In this podcast, Motley Fool analyst Tim Beyers and host Mary Long break down Salesforce's earnings and ask whether the stock's slump is warranted.

Then, Motley Fool analyst Asit Sharma and host Ricky Mulvey take a look at some companies that could ride the tailwinds of Nvidia's standout earnings.

To catch full episodes of all The Motley Fool's free podcasts, check out our podcast center. To get started investing, check out our quick-start guide to investing in stocks. A full transcript follows the video.

This video was recorded on May 30, 2024.

Mary Long: Salesforce slumps, but is the street missing something? You're listening to Motley Fool Money. I'm Mary Long joined today by Tim Beyers. We are both recording virtually, but from sunny Colorado. Tim, thanks for being here.

Tim Beyers: Thanks, Mary. Fully caffeinated, ready to go.

Mary Long: Love that. We are having a sunny day here, but not such a sunny day for a Salesforce. Today we're shining the spotlight on that company because it's down more than 20% this morning. That is largely due to the fact that the company just reported its first revenue miss since 2006. That sounds pretty bad, but I will say up top, bottom line results were solid. GAAP net income up almost 700%, free cash flow up 43%. Tim, we've got some good, we've got some bad. The street seems to think there's a lot of bad. What say you? What stuck out to you is most important in this report?

Tim Beyers: Well, the street doesn't like to be surprised, and they were surprised. First quarter revenue coming in at 9.13 billion. That was under the consensus forecast that I got from Capital IQ of 9.147 billion, so that's a mess. The street does not like a mess. They like it even less when you follow that up with a Q2 forecast, that is way below what the street wanted. In this particular case, the midpoint in the Q2 guidance, Mary, was for 9.225 billion, and the street wanted at least 9.34 billion. Man, that is a serious mess. Also, looking ahead to the full year, Salesforce said that their guidance on margins. This is operating margins. They issued operating margin guidance in Q4. This is three months ago, and they said, look, we're going to hit our gap, operating income is going to be a little over 20%. It's going to be 20.4%. Then this quarter, they said, you know what? We have maybe a little extra stock based compensation. We're going to take that down 50 basis points to 19.9%. Just surprising the street all the way around is really not a great look. Having said that, Mary, what stands out to me is that even when you subtract all of the things, so you subtract everything. You subtract stock based compensation, you subtract all of the capital expenditures, you subtract the dividends, you subtract a $2.1 billion buyback of shares. We can talk about whether or not they should be doing that buyback. But, let's just say it's fine, that's a lot of spending, and even with all of that, they still end up with 2.75 billion in free cash flow. This is a very healthy business, but it's not growing as fast as it was.

Mary Long: Management's explanation for the slower revenue growth is that enterprise customers are, they're working with tighter budgets. They're reevaluating contracts, taking a closer look at, like, every penny spent. Does that mean that this mess is more like a macro story, or is it indicative of a longer term problem specifically with Salesforce?

Tim Beyers: I'm going to use a dumb term that has been around the tech industry forever, so bear with me here, Mary. Elongated sales cycle. This is something that we hear about in enterprise software, everywhere. It's been around for decades. It's not new, but it happens to be hitting Salesforce right now. So is macro part of it, maybe? Also, just in a year of efficiency, our company is going to be a bit more judicious about how they spend their capital, especially when you have investors, board members, customers saying, give me AI. I want AI. Have you spent on AI yet? I'm sorry. I haven't heard you. I thought you said you were going to spend on AI. When that is the narrative that's coming out, Mary, it's a little bit harder to go forth and spend on something that is strategic, like a Salesforce product, but is fundamentally, a customer relationship management system. Now, it's more than that. It does more. It manages what we would call front office operations. Front office, everything facing the customer. Back office, everything that's not facing the customer. We know what these things are. Marketing faces the customer. Inventory management doesn't. So back office front office, and Salesforce is arguably the most significant and most important provider of that front office software in the world. But, there are other priorities here. You may have AI priorities, AI hardware spending, and so forth, just taking some money out of the budget.

