Matthew Eash — better known by his online moniker “muji” — is one of the world’s most forward-thinking technologists. After spending decades as a data architect, he’s now founded his hhhypergrowth newsletter to focus on the technologies powering today’s most innovative companies.

In this exclusive podcast, 7investing CEO Simon Erickson speaks with hhhypergrowth founder muji about several of the new technologies and investment opportunities that are arising in cloud computing.

Muji begins by describing the newest developments at Amazon’s most recent “re:Invent” cloud computing conference. The cloud king continues to roll out new products and features for AWS, though they are often confusing to customers and even its own sales team. While Amazon itself coined the term “serverless” to refer to on-demand service and usage-based pricing, there is contention on whether its newest products truly adhere to this term.

The two also discussed the potential implications of OpenAI’s recent ChatGPT open-source conversational (and controversial!) chatbot. Muji believes the true opportunity for AI in is writing and sharing code, which could be useful for Microsoft’s GitHub or its competitor GitLab. Usage-based software subscription licenses are quickly replacing per-user or per-device models within tech-heavy fields like cybersecurity or IT operations, though they’re also expanding into ad-supported media for companies such as Netflix. This allows software platforms to better monetize their power users, who generally rack up more hours of viewership or induce higher costs every month.

Silicon Valley is going through a round of layoffs, meaning there are fewer technology and IT workers today than there were a few months ago. This could impact software-as-a-service (SaaS) companies who price their products on a per-seat basis, such as GitLab, Zoom, or CrowdStrike.

Speaking of CrowdStrike, Simon and muji compared and contrasted the company to its up-and-coming competitor SentinelOne. Muji believes CrowdStrike recently hiring two of SentinelOne’s executives is a pretty big deal, though SentinelOne remains a very compelling option for small and medium businesses.

The two went on to discuss the go-to-market strategies of companies, and how it is often difficult to pivot in their sales approach. Companies like Okta often sold to larger enterprise accounts and are now trying to sell to smaller development teams, while companies like Twilio who typically sold to developers building apps are now looking to move upmarket and land larger deals. The transition in software sales is rarely easy.

In the final segment of the conversation, muji discusses three other publicly-traded companies he is a fan of: Datadog, Bill, and Zscaler.

Publicly-traded companies mentioned in this interview include Akamai, Alphabet, Amazon, Apple, Bill.com, Cloudflare, Confluent, CrowdStrike, Datadog, GitLab, Meta Platforms, Microsoft, MongoDB, Netflix, Okta, SentinelOne, Snowflake, Splunk, Twilio, and ZScaler . 7investing’s advisors or its guests may have positions in the companies mentioned.

Agenda

01:00 –  Amazon re:Invent conference recap

07:17 –  Cloud computing comparison

10:36 –  Digital assistants: The battle for your living room

13:37 – The coming impact of ChatGPT

17:43 – Open source for the enterprise: GitHub vs GitLab

21:07 – Usage based pricing models

32:07 – Cybersecurity: CrowdStrike vs SentinelOne

36:56 – Three companies muji likes: Datadog, Bill, and Zscaler

Transcript

Simon Erickson  00:02

Hello, everyone, and welcome to our 7investing podcast where it’s our mission to empower you to invest in your future! To learn more about our 7investing membership, where we pick our top seven stock market ideas each and every month, visit 7investing.com/subscribe and get started today for just one single dollar.

Simon Erickson  00:20

My name is Simon Erickson; I’m the founder and CEO of 7investing. I’m really excited today because you’re we’re going to be looking at the technology world’s most innovative trends. And who else could we possibly talk to about those developing trends than muji. Muji is one of the most innovative investors out there. He’s got his hhhypergrowth newsletter — that’s hypergrowth with three h’s dot com — if you want to follow along with a lot of his coverage. He also has a newsletter subscription. I always enjoy these conversations with you. Thanks for being a part of the 7investing podcast today.

Muji  00:50

Thank you. Yes, I think this makes podcast number three, always a pleasure to talk to you.

Simon Erickson  00:56

It does. And we always find a way to go in about 100 different directions. We’re going do the exact same thing here today, which makes it interesting.

Muji  01:02

Alright. Speed round!

Simon Erickson  01:03

We’ll talk cybersecurity. We’re going to talk about GPT. We’re going to talk about a whole bunch of different companies. We’re going to talk about questions that were posed to our Twitter account. But to at least frame this somewhat muji, let’s chat first about cloud computing. Because I know in your newsletter, you’ve been chatting quite a bit about the Amazon re: Invent conference where they kind of roll out a whole bunch of different product features. As we were chatting about right before this show, it seems like Amazon always has its own 100 different directions that it’s going in. But what are a couple of things that you’ve been following with AWS and their cloud presence lately that you think that we should be paying attention to?

