throttled.
AI companies, including OpenAI, are constrained by computing capacity. What’s interesting is that there will be enough constraints that will force much more efficiency and effectively lower the cost per token, which could significantly help the AI investment part, which then theoretically could really make the profit at every dollar of revenue created significantly increase. If you look at the cost per token, I think it’s come down to something like 100X just in the last two years. Everything we see is it’s likely to continue to get much more efficient.
The ROI of AI is that it does not take a lot of guesswork. It’s pretty clear the question’s going to be, it’s more about the displacement. What part of white-collar jobs are taken from, are going to be replaced with AI? What are the new jobs that are going to be created? That’s the piece that I think will be really interesting too. I do not know for certain, but there are definitely going to be some pretty interesting transformations occurring.
Jason, one thing I would like to ask you about is KPIs. Even though you have an accounting background, you’re not like a traditional CPA/CFO at this point in your career. When you measure success and report success to your colleagues, what are three or four of the most popular KPIs that you personally like to use that you find meaningful?
First, I would say you’re right. To be honest, I was not that great of an accountant. Luckily, in the spirit of what the two most influential words ever mouthed, which are known themselves, was that Socrates, or I do not know, some pretty smart philosopher. Pretty early, I realized that being around some genius accountants, I am not a genius accountant. I probably better figure out something else that I am better at. I would say when I think about where I am at now as a CFO, and I try to figure out, like, what are the KPIs that I use to measure success?
To me, one of the things that was really instilled in our culture at Amazon was this concept of bar raising. We always want to hire people who make us nervous because they’re better than we are. We know that we’re going to learn from those people that we hire. That was just such a core part of the Amazon culture and still is. It definitely created a mutation in my DNA. As a result, to me, the best measure is as a CFO I have now had five CFO roles. How many CFOs have been produced from the finance teams that I have been able to look at enough to manage?
As of a couple of months ago, I did a check. It was up to something like 30 plus CFOs that have been part of my team at Groupon, Jawbone, Opendoor, Splunk and now Arm. That means that clearly these are organizations that are providing the right development, the right exposure and the right growth to be able to launch careers that are fulfilling for folks. That to me is probably the most, to me, that’s the best metric that encapsulates. That, at least, the organizations that I have been able to manage are actually creating great opportunities. That’s the main metric I focus on.
The one other thing I try to always do is, it’s also a bit unique, I’m told, but I basically, for all of my direct reports, I know right now there’s a CFO shortage. I tell them, like every year or probably more like every six months, I just say, tell me, “Where do you want to go? What do you want to do, and how can I help you?” I probably get one solicitation every couple of days on CFO jobs.
I tell them, “The ones that are really good, I’ll share them with you.” I’ll have a discussion about whether this makes sense for you. Is it the right fit for you? Is this the right opportunity for you? I tell my team, as we go further down the organization, how are we always finding opportunities for folks that are going to help grow and develop? First, I hope that’s in the company.
If they’re outgrowing the company or outgrowing the role, and there is not an obvious opportunity for them, how do I help them be able to find that opportunity outside the company? I find that if you have that approach, you will get the best people and they will be loyal to you and the company because they know that you have their own interests at heart. That’s probably one other thing, I would say, it’s not really a metric, but that’s another core focus.
That’s great. At the end of the episode, I am going to share Jason’s email, and you’re going to be getting about 2000 controller CAO VP of finance resumes mailed directly to you. Everybody should have a boss like you at some point in their career. I have a theory, and you did not mention Amazon is one of the companies, but my personal theory, which in academics could prove or disprove, but I think that Amazon, as a company, has produced more CFOs amongst their alumni than any other company in the country. I meet so many people, including you, who are highly successful CFOs who worked at Amazon at some point.
Bezos early on said we’re a low-margin business and we need to hire world-class finance. Joy Covey, who was brilliant, had basically retired, and Bezos said, “I got to go hire the best.” He went to GE in 1999. That was the best place to hire CFOs from. He hired Warren Jensen. Warren was a very successful CEO, and then he brought in Tom Szkutak, another successful GE CEO. I spent a lot of time training under those folks, and the mentality that GE had built for so many years, as you know, was the CFO factory.
You created not just smart CFOs, but smart general managers who could really understand the business from all angles and then hopefully have the detailed orientation to understand all the core technical aspects. That was definitely apparent at Amazon, and you’re right. I did not count anyone at Amazon because while they may have worked for me, I was not the CFO, and I would give that to the CFO. I’ll let the CFO, Brian Olsavsky, know he can count those folks, but you’re right. There may be a hundred CFOs who came from Amazon now. I do not know. It’s a big number, though.