Mary Long: Let's talk about AI because I felt like listening to this, I counted 57 mentions of AI or Artificial Intelligence.

Tim Beyers: That's it.

Mary Long: Anyway, almost felt like every other word was AI or Artificial Intelligence. Just based on that alone, it seems like that's quite a priority for Salesforce that said, looking at this 20% plus slump, I've seen headlines that have attributed the revenue miss to Salesforce falling behind on AI compared to competition. What's the argument there? Benioff mentioned in the call that other AI players, he called them commodities, says, "That's not where the intelligence lies." His point was that there's an importance to data here, and Salesforce has tons of data. Some of the numbers that they threw out are mind boggling. They process two quadrillion records of data. Salesforce manages 250 petabytes of data. They're saying that that's a huge foundation for this AI game that is to come. Do you buy that?

Tim Beyers: Yeah, I do buy that. I think it's a little bit arm wavy, but it's also true. All of those numbers, yeah, that's a lot. There's a lot of data that are in Salesforce systems. But he's making an argument that I think is fundamentally correct that your AI algorithms, however you build them, whoever you use, whatever hardware you've got, is only going to be as good as the data that that hardware and those algorithms point at. Whatever's pointing at the data, can be fast, it can be resilient, it can be really expert. But if the data itself is terrible or not really useful, then what insights are you going to get? What he's saying is that in the fundamental AI revolution where you are trying to automate real good workflow, you're trying to gain insights from data. The things that Salesforce helps you manage is fundamental to that process. He's trying to position them in the middle of the AI revolution. Again, I think he's not wrong. Now, having said that, is he being hyperbolic? Newsflash, Marc Benioff is hyperbolic. No. That couldn't be true. Of course, he's being hyperbolic. But at the same time, he is right that what they're doing here is fundamental. I'll mention one other thing, and we could keep moving here. They did invent a product internally that is designed to make Salesforce a bit more sticky in this area. In other words, take the data that you already naturally put into Salesforce and then add other contextual data to it very easily, and they call that Data Cloud. They have a couple of features inside of Data Cloud that I think are interesting. First is, it's like a set of programming interfaces, APIs, so you're going to bring in data from, say, like a snowflake or you bring it in from AWS, and just these other sources that add to the richness of those customer records and sales records and marketing, all these things that you're keeping. Then there's another feature to it where they have what, I believe they call it no copy or something like that. No copy partnerships. What essentially what that means is the data comes in natively. You don't have to do any transformation of that data in order to get immediate value from it inside of Salesforce. That strikes me as interesting. It's not clear to me exactly how it works, Mary, but there is something smart about Salesforce saying, essentially what Tim White has said, and I think it's one of the smartest things he said. Tim says a lot of smart things, ''Data is the food of AI'', that's what Marc Benioff is arguing. He's making Tim's argument. Data is the food of AI, and the most important front facing customer facing data lives in Salesforce, so don't count us out.

Mary Long: I'm seeing two stories here. On the one hand we have a perhaps hyperbolic Benioff positioning this company as in the middle of the AI revolution, Data Cloud being a part of that, etc, basically, the growth story is not over yet. Then the other story that I'm seeing, right, is this, like, not insignificant slowing down of revenue growth. Even like as you pointed out, this is a company that's generating a lot of cash. They're returning a chunk of that cash to shareholders. In the form of share repurchases, dividends, they're anticipating announcing a quarterly dividend next month. Taking all this into account, what story do you see as being more true? This next growth story, or is this the beginning of Salesforce becoming a mature enterprise software company?