Muji  01:38

Yeah, you really nailed that they have so many different kind of point solutions that they’re building for some particular purpose. That some, sometimes from the outside observer standpoint, it looks like there’s very little rhyme or reason for some of these product lines to spin out to. And Amazon even has products that compete with each other. Like I’d say, Athena and Redshift are both building query engines around themselves that are starting to compete with each other, much to the confusion of their own customers. One being kind of a data lake, query oriented in Athena; and the other being data warehouse oriented.

Muji  02:22

But Amazon is definitely, you know, last year’s focus at re: Invent was really around serverless. They were adopting the serverless mantle somewhat incorrectly, much to the commenters on Twitter’s dismay. They were basically spinning out serverless versions of a lot of the infrastructural analytical platforms where they create a managed cluster for you, whether it be EMR for big data analysis, Redshift for data warehouse, MSK for managed Kafka clusters. They started spinning out serverless versions of all those things. And most of the comments around that is that, “Hey, wait a second, you invented the term serverless when you created AWS Lambda serverless functions.” Which are pay as you go, spin up, spin down services that completely turn off when they’re not in use. These new serverless categories that they spun up at last year’s re: Invent all have idle fees. And so they are basically, auto scaling, but not truly serverless.

Muji  03:43

And so, you know, a lot of the commentary last year from Cloud Computing observers was that AWS seems a little confused. However, that’s really starting to approach when you when you look at a service, like Redshift, which is their data warehouse that is finally starting to approach where Snowflake has always been, which is separation of storage and compute.

Muji  04:07

Redshift already had added that the specific instances that separate storage and compute and now they are starting to add serverless to their Redshift, again, with base idle fees. So not truly serverless but starting to move towards where Snowflake is going.

Muji  04:27

The other thing they added last year was a data marketplace. Capability was expanded into Redshift. And so with this year’s, they kind of continued that same trend. They had one last service that that turns serverless, which is their open search, which is their fork of ElasticSearch. So this is a competitor to Elastic Cloud. Went serverless again with idle fees. $700 of idle fees, as Twitter was pointing out to me. Per month. To just sit with an empty server. So not truly serverless. But auto scaling in that it can handle higher loads and higher capacity when needed. So it can scale up and down. Bottom automatically, which is something that a lot of these other cloud platforms can already do.

Muji  05:20

So I’ve been watching the engines of Snowflake versus Redshift and of Confluent versus MSK. Of Elastic versus open search. These are kind of the competition that Amazon has been building up in its infrastructural platform. Obviously, MongoDB is in there for a long time as well with Document DB and Dynamo DB at Amazon.

Muji  05:45

And so it’s been interesting watching that dynamic. They’re moving, they’re clearly watching what the competing platforms are doing, and starting to approach what Snowflake has already built in the separation of storage and compute. In serverless, turnkey solutions, in data sharing, now data marketplace.

Muji  06:10

And at the same time, I’m fascinated by how Amazon is promoting those same companies, through its ISV [meaning ‘independent software vendor], accelerate joint co-sell partner program, where it is working with those same ISVs in order to sell all the ancillary services. If you’re not going to pick Redshift, oh, you’re interested in Snowflake? Well, guess what, we’ll work with Snowflake, to co-sell their service so that we can sell you to use AWS Lambda. You can use private link for private internet working to your Snowflake instance, those sorts of ancillary services get bundled. And so Amazon really wins both ways. They can sell you the raw compute, if you can navigate through to figure out what exactly it is you need. How to not have your budget explode when you adopt these things. Or if you’re gonna go with the turnkey ISVs, they’ll co-sell with those ISPs as well and promote their platforms. So kind of a win win for Amazon either way.

Simon Erickson  07:17

“Frienemies” instead of “BFFs”? They are frenemies with a lot of these companies, right? Big time. We can chat more about that in just a minute. But let’s stay at the 10,000 foot level. Before we drill down into that, you know, it seems like this is infinitely complex. Cloud computing certainly is a huge trend that’s out there. And it seems like also the big three, when you talk about Amazon, Microsoft, and Google who all have their own cloud infrastructure platforms, they want to be a lot of things to a lot of different customers. But ultimately, it’s the customer that’s gonna say, “This is what I need. This is how sophisticated I am. What do you have for me here?” But that said, muji, how do you do broadly generalize the cloud platforms. Is there something that Amazon, if you see, as an independent observer of this, can you differentiate between AWS and Microsoft Azure and Google Cloud Platform? Is there certain things that they are trying to be or differentiate themselves from one another?

