RBC Capital analysts discuss AI monetization shifts, enterprise software resilience, and consumer adoption challenges
Summary
RBC Capital analysts argue that application software is moving in lockstep with hardware as investors seek certainty amid macro volatility. They highlight that cybersecurity and ad tech fundamentals remain strong despite valuation corrections, driven by outcome-based pricing models. The team identifies a 2027 inflection point where AI monetization flows into application software P&Ls, potentially shifting leadership from hyperscalers to names like Microsoft and ServiceNow. Conversely, they note slow consumer adoption of AI agents, citing low awareness among college students and disappointing early uptake for Meta's Muse app.
Key takeaways
- 1 Model orchestration is a critical competitive advantage; building infrastructure that supports multiple LLMs allows for cost optimization and resilience against vendor lock-in.
Unidentified speaker (speaker_3) · 14:21
Episode sections
- 2:48 – 4:18
Market Sentiment and AI Lockstep
Speaker_3 explains that application software was previously an 'anti-AI' trade but is now moving in lockstep with hardware due to uncertainty. Investors value certainty, and without it, stocks can decline continuously.
- 4:18 – 7:21
Cybersecurity Fundamentals vs. Valuation
Speaker_3 notes that while cybersecurity stocks were hurt by the 'SaaS apocalypse' and PHP vulnerabilities, fundamentals remain strong. Recent declines are attributed to valuation concerns after rapid prior gains, not fundamental deterioration.
- 7:21 – 9:45
Performance Ad-Tech and Outcome-Based Pricing
Speaker_2 highlights a shift toward consumption-based, outcome-based pricing in ad tech. Companies like Zeta and AppLovin benefit because they can prove tangible ROI (e.g., 'put a dollar in, get a dollar twenty out'), making them resilient during volatility.
- 9:45 – 11:55
AppLovin Volatility and Thesis Risk
Speaker_2 attributes AppLovin's sharp decline from highs to a 'false narrative' of thesis risk triggered by a single revenue miss. The stock recovered significantly as the market realized the miss was an anomaly and the long-term growth thesis remained intact.
- 11:55 – 14:01
Microsoft's Recovery and Misread Narratives
Speaker_3 identifies three misread narratives for Microsoft: the 'SaaS apocalypse' (vibe coding Excel), Azure growth concerns, and Office Copilot failure. All three have reversed, with Copilot engagement and Azure cash flow strength driving the recovery.
- 14:01 – 17:02
Model Orchestration and Multi-LLM Strategy
Speaker_3 argues Microsoft's decision to build a multimodal, model-agnostic infrastructure was smart. This allows them to leverage multiple LLMs and open-weight models, optimizing for cost and efficiency through model orchestration rather than relying on a single provider.
- 17:02 – 20:13
ServiceNow Consolidation and FDEs
Speaker_3 discusses ServiceNow's strategy of acquiring cyber capabilities to broaden its data moat. He introduces Forward-Deployed Engineers (FDEs) as a mechanism to help IT-constrained customers deploy AI and prove ROI, addressing the 'paradox of enterprise AI ROI'.
- 20:13 – 27:19
IBM's Evolving Thesis
Speaker_2 describes IBM's thesis as a 'broad-based bet' on AI infrastructure and consulting, allowing investors to avoid picking specific winners. However, the stock has been volatile as the market reassesses the timeline and S-curves of AI adoption.
- 27:19 – 30:27
IBM's CapEx Reprioritization and Server Cycle Pushouts
Speaker 2 notes that IBM faced a pre-announcement in early 2026 where CapEx was reprioritized and server sales cycles were pushed out due to elongated cycles and pulled-forward memory/storage deals. While software and consulting performed well, the infrastructure side faced headwinds, leading to a stock price reflection of uncertainty regarding value creation versus destruction during these pushouts.
- 30:27 – 32:43
Investor Preference for Simplicity and AI Budget Constraints
Speaker 3 observes that investors prefer simple stories over complex ones like IBM's, favoring names like Cloudflare. He highlights that CIOs face limited IT dollars and must prioritize spending amidst token bills, leading to potential budget flushes and a focus on 2027 budget alignment.
- 32:43 – 34:07
Token Budget Dynamics and Non-Human Traffic Growth
The speakers discuss how companies are balancing AI agent usage with ROI, noting that token pricing is dropping. Speaker 3 cites Cloudflare's Matthew Prince, who predicts non-human traffic will grow 1,000x in five years, shifting the paradigm for cybersecurity, ad tech, and application software.
- 34:07 – 37:29
Enterprise AI Ownership and Incumbent Software Value
Speaker 3 argues that high-growth AI-native companies continue to use incumbent software (Salesforce, ServiceNow) rather than building their own stacks, citing the prohibitive cost of building and scaling internal AI solutions. He suggests that using incumbents for agents allows companies to focus on unique value propositions.
- 37:29 – 45:26
Ad Tech Shifts: Attention, Viewability, and Agentic Buying
Speaker 2 explains that ad tech is shifting focus to formats like CTV and mobile in-app to capture attention and viewability, as bots are hard to monetize. He notes that agentic buying is increasing, creating opportunities for SSPs like PubMatic and Magnite due to compliance complexity and the rise of server-side targeting.
- 45:26 – 59:41
2027 Catalyst: AI Monetization in Application Software
Speaker 3 predicts a 2027 inflection point where AI monetization flows into application software P&Ls, moving beyond infrastructure and cyber. He expects investors to recognize that app vendors have innovated at the AI layer, leading to real monetization and growth, potentially shifting leadership from hyperscalers to software names like Microsoft.