I bet there are a hundred. One thing I would like to chat with you about is that you’ve had a front row seat as a participant in it, but just the simple nature of the CFO roles has changed so much over the last 10 to 15 years. How do you see it changing in the next few years? Are there any parts of it that you do not think will ever change?
The things that will never change, one of my mentors had told me, is that you have to realize that the CFO gets a pass on compliance and accuracy. That just has to be the case. Whatever it takes to make sure that the numbers are right, that the company and the market can trust what you say and the numbers you’re projecting. That just will not change. You get a pass or fail. You do not get a grade. It’s just a fail. There’s, I would say, the strategic piece, which is really about resource allocation. What are the mechanisms we use to try to have the best estimate of where we deploy our capital and our resources, and how do we get the optimal ROI from that over whatever time horizon?
That, of course, is like the core part of the job, but at every company, it’s so different based on industry and time horizon. It just depends on so many factors, but I do not think that changes either. The mechanisms and the tools that help you understand your resource allocation. We have never had better tools than we have right now. Someone said, “Should Arm consider looking at acquiring the XYZ company?” I could just go into ChatGPT and say, “Give me a quick SWOT analysis on whether or not we should acquire this company.”
Does it make sense? In five minutes, I can get an answer that I used to pay a consulting firm, a banker, or my corp dev team to spend two weeks to do. I can now get it in five minutes. It is phenomenal. The access to information that exists, which I think can really help accelerate that resource allocation process, which, when I started my career, was a long process. I got to go back to the AI thing. When I started my career at Anderson, I had a briefcase, yellow paper and pens.
I would go to the office, and I would call my voicemail, and it would tell me where I was supposed to go. Five years in, I got a shared notebook. I got my own notebook. I got Excel. I was like, “How many people in this world today live and die in Excel?” That did not exist more than 25 to 30 years ago. It didn’t even exist. When people say, “It’s this AI, it’s like it’s going to crush all these jobs.” It was just like a lot of these jobs did not even exist 25 years ago.
They were just completely different. What’s happening is the Excel spreadsheet jockey. That job is probably going to have a short life. It had a 25-year existence, but how you use these tools to get to an answer on resource allocation, I think, is incredible. I do think you could see CFO teams that could be a lot smaller than they used to be. I think the structure of those teams, the ratio of senior to middle to lower management, probably changes a bunch. Those two aspects of the CFO role just do not change.
It’s interesting because I think Excel came out when I was a senior in high school. When I decided to study accounting, a lot of smart people, by the way, just ended up being wrong. They told me to get out of accounting because information technology would replace accountants. The exact opposite happened. They created and empowered accountants and financial people generally. That’s a great answer. I always like to ask, because our audiences tend to be on the cusp of their first CFO job or maybe actually in their first CFO job, but what’s some advice that you could give to the next generation of financial leaders to help them achieve some of the success that you’ve achieved?
I do not think there is a better time to start your career than now. The uncertainty, and there is all this AI doom that you could find, at least in the US. Different countries have different views. It’s interesting. For someone who is curious and wants to work, has some aspect of tenacity and grit, and really just wants to go figure things out. You have more tools at your disposal than you have ever had, and they’re cheaper and easier to access than they’ve ever been. You do not have to have gone to one of the top schools. You do not have to have been born in the right place.
Meaning, a lot of the things that determined whether you’re going to be successful before are just mostly eradicated. I would say that the opportunity is just that the world has been flattened in a pretty massive way. I would say follow your curiosity. Every successful person I know has probably just made tons of mistakes. I have had five CFO roles, of which there have been a lot of nasty articles and messages that I had to get over bad decisions, or being in the wrong place at the wrong time or just making a mistake. It’s not like that will happen. The question is, how are you going to deal with it? Are you going to have the resilience to basically learn from it and then not make that mistake again?
I hear you. One thing I noticed about my first CFO job. It was from the Boston area it was like being the manager of the Boston Red Sox because everybody thought they could do the job better than I could. They did not really appreciate how hard it was and how you had to prepare for the role. Just anyone could walk in and be a CFO. That’s no longer the case, although there are still pockets of people like that. Anyway, I think that’s great advice. Jason, I know you’re a really busy guy at one of the really interesting companies in the world. I want to thank you for your time. Also, I just want to give you the final word. If you have any parting thoughts for our audiences.
I would just say, be curious. The world is changing so quickly. I would say embrace AI. I’m energized by what I see with AI and what I am learning myself by constantly using AI tools to help me with specific tasks. Of course, you also want to make sure everything is accurate. I feel like with curiosity plus AI, the learning and growth opportunities are unlike anything we have seen