Muji  08:10

Oh, absolutely. Amazon carved this path, obviously, with AWS. And as I mentioned before, really did invent the term serverless on demand compute. Now, they’re trying to abuse that definition, but it’s absolutely all of them provide, and are starting to focus upon data and analytics, including machine learning on top of that, security, and DevOps, you know. They’ve all got infrastructural platforms, in terms of the raw compute, you can buy the raw storage and their object stores, the private networking that you can do within them to kind of create your own cloud space for your internal applications to run within. And so they’re all fairly commoditized I’d say or have the same generalized platforms. Where Amazon really shines these days is in its custom silicon. It is has that advantage, and multi-year advantage it through an acquisition it made to create its Graviton line. It’s also got Nitro. It’s got its new training chips for machine learning. It’s got both a separate chip for machine learning, training versus inference. And so it’s really starting to carve a unique path there. Azure is absolutely paying attention and has hired that role, but they’re years behind. And so that’s where AWS is really putting the pedal to the metal in its advantage. Despite the criticisms I opened up with, which is that you’ve got all these product teams on the software side that are creating all these infrastructural products that somewhat compete with themselves, one product might eliminate you using some other product on their platform. So I fail to see sometimes a kind of a harmonious inner working between them all. But that said, it’s its advantages with silicon are massive, and it keeps iterating upon them with Graviton.

Simon Erickson  10:36

And I’m going a bit off script on this. But since you mentioned custom silicon and machine learning, I would love to hear your thoughts on Alexa? Who is listening in as I speak, as I say her name in the other room. But the focus that Amazon has had on inference and chips, you know, on the custom silicon to answer machine learning inference in terms of the queries that we’re asking in our living room. This seemed like such a big deal a couple of years ago, when Amazon had 10,000 engineers and who knows how many resources on that. But I don’t know if I speak for everyone, but I still feel like I’m using it for fairly limited questions. Where do you see Alexa and kind of kind of the home speaker or the home machine learning inference model that Amazon kind of pioneered going with this? Is this still a huge priority like it was a few years ago?

Muji  11:25

I don’t know anymore. And let’s let’s do a little test here, “Alexa, play Born to be Wild at top volume.” [laughs] Darn ah, can’t she can’t. I was gonna, I don’t know, do a drive by Alexa for you there for a second.

Simon Erickson  11:43

I hear it in the other room right now. [laughs]

Muji  11:48

So that was one where along had held Apple stock for a long time and had been a longtime observer of Apple. That was an interesting dynamic to see. Amazon Alexa versus Apple Siri, both of which got acquired and folded into the organization. And that was one where Apple seemed to really fumble the ball and Amazon got a huge lead into the living room. But that has been so underutilized. And has been such a disappointing technology all around in terms of that living room assistant, that these are Apple, oddly enough, had just released its homepod second iteration after completely discontinuing the first iteration. And so, you know, Apple doesn’t quite seem to know what it wants to do in the living room. I think Amazon got a good foothold. But yeah, beyond shopping lists, and you know, buy this thing on Amazon, it hasn’t really proved that interesting.

Muji  12:57

This is where, you know, the conversation can start veering into the generative AI, and conversational API’s that are starting to emerge, that are text based, where you can generate images from a prompt, and now full text. It’ll be interesting to see how these all merged from here. But I’m pretty disappointed by the state of things in the living room at the moment. And think that that’s ripe for disruption from here and this is where Chat GPT three or four could make inroads because of the natural language processing within them.

Simon Erickson  13:37

That’s a perfect segue to another chat topic we wanted to chat about here, which is, not coincidentally, chat. Chat GPT. We’ve certainly seen Microsoft now taking a huge investment; another huge investment at open AI chatbots. That conversational AI has been kind of the holy grail for a long time. It’s always been the question of, how are you wanting to train the data and the algorithms to respond. Microsoft tried this years ago and had certainly some problems when it released “Tay” out into the social media world. Yan LeCun is one of kind of the main researchers in this field. You know, he’s at Facebook right now. And he recently called GPT “not particularly innovative and nothing revolutionary.” He’s trying to look at the bigger world models of how we should train general AI, as opposed to something that’s just doing statistics and kind of all of the algorithms based on matrix math.

Simon Erickson  14:31

Do you have any thoughts about this? On just how to train the models and the infrastructure behind the scenes, muji?

Muji  14:36

I mean, I’m certainly fascinated by it all. I don’t have any. I don’t have the depth of research as the person you just quoted. Certainly, it’s what I think is exciting and that it’s easy to see how particular areas can get disrupted by the wave after wave after wave of these AI tools that that are coming out now. You see something like stable diffusion come out. Massively disruptive to the illustration field where you can just use a prompt, and it can go through what it’s already processed in the internet in terms of imagery to create something new for you. And that’s starting to be like, there’s a new startup that has prompt driven, artificial humans narrating videos for you. And so, oh boy, an illustration, game development, writing, it’s easy to see how these tools can be disruptive into those fields with chat GPT. There’s a whole, certainly there’s all the ethical discussions, it’s scraping the web, not necessarily applying the licensing and copyright of the imagery that it’s scraping to build up its model. So there’s a there’s a legal battle to be to be fought there. But it’s pretty exciting to see how are these going to replace illustrators.