Assets discussed
| Asset | Kind | Stance | Context |
|---|---|---|---|
| NVIDIANVDA | equity | neutral | Um, a lot of it has to do with, you know, everything in and around NVIDIA. And you know, the hardware front might be driving a bit of the train um as far as the application side of things. |
| MicrosoftMSFT | equity | mixed | And and so, Matt, I'm just trying to get a sense because your space, you you kind of lean in hard to security. You know, if you think about the Microsoft situation last week, you think about the FAA stop down. |
| AppleAAPL | equity | neutral | And now it's it's Microsoft and Apple is a two horse race. Both of those stocks, I know none of you guys care uh cover Apple, um, just astounding to me, the relative strength, you know, that they're demonstrating. |
| AlphabetGOOGL | equity | mixed | Yeah, you know, and you just mentioned Google in that quarter they just reported and obviously Amazon, that is a massively, you know, profitable business for them. |
| AmazonAMZN | equity | mixed | Yeah, you know, and you just mentioned Google in that quarter they just reported and obviously Amazon, that is a massively, you know, profitable business for them. |
| Meta PlatformsMETA | equity | neutral | Or Meta is doing a horrible job marketing it to the three billion people that are on, you know what I mean? Like, I just really like to me, I just thought that was |
“Stance” is how the speakers talked about the asset, not a recommendation.
Claims worth verifying
Software application names were previously used as an 'anti-AI' trade, gaining when AI infrastructure weakened.
Unidentified speaker (UNKNOWN) ·2:48
How to verify: Review historical performance data of application software ETFs vs. semiconductor ETFs during periods of AI stock weakness in 2026-2026. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Cybersecurity stocks like Okta, Palo Alto, CrowdStrike, and Cloudflare 'ripped' out of early May lows.
Unidentified speaker (speaker_3) ·5:34
How to verify: Check historical stock price charts for OKTA, PANW, CRWD, and NET for the period around early May 2026/2026. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
AppLovin stock rose approximately 500% from 2024 to its highs in 2025.
Unidentified speaker (UNKNOWN) ·8:39
How to verify: Calculate the percentage change in AppLovin (APP) stock price between January 2026 and its peak in 2026 using historical price data. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
AppLovin missed revenue for the first time in its history in Q2.
Unidentified speaker (speaker_2) ·9:45
How to verify: Review AppLovin's Q2 earnings report and press release for the specific year mentioned. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Microsoft's market cap was previously over $5 trillion.
Dan Nathan ·11:28
How to verify: Check historical market capitalization data for Microsoft (MSFT) to identify when it crossed the $5 trillion threshold. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Primary sourceAzure growth was approximately 40% year-over-year, while AWS bottomed at 12-13% and GCP grew 100%.
Dan Nathan ·13:31
How to verify: Review recent earnings transcripts for Microsoft (Azure), Amazon (AWS), and Alphabet (GCP) to verify reported growth rates. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Microsoft has 450 million enterprise users and almost 50 million consumer users for Office.
Dan Nathan ·15:12
How to verify: Check Microsoft's latest investor presentations or earnings reports for Office 365 commercial and consumer user counts. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Primary sourceServiceNow has approximately 30% professional services, down from historical highs.
Unidentified speaker (speaker_3) ·21:40
How to verify: Review ServiceNow's latest 10-K or 10-Q filings for breakdown of revenue by type (subscription vs. professional services). For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Primary sourceWorkday had 35% professional services at its peak.
Unidentified speaker (speaker_3) ·21:40
How to verify: Review historical 10-K filings for Workday (WDAY) to find the peak professional services revenue percentage. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Primary sourceIBM stock started at $25 in 2023/2024.
Dan Nathan ·26:16
How to verify: Check historical stock price data for IBM (IBM) for the years 2026 and 2026. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
IBM had a preliminary announcement in early 2026 involving CapEx reprioritization and a pushout of server sales cycles due to elongated cycles and pulled-forward memory/storage deals.
Unidentified speaker (speaker_2) ·27:19
How to verify: Check IBM's Q2 2026 earnings release or press releases from April/May 2026 for CapEx guidance changes and management commentary on server cycle delays. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
AppLovin had a Q2 miss in 2026.
Unidentified speaker (speaker_2) ·28:18
How to verify: Check AppLovin's Q2 2026 earnings report for revenue/earnings per share vs. consensus estimates. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Cloudflare's Matthew Prince predicted non-human traffic will grow 1,000x within five years.
Unidentified speaker (speaker_3) ·32:43
How to verify: Check Cloudflare's Q2 2026 earnings call transcript or investor presentation for Matthew Prince's comments on non-human traffic growth projections. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Cloudflare stated non-human traffic was equivalent to human traffic in Q2 2026, representing north of 20% of internet traffic.
Unidentified speaker (speaker_3) ·33:09
How to verify: Check Cloudflare's Q2 2026 earnings call transcript for specific metrics on non-human vs. human traffic percentages. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
CIOs report that much AI spending is 'net new' rather than coming out of existing software budgets.
Unidentified speaker (speaker_3) ·34:47
How to verify: Check RBC Capital's 2026 Enterprise CIO Survey results or similar third-party CIO surveys (e.g., Gartner, IDC) for 2026 data on AI budget sources. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
PubMatic and Magnite stock prices were up over 80% in 2026.
Unidentified speaker (UNKNOWN) ·49:14
How to verify: Check historical stock price data for PUBM and MGNI for the year 2026, calculating year-to-date or full-year returns. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
ServiceNow ACV hit new records in 2026.
Unidentified speaker (speaker_3) ·47:12
How to verify: Check ServiceNow's Q2 or Q3 2026 earnings releases for commentary on Annual Contract Value (ACV) records. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Primary sourceDynatrace reported consumption growth north of 20% in 2026.
Unidentified speaker (speaker_3) ·47:12
How to verify: Check Dynatrace's Q2 or Q3 2026 earnings releases for consumption growth metrics. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Okta is accelerating growth, but none of it is due to agentic capabilities.
Unidentified speaker (speaker_3) ·47:55
How to verify: Check Okta's Q2 or Q3 2026 earnings releases for revenue growth drivers and management commentary on AI/agentic contribution. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
RBC Capital uses an internal AI tool called 'Aiden AI' for summarization and diligence.