Tim Beyers: Well, [LAUGHTER] if you listen to the street, they will say, growth story is over. If we were to visualize it as a meme, you have algorithms as babies throwing cash into a big giant bonfire, and they're all upset about it. They're all bent out of shape. I think two things can be true here. I think that Salesforce can be maturing, that growth rates can be sliding into the high single digits, low double digits for a long period of time, where there's dividend growth over a long period of time, improving margins on a company that will just not go away because it's too sticky to remove, Mary, that's not just a good business, that's a great business. Can you name other businesses that fit that same profile? I'll give you one. Microsoft. Microsoft fits that profile. Now, is Salesforce going to be the next Microsoft? No, I'm not going to predict that. But if the contours of Salesforce are becoming more like the contours of Microsoft, why would you be not OK with that? That's a great place to be. I do think that you should be very skeptical of AI as a catalyst to return Salesforce to 20% plus growth rates. I don't buy that for a second, and I don't think Benioff is arguing that. What I think he's arguing is we have never been more important. If he's right about that, and they're sticky, and they can keep disciplined, and keep solid margins and grow them just a bit every year, half a percent, 1%, just get a bit better, a little bit more, they are going to be a cash machine for years. That'll be something interesting. I'm looking forward to seeing how the street rerates this stock. If they just keep crushing it such that the free cash flow yield just makes it look cheap, then I may be interested in buying some more and I already have a lot of Salesforce stock.

Mary Long: Yeah. That works out well for us.

Tim Beyers: Yeah, I am OK with that.

Mary Long: We were chatting yesterday afternoon about this a little bit, and you said that 20 years ago, if someone had said Salesforce would become a player in enterprise software, most people would have laughed at that comment.

Tim Beyers: A lot of people did. Not just would have, they did.

Mary Long: They did. It happened.

Tim Beyers: It did happened.

Mary Long: Spoiler alert, that's what Salesforce has become a player in enterprise software. That's maybe even an understatement. Why was the idea of Salesforce becoming an enterprise software player once so difficult to believe?

Tim Beyers: Because it required a paradigm shift. You had to move from this process of buying software, and you would buy the software itself. You would own the code. You would own all of the upgrades. You would install it yourself. You would maintain it yourself. You had poor souls, like my friend, Tim, like going into work on weekends to do things like major upgrades because the network had to be taken down, and then the new upgrade would be put in place. That was a nightmare. That used to happen, IT people would be like, everybody out by like 3:00 PM, on whatever day, and then the network is going to be taken down. All these things are going to happen so that you could stand up the new software or the new network upgrade or whatever it is. You had to believe that that paradigm that had existed for such a long period of time, would be replaced by, you know what? You don't have to do any of that. You can just rent the software from us. We're going to deliver it over the Internet. All the upgrades, they're going to be handled on our end. If you have Salesforce, you don't even see the upgrade. You might see a new field, but you don't experience the upgrade. You just log into your system here. You had to believe that that was something that people would trust. A lot of people said, that's insane. Maybe there's a couple of small software apps that can be delivered via the Cloud, but enterprise software? Are you out of your mind? Four hundred billion plus dollars worth of software sales are going to move to a totally different paradigm? You have lost it, son. That was the argument. My argument at the time was that, well, it doesn't have to be all of it. Even if it's just a small portion of it, this is going to be a much bigger company. It turns out that that actually has become the new paradigm, that virtually all software of any meaning, is delivered via the Cloud now. But that 20 years ago, Mary, a lot of people didn't believe that was going to happen.

Mary Long: Tim in preparation for this, I had asked you if you had anything that you really wanted to rant about, and you had said, not really. I don't know that I have anything, but I think we got you on a rant about Salesforce, and why it's maybe not as down and out as the street wants us to believe.

Tim Beyers: Algorithms throwing out the stock, like babies chucking cash into a bonfire is not the meme I expected, but feels like it fits.

Mary Long: But it's the meme that we got. Thanks so much.

Tim Beyers: It's the meme that we got.

Mary Long: Thanks so much, Tim for diving into this with me today. I always appreciate having you on.

Tim Beyers: Thanks, Mary.