Muji  16:04

Are these going to replace writers? Not necessarily, I, personally am more excited from an investment standpoint in what they’re doing in the code world. In terms of assisted programming, where it can go out and find examples in open source solutions to the problems that you have and can help you get, you know, 80 plus percent of the code in place in order to solve some problem. I feel like we as a society are going to move from “Oh, this AI is just going to do it for me” to being more handheld. It’s going to assist me to improve my human driven output. Like, is it going to replace me as a writer? Probably not. It isn’t gonna make me a better writer like I think it could. Is it gonna make me a better programmer? I think it could. And so that’s what I’m excited about. Personally, as a technologist, investment wise, it’s really hard to know where any of these are going to have an impact and drive profits. So I, you know, from the investment side, I’m always the cautious one, I want to see it actually create a product and drive revenue. But you see it firsthand right now. Microsoft already has GitHub Copilot, which is that assistive programming. And so, you know, these are starting to emerge now as products. And as an investor, I want to see where they go from here.

Simon Erickson  17:43

Let’s chat about that next. Because Microsoft has embraced open source in the last several years. That might not have been what Microsoft would have done in decades past. But we did see that they went on, they paid seven and a half billion dollars back in 2018 for GitHub. Open source, where developers can share code. And like you said, suggestions for how to write more effective code.

Simon Erickson  18:02

We now have another publicly traded player in this space called GitLab. And muji, I know that you don’t cover this one as closely, but I still would love to hear your thoughts on this company. GitLab went public last year at a 16 and a half billion dollar valuation. Now, we know that…I’m sorry, this was 2021. I keep saying last year as though it was last year. We’re now in 2023. It went public in October 2021, year and a half ago. But still at a valuation that was twice what Microsoft paid for GitHub.

Simon Erickson  18:33

Do you have thoughts about about GitLab versus GitHub? We know they’re not completely direct competitors. But any thoughts about those two solutions and how they’re being used?

Muji  18:42

Oh, I mean, they are, absolutely, direct competitors. The overlap is tremendous. Now, certainly GitLab maybe has a broader view of the overarching what’s called the value stream delivery platform, which could be a tool for the enterprise to manage all of its separate disparate development efforts in one platform. GitHub for Microsoft is certainly moving stronger into…well…let’s start with a 10,000 foot view.

Muji  19:16

GitHub was a fantastic purchase by Microsoft. It has become the de facto for open source software tracking. So it’s a place where you can have collaboration between all of these different programmers. There’s issue tracking. You know, the branch methodologies within the underlying source code repository, we can merge all of these different development efforts into one place and have scheduled releases out of it. GitHub really won the battle with open source. GitLab is winning the battle with enterprise options, because you could deploy it on premise. And then they started a cloud service to have the managed instances in the cloud. And you know, why would you pay someone to manage what is essentially free? It’s that you’re freeing up your IT resources. And so, we always talk about enterprise customers, which are customers that spend more than $100,000 a year. That’s one employee. If you can shave one employee from not managing the infrastructure, it pays for itself. And so I think GitLab does have a great model of going after the enterprise and overall platform vision that they have. I find them a very open company, you can actually go look at their entire roadmap of where their platform stands today and where they think it can go. And so I’ve been pretty fascinated by them haven’t specifically covered them in depth, and I don’t own them. But I think they have a lot of potential.

Simon Erickson  21:06

I think that brings up an interesting question, too, that investors are aware of certainly is the tech layoffs. Right? We’re seeing now that Meta Platforms is cutting heads. We saw Amazon cutting heads. We’ve seen basically every every large Silicon Valley based tech company is is cutting employees and full time equivalents. And you brought up a good point that a lot of these solutions, you know, are becoming more efficient. If you’re hiring somebody else who used to have somebody’s FTE full time employee, or full time equivalent doing that work, you can now hire that out. Especially something that’s open source, it’s very efficient.

Simon Erickson  21:38

But for software companies, which are a lot of the companies that you and I tend to follow, there used to be this per seat model. The subscription model, based on how many people did you have using the subscription of the platform? We’ve seen, including a lot of the companies we just mentioned earlier in the call, whether it’s Snowflake, whether it’s Confluent, whether it’s others, Splunk, wherever it is. In usage based pricing, it doesn’t matter how many people or how many seats you have, it’s how much usage are you driving out of the platform? Is this kind of the new model that we’re going to see for everything that’s built in the cloud?

Muji  22:18

It’s been pretty fascinating to watch the explosion of growth through 2020 and 2021. Now, the kind of recessionary economic impacts of 2022; seeing massive layoffs now to start 2023. It’s been interesting to compare consumption based growth, kind of subscription tier growth, and usage based growth. Absolutely. Those three different areas are going to have, you know, different headwinds and tailwinds at different times. And I think we are absolutely approaching some headwinds for usage based growth.