Unidentified speaker (UNKNOWN) ·51:40
How to verify: Check RBC Capital Markets internal communications, press releases, or employee testimonials for mentions of 'Aiden AI' or similar internal tools. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Primary sourceApple is a 'head scratcher' and hardware is 'table stakes'; the differentiator will be AI tools within devices.
Dan Nathan ·55:18
How to verify: Review Apple's recent earnings calls and product announcements for AI-specific software features and revenue attribution; compare with Google's hardware/AI integration announcements. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Primary sourceGoogle is in a 'penalty box' due to search and ad reasons, but Gemini embedded across its OS is 'amazing.'
Dan Nathan ·55:57
How to verify: Check Google's recent quarterly earnings reports for search/ad revenue trends and press releases regarding Gemini OS integration. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Meta's Muse app had 'not particularly exciting' early uptake numbers.
Unidentified speaker (UNKNOWN) ·57:50
How to verify: Review Meta's Q3/Q4 2026 earnings reports or investor presentations for user engagement metrics related to the Muse app. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
25 college seniors visiting a CNBC show at NASDAQ had never heard of or used the Muse app.
Dan Nathan ·56:37
How to verify: This is an anecdotal claim by the speaker; verify via the specific CNBC show episode or social media posts from Dan Nathan around the time of the recording (approx. one month prior to 2026-11-17, i.e., Oct 2026). For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
The Nasdaq has had '20% plus years' of growth over the last four years.
Dan Nathan ·59:04
How to verify: Calculate the total return of the Nasdaq Composite Index from 2026-11-17 to 2026-11-17 using historical data from Yahoo Finance or the Nasdaq website. For recent events, verify 2026-08-17 through 2026-11-17. Use another period only when the speaker explicitly names it.
Do your own research
- Verify Microsoft's Azure growth rates and Copilot engagement metrics via recent earnings calls and investor presentations.
- Check ServiceNow's Q3/Q4 2026 earnings for evidence of AI-driven revenue growth and FDE adoption rates.
- Review Cloudflare's investor materials for data on non-human traffic growth projections and token pricing trends.
- Analyze ad tech companies like AppLovin, Zeta, PubMatic, and Magnite for shifts toward outcome-based pricing and CTV/mobile in-app revenue.
- Assess consumer adoption data for Meta's Muse app and Google's Gemini integration through third-party usage surveys and app store rankings.
- Examine IBM's Q2 2026 preliminary announcement and subsequent guidance for CapEx reprioritization details and server sales cycle delays.
- Compare incumbent software vendors (Salesforce, ServiceNow, Workday) against AI-native startups for cost-to-build vs. ROI arguments in enterprise AI stacks.
Full transcriptexpand
- 0:00
- Dan Nathan: How do you position capital when geopolitics is permanently repricing risk? Is AI structural or cyclical? Or does it even matter? And with MA accelerating, where's the value actually being created? Strategic Alternatives, the RBC Capital Markets Podcast, gives investors and deal makers the insight to find the answers and stay ahead of what's reshaping business and markets.
- 0:19
- Dan Nathan: Listen and subscribe to Strategic Alternatives on Apple, Spotify, or wherever you get your podcasts.
- 0:25
- Dan Nathan: iConnections is the largest. membership-only platform for the alternative investment industry, bringing together thousands of fund managers and allocators in one powerful hub. iConnections reimagines how the industry connects through its digital offerings and premier in-person events, empowering allocators and managers to meet, build relationships, and do business anytime, anywhere. To explore more about iConnections, its events, and gain access to
- 0:45
- Dan Nathan: the members-only platform, visit iConnections.io.
- 0:50
- Dan Nathan: Welcome to the OK Computer Podcast. I am Dan Nathan. Today I have a
- 0:54
- Dan Nathan: conversation with the RBC Capital software equity research team. The guys are in town for their global TMT conference. I have Reese Drularia, I have Matthew Hedberg, I have Matthew Swanson. They cover all different subsectors within software. It's security, it's infrastructure software, it's some of the ones exposed to AI
- 1:14
- Dan Nathan: data center. We just hit all the major themes AI disruption, and which ones are using AI to benefit their customers, their own processes, and the like. So stick around. We have a great conversation.
- 1:28
- Unidentified speaker (UNKNOWN): All right,
- 1:28
- Dan Nathan: welcome back. Uh, gentlemen, this would be Matt One,
- 1:31
- Unidentified speaker (speaker_2): Matt Two,
- 1:32
- Dan Nathan: and Rishi. Yes, you're the only Rishi here. Um, uh, these guys, as I just kind of mentioned, they're in town for the their uh team. TMT conference here in New York City. And we thought we'd just do a little bit of a sit-down, kind of get the state of uh software as it relates. You guys all cover different subsectors. There's some sort of overlap in some of the themes, I I guess. And we just thought that like prior to the
- 1:52
- Dan Nathan: conference, um, and maybe we'll do a follow-up. Also, you guys have what dozens, maybe a couple hundred uh companies presenting, doing a lot of one-in-ones. You'll have hundreds of investors um here in New York, but obviously folks come in here a little bit. And maybe we'll just kind of start out with the mood. You know, we're recording this um on Monday afternoon. The market's down
- 2:12
- Dan Nathan: again. Um, a lot of it has to do with, you know, everything in and around NVIDIA. And you know, the hardware front might be driving a bit of the train um as far as the application side of things. And um, so I'd love to kind of drill in a little bit and just maybe let's start. with sentiment. Um, you guys are gonna, again, you're gonna talk to a lot of customers and a lot of companies over the next, you know,
- 2:32
- Dan Nathan: few days, but give me a sense what this market, maybe Rish, I'll start with you in a way. What is this recent downward volatility? Like, how is it kind of um being perceived by investors? Because it looks like there's I think very few places to hide right now as far as tech investing.