Ricky Mulvey: Today's show is sponsored by public.com. That's where you can earn a 5.1% APY with a high yield cash account. While we can't say for certain, it's the highest interest rate there is, we can say this, it's a higher rate than SoFi, a higher rate than Marcus, a higher rate than Wealthfront, a higher rate than Betterment, frankly, a higher rate than Capital One, a higher rate than ally, a higher rate than Barclays, a way higher rate than Bank of America and Chase, a higher rate than City Wells Fargo, Discover, and it's a higher rate than American Express, too. If you want to get started earning 5.1% APY on your cash, check out public.com. We can't say it's the highest interest rate for your cash, but it's up there. This is a paid endorsement for public investing, 5.1% APY as of March 26, 2024, and is subject to change. Full disclosures and terms and conditions can be found in the podcast description, US members only.

Mary Long: Last week, NVIDIA recorded some stellar earnings again. Up next, Asit Sharma and Ricky Mulvey, look at three companies that stand to benefit from the chipmakers record results.

Ricky Mulvey: Last week when we were talking about NVIDIA , I had a question about the ripple effects on other companies. Then we just straight up ran out of time, but I still wanted to talk about it with you. There is this boom in data center spending, and NVIDIA is not the only beneficiary, and in fact, there may be some other large and small tech companies that are riding the co-tails of this wave a little bit. The one company when we were Slacking about this, that you "def want to hit" is AMD. In terms of this data center spending. How does the boom? How does this data center spending boom affect AMD?

Asit Sharma: Wait a minute, Ricky, before I jump in, I really take issue with you revealing what my normal speech patterns look like to a wide audience.

Ricky Mulvey: I think it's lovely and personal.

Asit Sharma: I definitely want to hit AMD. This is a company that has a wide playfield playground in the chip industry. It doesn't only make data center GPUs. It makes chips that we use in laptops and computers, field programmable gate arrays, FPGAs, etc. They are really widespread when you think about computing both in data centers and just in everyday life. But AMD, is going to have a spillover effect from all of this investment in data centers. It has a competing product, a very advanced GPU that goes up against NVIDIA's omnipresent, all powerful, great GPU stacks, and that's starting to get a little bit of traction in the marketplace. They're only doing about $4 billion a year in GPU data center business, AMD. By comparison, NVIDIA as we all know sells in the tens of billions of GPUs each quarter. While it is a distant Number 2, it's starting to see a little bit more of business from just the amount of demand that's out there. I think what's good for NVIDIA remains good for AMD.

Ricky Mulvey: This is where AMD still has a strong growth story in spending for Data Center GPUs, even though it's a smaller part of their business. Let's say, I have a large language model, Asit, and I want to run it somewhere. Why might I put that on AMD servers instead of in NVIDIA? Is there a quality difference or is this a pricing game?

Asit Sharma: It's somewhat of a confluence of factors, Ricky, and you hit on one. The AMD processors are a bit cheaper right now than NVIDIA. NVIDIA can price as it wants. Also, they're accessible. If NVIDIA GPUs are in such demand in the Cloud provider that your AI is going to use, let's say that we're a start-up that needs to develop our own LLM and train it, right now, you might have, depending on who you're using, a bit of a weight to even get to the specific GPU setup that you want with NVIDIA, so there's that, and I would say they're comparable. There are some start ups now who are trying to test the benchmarks of the NVIDIA GPU stack versus AMD, and really the message is, sure, NVIDIA on most benchmarks is a little bit faster than AMD. The next generation that you and I talked about last week is going to be even faster, but for a comparable price, for something that's available, really, you don't have a ton of difference if you're working to train a large language model, and also for the inference part when people start interacting with that model. You've got an alternative that people want to use, you've got one that is being promoted by various Cloud providers like Microsoft. The implication for AMD is that it is going to be able to lean a bit into this demand. We should see that $4 billion annual run rate with its GPUs start to expand in the next couple of years significantly.

Ricky Mulvey: You're telling me that AMD is store brand GPUs, we don't need to go out for fancy GPUs, we have GPUs at home with AMD?