Muji  23:05

What’s funny, we’re just talking about GitLab. For instance, one of my hesitancies with developer tools in particular is that they are per seat, which means per developer. They’re generally not cross applicable across the entire enterprise. I’m more interested in platforms that have scale in that they don’t necessarily limit themselves to one audience. And so one of my hesitancies with DevOps tools is that it’s per developer, and there’s only so many developers within an organization. Now, are those the ones necessarily getting laid off? Probably not. So, you know, maybe that’s not quite facing the headwinds as operational tools, security tools, where you’re, you know, something like CrowdStrike, for instance, you’re buying per device, which is per laptop, which generally means per user. And so that, we’ll have to see what kind of headwinds or some of these usage based platforms are going to get from here.

Simon Erickson  24:13

Yeah, and it hits even outside of that. You know, we’re talking in terms of IT tools or cybersecurity. I mean, even outside of that, we’ve seen Netflix now even embracing an ad supported tier, right? So power users, if you’re watching one show a month and I’m watching 10 shows every day, even though we’re paying the same amount for the base subscription plan, Netflix will be making more on ad supported revenue from a user that’s a power user. It certainly is something that’s catching on.

Simon Erickson  24:42

Outside of that, let’s change gears a little bit to something we mentioned earlier, which is about the go-to-market approach. With all of the changes to the cloud and the usage based models and everything else, we’ve also kind of seen a lot of companies pivoting and how they’re approaching their own customers. It seems like there have been companies that have been going after the Enterprise. Nice dinners, get as high level of an executive into the meeting as you possibly can, and then have that permeate through the rest of the organization. Versus one that might be working with a small team of developers and let that catch on organically to build a larger following with more seats or more people embracing the solution we’ve seen. Two examples of this Okta was typically a kind of a top down, right? Like it was going up to a very large accounts with a direct sales force. Then, you know, through Auth0 and kind of acquisitions that they’ve made, they’ve tried to get into more of the developer teams. And then the opposite is true, too. We’ve seen Twilio started as a solution that was embedded and apps in things like Uberand then try to grow now into larger enterprise accounts. Now we can take this any direction you’d like to muji. But maybe my question for you is, does this work? Can you change your go to market strategy effectively out there?

Muji  25:55

Yeah, it’s the examples you gave; another one is CloudFlare. That was typically bottom up and has switched or is switching to top down to sell some of its new products. It’s difficult to say the least, if only because…like an example like CloudFlare, they disrupted the giant of Akamai in what it did, which is DDoS protection application services, CDN, in its core platform. It disrupted by going bottom up instead of top down. And so you know, now that it’s selling SaaS-y services, it can sell that directly at the CIO level, in order for an entire enterprise to adopt this next gen networking security solution. And so it has to adapt to that kind of change in the level that it sells to. And so it has to shift from not from one to the other. It’s not shifting from bottom up to top down. It’s adding both in that particular case.

Muji  26:03

And the exact same with Okta. Okta sold itself as an enterprise wide solution that you would adopt to track your workforce and identities across your employees, partners, vendors, whoever needed to access your systems. They sold two components to that platform, their workforce tracking and their consumer tracking, which means you’re going to embed Okta within your own application stack to track the users within your application. So it will be as popular one they like to call out. Everything in Adobe Cloud is tracked through Okta, it’s the consumer side about them. When they acquired Auth Zero, it was almost an admission that “Oh, wait, that consumer user tracking embedding within the application itself is something that the DevOps team normally picks not the top of the chain at the CIO level, CTO level, or whoever in the C suite is making those decisions.” And so when they purchased Auth Zero, at first it seemed like an admission that maybe we should sell this bottom up. But they got very confused, kept both around. And where selling both at the same time. Meaning to me, they were having two competing elements within their company, two competing segments that were going out to the same deals and trying to sell them. It was a really odd scenario. And ultimately, they admitted that they really fumbled that acquisition, and I think are starting to improve things. We’ll see.

Muji  28:50

But that was one where they feel like put blinders on and didn’t really adapt well. Or didn’t really recognize that, hey, these are two very drastically different go to market motions. And so it’s been interesting to watch the fumbles and stumbles of some of these public companies as they tried to do it. I think Cloudflare has been a lot more successful in selling its into the enterprise level and putting an enterprise team together. And so yeah, it’s been interesting, especially in data insecurity. When you have point solutions, it’s the developers that are making the decision about what database we’re going to embed. What is the streaming platform we’re going to embed? Are we going to go with AWS or Confluent for a streaming platform, for instance? Are we going with Redshift versus Snowflake? That’s typically, you know, head of engineering level. But bottom up, it’s the developers who are using that tool deciding to use it. Now with CIOs. As the CIO role, that’s really what they’re there for. Is to determine what’s the best platform for us to adopt enterprise wide. And so you have a lot of these decisions, especially as a company like CrowdStrike who is moving beyond just endpoint protection to become a fuller security platform that needs to be sold, you know, top down. Luckily, they’re already one that focused on top down, kind of from the start. So it’s been interesting watching these go to market motion shifts. And not everyone’s been successful, because I think of the difficulty in doing so.