- 2:48
- Unidentified speaker (UNKNOWN): No,
- 2:48
- Unidentified speaker (speaker_3): I think I think that's exactly right. And you look what I think is is super fascinating. is you had a period of time in which software was especially application software, was used as the anti-AI trade. And so when you had any sort of weakness in the quote AI trade infrastructure, et cetera, it actually was helping those application names. Now I think with maybe the potential scares around it, and you know, it's been confounded by what's going on
- 3:08
- Unidentified speaker (speaker_3): in NVIDIA, is bringing out worries that this is just gonna have broader economic implications throughout the tech ecosystem. And I think that's why you're seeing everything kind of move very much in lockstep until there's a level of certainty. And I think that's something that we've seen really over the past year is investors really value certainty. And when you start to remove the floor of certainty, these
- 3:28
- Unidentified speaker (speaker_3): stocks can just keep going down and down.
- 3:30
- Dan Nathan: Yeah. Do you think though it's it's interesting, like the relative strength that some software, if you just look at the IGV, has shown, you know, last week, you know, semis and memories. memory and storage got really creamed, right? But I think the IGV closed up, you know, one or two percent um on the week. And and and so, Matt, I'm just trying to get a sense because your space, you you kind of lean in
- 3:49
- Dan Nathan: hard to security. You know, if you think about the Microsoft situation last week, you think about the FAA stop down. I mean, we start saying to ourselves there's no shortage of issues as it relates to our critical infrastructure, then throw in capabilities in and around AI. I was kind of surprised last week that some of the security names didn't trade
- 4:09
- Dan Nathan: better.
- 4:10
- Unidentified speaker (speaker_3): Yeah. Uh by the way, thanks for having us on, Dave.
- 4:12
- Unidentified speaker (UNKNOWN): Of
- 4:12
- Unidentified speaker (speaker_3): course.
- 4:18
- Unidentified speaker (speaker_3): It got hurt earlier in the year during SaaS apocalypse and then it had a really sharp recovery. Uh uh you know, a lot of times these things trade in baskets. And so uh, you know, we still think especially as organizations roll out AI in the future, security, it has to go hand in glove with what's going on. You m the other big one was the PHP invulnerability from a couple of weeks ago. And we think as these AI models get smarter, there's gonna
- 4:38
- Unidentified speaker (speaker_3): be more and more of these vulnerabilities. And so, you know, sometimes these these stocks move in in tandem with each other, but we continue to think fundamentals from some of the big cyber consolidators are are is very, very strong.
- 4:48
- Dan Nathan: Yeah, what what about sediment though? Um, you know. it seems like again, like a Powell Alto and I know the report, what, in a week from now?
- 4:54
- Unidentified speaker (UNKNOWN): Oh, this week. Wednesday.
- 4:56
- Dan Nathan: Yeah. And um, you know, it it seems like these are places from a sediment standpoint, people think exactly what you said. There's gonna be no shortage of kind of risk as it relates across the entire economy, globally, geopolitically, that sort of thing. Um, why do you think they just kind of lost some of that upward momentum? Is it a valuation thing? Because that's like a lot of
- 5:15
- Dan Nathan: these names in all of your kind of subsectors, you know, the valuations are starting to come in fairly dramatically. And if you believe the 2027 numbers, and obviously you guys haven't printed and so I'm assuming right now, you know, like this is kind of they're getting cheaper by the day,
- 5:29
- Unidentified speaker (UNKNOWN): right?
- 5:30
- Dan Nathan: So like, you know, that wasn't the case or has not been the case in cybersecurity.
- 5:33
- Unidentified speaker (UNKNOWN): For sure.
- 5:34
- Unidentified speaker (speaker_3): Yeah. I
- 5:34
- Unidentified speaker (UNKNOWN): mean,
- 5:34
- Unidentified speaker (speaker_3): you know, coming out of the early May lows of the of the market, uh, some of these cyber names just ripped, including like an Okta, Palo, Crowd, Cloudflare. And so in some instances, it was a little bit of valuation. When we looked at tier one AI winners, those, you know, the PALOs and the crowds and the Cloudflare were clearly in the tier one AI winners. And we got to this point really before three Q
- 5:54
- Unidentified speaker (speaker_3): earnings when there was a little bit of this appetite for investors to kind of move down the curve a little bit to some tier two names that could start to see some AI benefits. And and the reality is we hadn't seen those manifest as quickly as I think we thought. And so there was a little bit of a rotation back to these higher growth names. But yeah, it it's largely been this valuation thing with some of
- 6:14
- Unidentified speaker (speaker_3): these names that are that are still kind of off their highs, but you know, not cheap, not as cheap as the broad software market.
- 6:19
- Dan Nathan: What's also interesting that the growth rates for twenty 2027 twenty seven and some of the names, they're they're not like eye popping. They're like high teens,
- 6:26
- Unidentified speaker (UNKNOWN): you know.
- 6:27
- Dan Nathan: And when you think of the multiples, I mean, that's something that I I know a lot of your clients, you guys in particular, as you start thinking about 2027, are really exciting, how durable
- 6:35
- Unidentified speaker (UNKNOWN): right
- 6:36
- Dan Nathan: are are some of these, you know, growth rates, that sort of thing. So um again, not one area, you know, Matt Swanson, where it just seems like there should be no shortage of sensitivity, um, would be, you know, you're you have a focus within like digital ad space, right? And I look at some of what went on in the SaaSpocalypse as it relates to, let's say, a tech data or
- 6:56
- Dan Nathan: an app loving. And then on the flip side of that, you have a Zeta Global that um you know just reported last week and the stocks making new all-time highs, you know. And and full disclosure, David Steinberg, CEO, founder of the company, good friend of mine, been on the pod uh of late. But I bring the name up. I was with him last week, and then he reports his earnings and the stock's making. new all-time highs,
- 7:16
- Dan Nathan: you know? And so you don't find too many scenarios like that within your space here.