Asit Sharma: We want the expensive GPUs precious. We can't reach them just now, so there is an alternative out there. Let me throw in this quick example, which has seemingly on the surface, Ricky, nothing to do with AMD, but it has everything to do with AMD. We just heard that xAI, Elon Musk's AI start-up using a chat bot called Grok G-R-O-K, not to be confused with a large language model called G-R-O-Q. They're going to purchase, this start-up that Elon is backing, it's going to purchase up to 100,000 of NVIDIA's H100 GPU, so billions of dollars of the H100. Well, that's an older technology now. NVIDIA has moved on to the H200. It's got the Blackwell architecture coming out. Why is this start-up backed by Elon Musk investing in this older technology? It comes back down to this confluence of factors I was talking about. Price, they're cheaper than the next generation availability. They're here now. Comparable, they're comparable to the future generations of NVIDIA's own product. If Elon is starting to think along these lines, you can just picture drilling down to small start ups who can take that to the next level of logical reasoning and go ahead with the AMD processing stack.

Ricky Mulvey: Perhaps the comparison to store brand ice cream is a little mean, especially when the products that AMD are making are good enough for Microsoft, that multi trillion dollar tech giant. What's going on between AMD and Microsoft? Who does have a lot of partnerships with different semiconductor companies?

Asit Sharma: Sure. Well, AMD, was actually backed by Microsoft early on in their race to catch up with NVIDIA. This is something that I think many investors get a little confused by. Why would Microsoft even bother with AMD since they're so deeply intertwined with NVIDIA's platforms? The reason is that they want an alternate source of supply, as does Oracle, as does Amazon. All these Cloud titans don't want to be dependent on NVIDIA at the end of the day. They like the business, NVIDIA is helping them realize billions in revenue, but they don't want to be beholden to this one provider. That's also the reason, of course, why they're all developing their own chips and platform. AMD plays right into that with their MI300X, their GPU accelerator, and Microsoft is looking at the prevailing winds and understanding the same factors are at play. They're starting to reach some capacity limits in what they can offer to customers. They almost have to move forward with this little relationship they started with AMD for Cloud center GPUs. It makes sense when you think about those dynamics of demand and supply.

Ricky Mulvey: Let's talk about Micron, which is a name that may be a little bit less familiar, but they do memory and data storage chips, basically. A lot of spending on data centers. Why is Micron feeling the tailwind of all of this spending?

Asit Sharma: Micron Technology it's been around for a while. They're one of the largest memory providers in the world. Ricky, as you pointed out before the show, they're like a commodity business, not a terribly exciting business, but they offer a very interesting memory component. It's called the HBM3E. This is high bandwidth memory. What's so interesting about this component is that it works well with a certain type of chip architecture used by both in NVIDIA and by AMD. When these companies build their GPUs, they no longer have these single dye components. They're stacking a lot of components. They're stitching together memory compute into one really dense package on a slab of silicon. They're building Lego-like structures to make all this function with the densities of these products, you need memory that can interact with the different components on these dyes, and Micron has these components that can also be stacked and configure right into what NVIDIA and AMD are building. The HBM3E is the latest iteration of this memory module. It's going to be part of NVIDIA's H200 tense record GPU. So that second generation of GPUs that we're talking about last week, I think it's in production now. It's going to be available a little bit later this year. It may already be finding its way into some components. This is a tailwind for Micron, and I think the demand that we're seeing in data centers shows that it's a new part of the market the company can play in. The product consume management has about 30% less power than other offerings, and there really aren't many offerings that compete with it. Stretching beyond 2024 and 2025, they are capacity-constrained. I think you'll see here a chance for Micron to have this higher margin higher volume product that makes them a little bit less of a commodity business than before.

Mary Long: As always, people on the program may have interest in the stocks they talk about, and the Motley Fool may have formal recommendations for or against, so don't buy or sell stocks based solely on what you hear. I'm Mary Long. Thanks for listening. We'll see you tomorrow.