Simon Erickson  30:34

It is. And it’s something I think we should be aware of, as investors, as you see the investor presentations about how much the addressable market is going to grow. Because we used to see this elite group at the top of the pyramid, but look at how many more opportunities there are at the bottom of the pyramid. That is not easy. You’ve seen companies like Okta and FireEye is another one that comes immediately to mind. If you’ve got the highest margins, the highest top of the food chain, the highest paying customers, it’s very hard to pivot from that. But certainly also harder to pivot from smaller engineering teams to something that would be embraced across an enterprise. But it’s certainly hard.

Muji  31:16

I’d even add a different pivot. Certainly, we talked about go to market shifting between bottom up and top down. But you also have a maturity of market going on with some of these companies. Now you have CrowdStrike and Zscaler, that have been sold to large enterprise. And they’re starting to move down, that kind of customer segment into just kind of other enterprise sizes, and they’ve all got their different cohort sizing charts for what they want to name these things. But they’re moving from, you know, global 2000 to the rest of global enterprises, and then smaller into SMB. And then you’ve got a company like Sentinel One that really focused on setting up SMB and is starting to move up market. So you’ve got that as an extra motion on top of these things. And so it’s been been fun to watch all that play out.

Simon Erickson  32:07

Let’s chat about that one. Let’s another topic I’d love to chat about is CrowdStrike versus SentinelOne. Both in endpoint detection; XDR is embraced by both of them. I think most people are perhaps quick to call CrowdStrike the lumbering dinosaur, and then SentinelOne the new nimble innovator that’s looking to steal share. But we did just see CrowdStrike woo two of SentinelOne’s executives, both their Chief Marketing Officer Daniel Bernard and then their Chief Product Officer Raj Rajamani. Leaving SentinelOne for CrowdStrike. Do you make anything of that move? We know that there’s people moving from company to company a lot in the IT world. But is this a big deal?

Muji  32:47

Probably. I haven’t fully processed that change yet. But yes, it is a big deal when you steal C suite employees actively from your one of your competitors elbowing you for, for your clientele, in in a couple of different fronts. And so yes, it’s been interesting to watch SentinelOne and CrowdStrike, for sure. I don’t know about the lumbering dinosaur. I’d say CrowdStrike is top of the leaderboard in selling its solution for endpoint protection, and then is expanding into sec ops and IT ops. Other capabilities, cloud security, and identity protection, in particular, are pretty exciting new directions for it. So it’s really starting to become the the Cadillac of security platforms. It’s trying to provide all of these extras, so you can just go with them. And so, you know, certainly one is obviously earlier and it’s S curve. It’s been, although is not altogether, maybe there’s like a year, year and a half between when they were truly founded and launched. So there’s not that much, or they’re both, you know, many years old at this point.

Muji  34:12

But SentinelOne’s claim is to autonomy. So, you know, CrowdStrike requires a SOC team, a security operations center, in order to run it. And SentinelOne’s claim is that it’s an autonomous solution that can find and remediate things for you. And you can supposedly use less security personnel in order to run it. But yeah, it’s that dynamic that’s at play. SentinelOne got successful because it was going after the SMBs that CrowdStrike wasn’t going after. And that was its entry into this and now was moving up market and absolutely is moving up market with some large lands. And so it’s been interesting to watch that from a security front, because of the how prevalent it is in the press. In terms of breaches, it’s an everyday occurrence with ransomware, that you have to have solutions. And this is moved up to a board level decision to be proactive with security, even though it’s not a contributor to growth and revenue. It’s an expense, that just kind of becomes a black hole, and you can continue to spend in it. So you need to limit how much you spend in it. But it’s become a board level decision to do so for risk management purposes. And so it’s gotten a lot of attention. I think there’s still a lot of success here for these platforms. But it, that poaching of employees  is pretty potent, to get them to switch much less. There’s other employees that have left for their own personal reasons of late. And so there’s been a pretty huge management shuffle on the SentinelOne side.

Simon Erickson  36:05

Muji, from a technical perspective: I know you’ve been a Data Architect for decades. Now, if you’ve got a chief technology officer or chief information security officer coming up to you and saying, “Hey, we need you to pick one or the other for endpoint security, would you recommend CrowdStrike or SentinelOne?”

Muji  36:21

Ah, I haven’t done an a product evaluation. I do the overview of platforms. And so I don’t have a particular recommendation of which one I would prefer. It’s, you know, I think they both get the job done enough. They both are scoring 95% plus on all the discoveries. You know, I think SentinelOne does have slightly better test scores. But that’s only part of the evaluation.