- 7:21
- Unidentified speaker (speaker_2): Yeah, no, I think it's a really interesting kind of byproduct of SaaSpocalypse. And one of the things we've been talking a lot about is the evolution of pricing models that go along with Gen AI. And one of the focuses that has really come up is moving towards outcome-based or kind of these consumption-based pricing models that really focus on the value. of the product that you're
- 7:40
- Unidentified speaker (speaker_2): creating. And so historically, relative to these SaaS companies that had all these warm, fuzzy metrics like ARR and CRPO deferred revenue, these consumption-based businesses around ad tech were always kind of seen as you know lesser because of the revenue visibility. When you fast forward to today, and all of a sudden, these businesses that are really built
- 8:00
- Unidentified speaker (speaker_2): around generating outcomes and especially names like you. mentioned, a Zeta, an App Love in, um, Liftoff, which was a recent IPO that had good quarter, they're all really focused on this performance ad spend. And anytime you can really go to your chief marketing officer and your CFO at the same time and say, you put a dollar in the this machine, it gives you a dollar twenty out. Well,
- 8:20
- Unidentified speaker (speaker_2): you're gonna keep putting a dollar in until you don't get that back. And so I think there is just really tangible value proposition that a lot of these companies are delivering right now.
- 8:28
- Dan Nathan: Let's talk about an app level for a second. I mean, here's here's a stock that was a monster. You know what I mean? It was just like literally was up like I don't know 500%. I want to say like from 2024 to you know the highs in 25. Does that sound right? I mean,
- 8:39
- Unidentified speaker (UNKNOWN): I
- 8:39
- Dan Nathan: might be
- 8:40
- Unidentified speaker (UNKNOWN): it might be
- 8:40
- Dan Nathan: off by a year. And um, I can't remember his name, but I
- 8:43
- Unidentified speaker (UNKNOWN): What's a
- 8:44
- Unidentified speaker (speaker_2): few decimal points
- 8:44
- Unidentified speaker (UNKNOWN): among
- 8:45
- Unidentified speaker (speaker_2): friends?
- 8:45
- Unidentified speaker (UNKNOWN): Exactly right.
- 8:46
- Dan Nathan: And and the CEO founder of that company, I I've seen him twice at this like private tech conference. It was my friend Imran Khan, who runs uh ProM Asset Management. And he was there, I want to say, in early October. And the CEO of this company put on a freaking masterclass, like two years in a row, about you know, how we built this company, the culture there. I mean, all
- 9:06
- Dan Nathan: these, you know, kind of metrics that he was solving to as he was kind of bootstrapping early on, and then how he really kind of latched into the, you know, some of the momentum that he was seeing in the space. But that stock is down what 60, 65, maybe more, you know, from its recent all-time highs. Like what's going on? And you don't have to get so specific, I guess. I'm not trying to like kind of tee you
- 9:26
- Dan Nathan: up here, but like that sort of sort of um reversal in such a short period of time, it brings me back to periods where I'm like, this is just the start of something. It's not uh what I'm saying is like broadening out a little. bit because that's a company went from a very small market cap to a very large market cap and it's on its way, you know, back.
- 9:45
- Unidentified speaker (speaker_2): Yeah, no, absolutely. And so App BLove and again, it's a demand generation platform. So a lot of the thesis that investors bring into that company is is really based on the fact of how good they are generating revenue. So you have a Q2 where it's the first miss in the company's history from a revenue standpoint, and I think for some of the newer investors, which again, like you mentioned, when you
- 10:05
- Unidentified speaker (speaker_2): see that kind of multiple expansion, like you have a lot of new investors. You have a lot of people who, you know, haven't seen the company miss before. And I think it created almost this um what right now to me feels like a false narrative of thesis risk, where people really extrapolated one data point into all of a sudden saying, you know, the TAM's too small, they're
- 10:25
- Unidentified speaker (speaker_2): too successful. essentially. They've taken over the whole TAM already. Q three, we get to, and you know, this huge dead space in between Q two and Q three, people's imaginations run wild. And they throw up, you know, the best beaten company history, the biggest raise in company history. And like you mentioned, we went from seven hundred to two fifty and now we're back to four hundred. And it and it feels like a lot of uh a lot of
- 10:45
- Unidentified speaker (speaker_2): that uh chatter is definitely quieted down.
- 10:48
- Dan Nathan: Yeah, you know, that sort of uh move I saw it was up thirty 30% percent. the day after it followed through um a little bit after those earnings and you know, after the first one day pop. I look at that sort, I mean, on a chart like that, it looks like a blip, by the way, like a 30% move. And those are the things that I don't see that price action as particularly bullish. But Risha, I want to go back to, you know,
- 11:08
- Dan Nathan: earlier this year, like Microsoft's a great example. I mean, for a bit it had that, you know, five 5 trillion plus. I think it's probably back there. Like my market cap. And now it's it's Microsoft and Apple is a two horse race. Both of those stocks, I know none of you guys care uh cover Apple, um, just astounding to me, the relative strength, you know, that they're demonstrating. Microsoft, you know,
- 11:28
- Dan Nathan: had a lot of volatility this year. At one point it was down 30 some percent from its all time highs. Now it's gotten basically um all the way back up. What do you think some of the earlier misplaced, you know kind of negativity was about the story and how it came back? It is unique among the hyperscalers. It doesn't have that consumer um exposure as
- 11:48
- Dan Nathan: we think about some of the Frontier Labs and some of the other related names. But what did investors get so wrong? Let's call it, you know, five, six months ago?
- 11:55
- Unidentified speaker (UNKNOWN): Yeah,
- 11:56
- Unidentified speaker (speaker_3): absolutely. Look, I think there are three things uh that an investors fundamentally got wrong. Number one was, and and you know, I know we talked about before the show, just hate the term SaaSpocalypse, but it's in the common vernacular, so let's use it. So when we were at peak SaaSpocalypse narratives, you were at a world where, hey, we're just gonna vibe code our own Excel, which is a terrible idea for you know, half the global economy runs on Excel. So that was number one, and
- 12:15
- Unidentified speaker (speaker_3): obviously those concerns have faded. Number two is there was this viewpoint that because Azure growth wasn't quite at the same levels or wasn't inflecting at the second or third derivative, the way their competitors. were, that they were, you know, left behind in the hyperscale cloud wars. That was always wrong, right? Microsoft was just very like deliberate about their capex, and that's why they've been cash
- 12:35
- Unidentified speaker (speaker_3): flow positive, where everyone else has been burning a lot of cash and had to actually tap the markets. And I think number three, and this is where I think the next leg of the story comes, is around Office Copilot. There was this viewpoint, even six months ago, that office copilot was completely a failure. Now look at the metrics, look at the engagement metrics, look at the net paying ads, look at the fact that we've seen the office line actually accelerate already.
- 12:55
- Unidentified speaker (speaker_3): And really importantly, customer feedback to Office Copilot is meaningfully better than it was even a year ago. And that's when Satya really stepped back into the product org. So I think those are the three things that were really holding Microsoft's stock back earlier that have now kind of deviled devilly trended trend in the opposite direction.
- 13:11
- Dan Nathan: Yeah, and aside from Azure, and and that was a big story, I guess, with AWS. and GCP, the growth rates there were like like reaccelerated in a massive way. And when you look at, I think, what was Azure up? Maybe 40% year over year, something like that. Maybe high thirties, that sound right. And and and again, you know, versus GCP, which is obviously much smaller, um, but up a hundred percent,
- 13:31
- Dan Nathan: and then Amazon, AWS, which I think bottomed out at like twelve, 12, thirteen 13%, percent, you know, year-over-year growth. So that sort of reacceleration, it kind of validates a lot of what's going on, I think, you know, in and around their I guess their AI strategy away from the infrastructure side. How do you think about there was an article in the information today, you
- 13:51
- Dan Nathan: know, when Microsoft, and maybe this kind of is something lines up with when the stock bottomed out to some degree, when they started moving away from the reliance on open OpenAI AI within Copilot, is there anything to that? Yeah,
- 14:01
- Unidentified speaker (speaker_3): absolutely. And without commenting on any specific LLMs out there, what I will say is Microsoft from the get-go, I think, made a very intelligent decision, even when OpenAI was the only game in town. And that was to build everything in a multimodal perspective so that it could be used with any LLM regardless of who it was, and even open weight models. And as we've seen more competition, And as we've seen open
- 14:21
- Unidentified speaker (speaker_3): weight models, you know, do well on benchmarks, Microsoft's been able to leverage multiple models. simultaneously, not just let's switch out one for another, but how do we actually use some of these tasks within co-pilot? And I think this is a really important bull case for Microsoft and others is that model orchestration piece, how do I pick the right model for the right job that does it in the most efficient and cost
- 14:41
- Unidentified speaker (speaker_3): effective manner? And I think that's what's starting to shake out. So it was definitely a really smart decision of Microsoft to really further embrace the fact that we are gonna going to be in a world where we will be. using multiple LLMs.
- 14:52
- Dan Nathan: All
- 14:52
- Unidentified speaker (UNKNOWN): right.
- 14:52
- Dan Nathan: So aside from Azure, when I think about you, you know, you made this point, I guess, you know, the Office line within Microsoft. When you look at like Office 365 globally, um, 450 um, you know, users on the enterprise side and and almost 50 million on the consumer side. Those numbers and the get-go, get go, I think when you were on with me
- 15:12
- Dan Nathan: maybe I want to say six months ago, you know, they were talking about like 10 million up uptake or something like that. And it seemed kind of small, but then it started, you know, it's gonna have great growth off a low number. Is that a big part of the story for you? And does it kind of isolate it a little bit to some of the other concerns as it relates to those that are more focused on consumer? Um and and so
- 15:32
- Dan Nathan: just help me think about that because um right now Microsoft feels pretty defensive in this market.
- 15:37
- Unidentified speaker (UNKNOWN): No,
- 15:37
- Unidentified speaker (speaker_3): I think that's exactly right. And so the look, Office is obviously very well penetrated in the global economy. There's still room to run in SMB and frontline and et cetera. But really, I think the story starts to become the co-pilot story. How do people get even more value out of what they're getting with the their current office suite? How do they actually drive better outcomes, whether that's efficiency, whether
- 15:57
- Unidentified speaker (speaker_3): that's better revenue outcomes, you know, just greater productivity. And that's something that customers are willing to pay for. Whether it's on a subscription fee, whether it's on some sort of a consumptive element, Microsoft is still toying with that. But I think that is very much beyond Azure. That's the next leg of the story is how do you drive that ARPU uplift within that, you know, call it half a billion
- 16:16
- Unidentified speaker (speaker_3): odd users that are paying for uh Microsoft Office today? And I think that starts to lead to further growth inflection upwards.
- 16:22
- Dan Nathan: Yeah. So how do you guys, and and we'll get a I got a lot to talk about because this is a a good segue, you know, your service now, you'll your Salesforce and Workday, and you know, a lot of Matt Swanson's names kind of interact with with a lot of these others. How do you guys think about it, the way that some of these, you know, subsectors are divided up, that sort of thing, because I know you guys
- 16:42
- Dan Nathan: coordinate a lot. So Matt, away from like say some of the cybersecurity names, like a service now, like how do you guys, how do you guys cover? it? Because there's a lot of like similarities obviously between these companies, but some of them you probably are a bit more pessimistic on than the others. And I'm just curious how like a service now fits into your coverage.
- 17:02
- Unidentified speaker (speaker_3): Well, yeah, I mean it's uh the the reality is is the the lines are blurring in a lot of software categories and and internet for that matter, right? You know, what what is an internet company, what is a software company?
- 17:11
- Unidentified speaker (UNKNOWN): Yeah
- 17:11
- Dan Nathan: why couldn't Brad before beyond you
- 17:12
- Unidentified speaker (UNKNOWN): know it? I mean like what is
- 17:13
- Dan Nathan: he doing You know what I mean?