Simon Erickson  36:56

And then kind of the last segment I want to talk about was just kind of some companies that you generally follow. I’m not sure that I have specific questions about them. But I would like to maybe we call this a lightning round, where I just spot you up with a few and you can share some thoughts about you know, whether you really liked them, whether there’s still a best in class solution in the areas that they play in. Any direction you might take this. But the first one is Datadog. You’ve called out that you follow Datadog for several years. Where do we stand with Datadog right now?

Muji  37:25

Datadog is the what to watch in the market for. Datadog is how the clouds are doing. And they’re starting to slow a bit. And so you know, it wouldn’t be unusual for Datadog to start to slow a little bit because of it. It’s been interesting, because how tied it is to cloud consumption. And so, you know, observability, over that infrastructure that you’re creating is kind of a per host charge, really. Let’s call it not exactly usage based, but it’s capacity based charging, I guess I’d say. And so as companies, we saw it with COVID. At the start of COVID, all of these companies pull the levers within their cloud, to make requests take a little bit longer, the queues stored a little bit longer. We turned off some of our load balancing. So that process things were processed slower in applications. So those are the levers that you can pull in modern cloud architectures. That got pulled, which meant fewer containers, fewer serverless functions called smaller cluster sizes. And so that led to a few weak quarters for Datadog. But we were talking about all that. And then it rebounded, when everything kind of opened up with the market afterwards. And so we could see that same kind of dynamic here in recessionary spending, companies all pulling those levers, roughly in the same time period last six months, let’s say of trying to reduce their cloud spend, especially on the retail and consumer side of cloud services. But what I like about Datadog is that they’re really expanding, other ways to do to the billing. And so their pricing is not just per host based, it’s there is some usage based pricing, there’s some per user base pricing, depending on the product. And so they’ve really are starting to spread their their pricing in different ways across their platform.

Muji  39:32

And so when companies started to pull back spend with recessionary concerns, they saw it on the usage based pricing, which meant the amount of logs being processed in their log management system, you know, where you’re billed on the number of gigabytes per time period. And so, you know, that’s firmly usage based pricing, they saw impacts there not necessarily in the host based pricing. And so I think they’re pretty resilient of a company, in that they can handle whatever stage and level of spin companies want to commit with their cloud. And then once you’re in the cloud, as you want to modernize your applications, you need a company like Datadog to instrument your application, your application stack to see where it is now, as you make improvements where it is tomorrow, and so you need a Datadog, or a observability platform on both sides of that, much less the fact that they’re across cloud, regardless of which cloud you pick, regardless of your on prem or cloud. You’re going to use this system like Datadog. They’ve got that kind of universal reality, I guess I’d say, of of how they apply to any modern company. Every company has to have eyes on its own infrastructure, and especially in the cloud, that’s more true because you can’t, you don’t actually own the hardware at all. So Datadog provides the eyes, I still really like them as a company.

Simon Erickson  41:03

Okay, how about speaking of billing in other ways, especially if it’s volume base, how about Bill.com? Which I believe is now just called Bill. They’ve dropped the.com part off of that.

Muji  41:14

They have they’ve left 1999, behind.

Simon Erickson  41:18

The new millennium?

Muji  41:19

Yeah, we were just talking about management change, and they’ve gone under undergone a little bit of a management change over the past year or so. They’ve replaced a lot of roles. And so I’ve got some slight caution. They’re much less they’re SMB focused platform for managing cash flows. So this is a company that helps companies with their AR and AP side of things, accounts receivable, accounts payable, and managing your ongoing cash flows, and schedule of payments, in and out. And that is vital. I’d say. I think we got a glimpse at the start of COVID. March through May of 2020. Of how economic uncertainty can impact bill. And so I don’t have many fears that the you know, certainly, they’re gonna have some churn in any kind of recessionary environment where SMBs are coming and going. But it’s such a vital tool for the ones that are using it, that I don’t, I don’t think we’re going to have churn on they’re going some interesting directions with their Divvy acquisition invoice to go, they’re moving into where ramp is right now as ramp is moving towards them, which is a startup that is a competitor. And so that’s kind of one thing to watch, I’d say is that I don’t necessarily see a lot of innovation happening at Bill.com. But they are providing some, a lot of solutions for international payment handling faster ways to get vendors their money. And so it seems to be excelling at what it’s doing.

Simon Erickson  43:10

It’s a good one to think about right now, especially as we’re seeing a lot of valuation multiples contracting in the tech space companies that were previously selling at 25 or 30 times sales now selling for less than 10 times sales. But it’s also very hard to displace some of those mission critical platforms certainly seems like Bill is is one of those, you’re gonna keep around whether the economy is good, or the economy is bad.