- 17:15
- Unidentified speaker (UNKNOWN): Maybe it's an
- 17:16
- Dan Nathan: FAA thing. He's on his way
- 17:17
- Unidentified speaker (UNKNOWN): into the West Coast. We'll
- 17:18
- Unidentified speaker (speaker_3): figure it
- 17:19
- Unidentified speaker (UNKNOWN): out
- 17:19
- Unidentified speaker (speaker_3): later. Uh you know, so I mean the lines are blurring on some of these things. And and you
- 17:22
- Unidentified speaker (UNKNOWN): know,
- 17:23
- Unidentified speaker (speaker_3): Rishi and I in particular, we've had this view that sometimes different perspectives on different subsectors are kind of interesting. Um I cover Cloudflare, he covers Akamai. And so, you know, having a couple different perspectives on a category is is interesting for investors. Uh, you know, for service ServiceNow, now, uh part of it is I've, you know, have been around the hoop. for a long time on that, on that name and seen several CEOs uh over the last
- 17:42
- Unidentified speaker (speaker_3): several years. Uh and and it's just interesting to sort of see that the maturation of that company. Uh but but yeah, I mean, there there's, you know, the the lines are blurring on some of this stuff. Everybody seems like they're AI these days. Everybody's, you know, even like ServiceNow is adding cyber capabilities. In fact, Dan, like uh it wasn't long ago when I asked some investors what's your favorite cyber name? And this is when like Palo and Crowd were trading.
- 18:03
- Unidentified speaker (speaker_3): at like, you know, super, super high sales multiples. And people are like, ServiceNow is my favorite cyber idea. And it's like, okay, well, you know, clearly it's not that, but they're they are doing more cyber acquisitions. And and I think to the extent that a lot of these enterprise in Workday just bought uh an identity vendor. Um, so you're starting to see some of these enterprise software companies add in more cyber capabilities. And the answer is
- 18:22
- Unidentified speaker (speaker_3): why? Well, it because they understand that when when their customers are thinking through deployment. of Salesforce ServiceNow Workday, cyber goes hand in hand with a lot of that. So I would expect that there's even going to be more of this kind of cross-pollination.
- 18:34
- Dan Nathan: All right. So w when I look at a service now and I look at the 27 estimate, I mean the stock looks really cheap.
- 18:39
- Unidentified speaker (UNKNOWN): And we haven't seen the
- 18:40
- Dan Nathan: numbers come down like in a meaningful
- 18:42
- Unidentified speaker (UNKNOWN): way
- 18:43
- Dan Nathan: in 27. So what what would you say like the durability will be? You know, you're going gonna to have these tuck-in acquisitions as you just mentioned and are these the sorts of things that customers will recognize? Do they have the ability to upsell them, or is it more just giving them more confidence in not switching, if you will? You know, that's
- 19:02
- Unidentified speaker (speaker_3): yeah. I mean, I think that the challenge for a lot of these software vendors is if if especially if you're a system of record or ServiceNow talks about a system of action, you know, part of the defense mechanisms to prevent maybe TAM encroachment is to just get bigger data. estates. And so ServiceNow has bought a bunch of assets to I think broaden out their really their data moats that they think about. Uh,
- 19:22
- Unidentified speaker (speaker_3): and part of that cyber, part of its data. And so, yeah, I I think that you know, that's going to be an ongoing motion for not only ServiceNow, but a lot of these companies as they try to get broad. Uh, we've seen this trend towards consolidation. And I think, you
- 19:34
- Unidentified speaker (UNKNOWN): know,
- 19:34
- Unidentified speaker (speaker_3): ServiceNow is trying to do that. Salesforce has bought a number of companies recently as well. Uh I think the question. becomes, though, like you said, ServiceNow at its face value looks cheap from my perspective. There's still a lot of debate amongst the investors is that growth durable? Is it starting to gonna come, you know, is the market trying to discount in a deceler, an organic deceleration? Or
- 19:54
- Unidentified speaker (speaker_3): could there be a little bit of stabilization, if not reacceleration? And just the one thing and last thing I'll say on ServiceNow, it's an interesting development from this last quarter. There's been a number of companies, including ServiceNow. ServiceNows, that are that are starting to talk about having these forward-deployed engineers, F uh D E's, where to me it it what that tells me is that we're we're, you know, a lot of these companies have
- 20:13
- Unidentified speaker (speaker_3): converted a portion of their base to their SaaS or their excuse me, their AI suite. But we're to this point where they have to push a little bit harder to convince these customers in a IT constrained environment to deploy more AI capabilities with ServiceNow in this case. And so that'll be an interesting trend to watch. And if that actually does translate to more durable growth, then the stock could be interesting.
- 20:33
- Dan Nathan: Yeah. You know, and Rishi, I was on the FDE front, um, you know, it's something I I I was asking a CEO recently about. It was like, you know, what about those upfront costs? Is that like a really, you know, sort of expensive sort of thing if you're deploying them across hundreds of customers, probably kind of more. And they're like, no, no, no, it's it's great for us. because it actually helps us get over the hump
- 20:53
- Dan Nathan: to get these, you know, products in there, get their folks, you know, comfortable with them, that sort of thing, and the integration. And so is that like a big theme that you think is gonna like be a 2027 thing? We think about all these sort of acronyms. Is FDE gonna be one where it's basically going to give again more confidence for investors that they
- 21:13
- Dan Nathan: are gonna be able to kind of deploy? their whatever their product suites are and and not have the risk of being, you know, swapped out.
- 21:20
- Unidentified speaker (UNKNOWN): Yeah,
- 21:20
- Unidentified speaker (speaker_3): look, I think the fact of the matter is every new technological wave requires a lot of hand holding for enterprises to get through it. And we saw this actually during the early days of the SAS wave, and maybe it you know, inconceivable to think about it, but like workday at its peak was like 35% professional services. Service now is like 30%. I think it's like what less than five 5% percent of professional services today. Viva used to be like 45% professional services.