Muji  43:32

Yeah, Bill had some euphoria there for a while it was at the top of clouded judgments, the valuation list for cloud stocks for a number of months. Last year, and didn’t it the way its growth was impacted was a little bit hidden, you had to look kind of between the lines a little bit to realize that it was float revenue was disappearing, but being replaced by the explosion of its transactional revenue as it handles payments through its cloud. And now with the the the Fed rate being increased, that float revenue is actually coming back. So it’s got some resilience.

Simon Erickson  44:14

I feel like it’s so interesting, though, right? I mean, like, if you go back a couple years, you remember Bill Gurley, one of the most famous venture capitalists in America, not wanting to be a stickler about valuations, thinking it was crazy. The companies were selling for 10 times sales. And then, you know, a couple years later, you’ve got interest rates low. You’ve got “step on the accelerator” growth. And companies like Zoom are selling for 45 times sales, with the stability of those future cash flows kind of baked into this subscription model.

Simon Erickson  44:44

But we’ve kind of retrenched, I think we’re back to a lot of these names that are best in class. I don’t know if we can call them bargains or deep bargains right now. But certainly selling for a much more attractive valuation than they were two or three years ago. Thoughts on valuation in the markets right now, muji? Especially in the NASDAQ with tech companies? Are we at reasonable valuations for long term investors?

Muji  45:09

I think so personally, with the way I invest. It’s, you know, valuation is certainly not my strong suit. And for me, it’s more about finding the quality company, which tend to be the ones that are in the top 10. Clouded judgment from Jamin Ball, his blog that tracks SaaS valuations tends to be in the top 10 are the names that I tend to cover, because they are the highest quality companies, most resilient. Their platforms are going into exciting directions from here that I think keeps the growth continuing. And so you know, that’s what I focus on is how well the companies are performing.

Simon Erickson  45:57

Perfect. And one last thing, muji…

Muji  45:59

Yeah the value is now much, or I guess those companies are a lot cheaper than they used to be, through a hindsight bias. But I think they continue to have the most potential going forward.

Simon Erickson  46:13

And one more as we close this out, you know, we’re talking about best in class companies. Maybe let’s go with the alphabetically probably the last company on most people’s lists is Zscaler. Certainly a high quality company in cybersecurity that you’re a fan of too.

Muji  46:25

It is yeah. So Zscaler is providing basically security platform over user access, whether that’s user accessing SaaS applications like Office 365, or Salesforce, or internal applications, which is two sides of their coin is protecting external SaaS apps and internal SaaS apps. Together, those are called SSE, using Gartner’s methodology, which is security edge secure. Yes, secure security edge. No, I’m missing some of them either way. But what most call secure web gateway and zero trust put together in one platform.

Muji  47:11

And so this is all about protecting user access to applications that an enterprise sanctions them to use. And they are, they really pioneered both of those directions in one platform, and are still continuing to have a lot of success. You know, they may certainly have their user seat based pricing. And so there, there may be some slight headwinds in that. But overall, I think companies realize you cannot maintain that hub and spoke model of security. You cannot. Or let’s call it castle and moat. You can’t draw a moat around your enterprise network anymore, to have a firewall protection layer that doesn’t exist anymore. You’ve got what’s called the software defined perimeter. And you have to use tools like Zscaler in order to protect your users from the malicious websites that exist out there; the malware and breach attempts. And what I like about Zscaler is that they partner very heavily with Endpoint Protection platforms like CrowdStrike, like SentinelOne, in order to have kind of a harmony between not only this user who can access this application, but you can then tie the security to a particular device and improve the risk posture overall by using these tools in tandem. So still pretty bullish on Zscaler. As well as there’s a lot of runway for the next few years as companies start adopting this next gen security.

Simon Erickson  48:55

Love it. Next gen security. Zscaler a good pick there.

Simon Erickson  48:57

As always, probably one of the most insightful hours of the year. It’s always a pleasure chatting with you, muji. Thanks for joining us on the 7investing podcast today.

Muji  49:05

It’s a dance with that lightning round! We just got to hop all around. But that was a fun journey. Thanks, Simon.

Simon Erickson  49:11

I certainly always enjoy chatting with you. And like we said, and promised at the beginning, we’re going to hop all over the all over the place for the lightning round. Muji just has such a great perspective on everything that’s going on in innovation in the tech world. You can follow along with him at hhhypergrowth.com. That’s three H’s, “hhhypergrowth.com.” And he is @hhhypergrowth on Twitter if you want to follow his analysis there.

Simon Erickson  49:34

So that’s a wrap of today’s episode of our 7investing podcast. We hope you enjoy it. We’ll have it available on 7investing.com/podcast or you can subscribe to get these delivered directly to your inbox at 7investing.com/email My name is Simon Erickson. We appreciate you tuning in!