- 21:40
- Unidentified speaker (speaker_3): And so there was that recognition at that time. And I think the same thing is happening, right? Like there's still a lot of confusion in enterprise about how do we actually maximize what we can do with AI, especially because LLMs are a blank canvas and they're a blank slate. So I mean, I think that that trend has already happened. It's just that we've said, how do we up level what we do traditionally with professional services, make it more understanding
- 22:00
- Unidentified speaker (speaker_3): the customer problem, actually deploying into the actual customer environment in order to get that ROI. Because we're still at this point that even though enterprise AI adoption is off the charts and enterprise AI spending is great, there's still this paradox of enterprise AI ROI, where they're still debating okay, we've spent all this money, where are we getting the cost
- 22:20
- Unidentified speaker (speaker_3): savings that are tangible in metrics? And I think that's the other goal that the uh these FDEs are trying to at least do on more of a microscale. micro scale. How do we actually create those tangible metrics that companies can look at and justify their AI deployments? Well, it'll be interesting on this on this part, because one of the things that I've asked ServiceNow and other vendors is, you know, are we gonna be a point in 2027 or
- 22:40
- Unidentified speaker (speaker_3): 2028 where agents can just do this and we don't actually have to deploy humans to to help get some of these deals through? And so I and I don't know if those capabilities are available yet, say in a ServiceNow, service now, but I I would suspect that they're working on more agenda capabilities that could, in a more cost friendly manner, get some of these deals to accelerate.
- 22:58
- Dan Nathan: Yeah. And um, you know, one of the things we share, I just want to hit on Salesforce. So
- 23:01
- Unidentified speaker (UNKNOWN): when
- 23:01
- Dan Nathan: they reported like a month ago, and I remember the quarter, the guidance that they gave was like kind of uninspiring. You know what I mean? It's fine. But I remember sitting on the desk at Fast of Money, it was probably like five 515 fifteen or something, and they were cutting through to uh Benioff. who, you know, I give him a lot of credit no matter what they print, up, down, guide, you know, that that sort of thing. He always shows up on
- 23:21
- Dan Nathan: CNBC. Um, and you know, again, he's obviously a very promotional guy and he's a good salesman, great salesman. Um, you know, maybe not in the last year and a half or so. But when they cut to him, he's sitting there in in a conference room with Dario. And then all of a sudden the stock starts going higher and higher. And I'm just curious: like, what is the thought of some of those sorts of partnerships?
- 23:40
- Dan Nathan: How are some of these companies, these SaaS companies, gonna partner with the Frontier Labs, which might give them, I don't know, if it puts a floor to some degree from a sentiment standpoint? I'm I'm just curious like how you're thinking about that and what was your thought when you saw some of their announcements?
- 23:55
- Unidentified speaker (speaker_3): Yeah, I I think there's two pieces there. Number one, I think having these partnerships being very explicit, we saw a lot of them at Salesforce Dreamforce. for example, where the leaders of the Frontier Labs were there alongside. And I think it starts to put a little bit of a damper in the hey, uh, you know, the AI companies are gonna try to destroy Salesforce entirely, right? That there's some level of at
- 24:15
- Unidentified speaker (speaker_3): least coopetition going on here. I think the second thing, which is incredibly important, is meeting the customer where they're at. Right? We can we can have this debate are we gonna going to be in a purely headless world where everything's orchestrated via chat or natural language, or are we going to remain in this gooey interface that is, you know, at least easy to use and where you we're we can navigate? I think this should be a
- 24:35
- Unidentified speaker (speaker_3): combination of both, right? But I think Salesforce recognized early on that I think a lot of other companies are starting to recognize recognize right now, which is you have to be flexible there. You can't say we're gonna shut off Salesforce access to whatever AI model that our customers want to use. We have to have that integration. And the idea is over time you will be able to monetize that.
- 24:57
- Dan Nathan: Markets are moving faster and decisions are getting harder, but the companies that win aren't waiting for clarity. They're having strategic conversations early. RBC Capital Markets brings the insight, advice, and execution to help clients find paths to act with confidence when it matters most. From capital raises and IPOs to transformative MA. RBC's global investment banking
- 25:16
- Dan Nathan: team advises the most ambitious companies through their most defining decisions across industries, capital structures, and geographies. They know where capital is heading next because they're helping put it to work. Your capital needs a strategy. RBC Capital Markets Insights deliver results.
- 25:33
- Dan Nathan: Largest membership only platform for the alternative investment industry, bringing together thousands of fund managers and allocators in one powerful hub. iConnections reimagines how the industry connects through its digital offerings and premier in-person events, empowering allocators and managers to meet, build relationships, and do business anytime, anywhere. To explore more about iConnections,
- 25:50
- Dan Nathan: its events, and gain access to the members only platform, visit iConnections.io.
- 25:56
- Unidentified speaker (UNKNOWN): All right,
- 25:56
- Dan Nathan: Matt Swanson, you and I are gonna get away. We gotta talk as a little IBM here. Oh, because this is this is one that you know it sticks out like a sore thumb in your coverage. It was also, you know, one of those look really stealthy, at least the way through investors' eyes, you know, I I guess 23 into 24, start at 25. That was a you know, it was a bit of a
- 26:16
- Dan Nathan: stealthy sort of AI play. Maybe that has something to more do more with that sort of consulting aspect of it. But help me think about the last two years in IBM, because this last year. has been um I I would say frenetic. You know what I mean? You know, and that might be conservative a little bit, but it seems like a lot of the error has come out
- 26:36
- Dan Nathan: of the story over the last kind of six to nine months.
- 26:39
- Unidentified speaker (UNKNOWN): Yeah.
- 26:39
- Unidentified speaker (speaker_2): So if we back up to 2025, I I think one of the things people really liked about IBM was not having to choose anything about what the timeline of generative AI was going to be. It felt like it had a play for the picks and shovels from infrastructure and consulting. It had some of their own LLMs. It had a lot of the governance and
- 26:59
- Unidentified speaker (speaker_2): compliance features. It kind of felt like a way that you could play the theme without having to ultimately be right on the potential winners of the top-line application layer and still can kind of benefit from maybe like a broad-based broad based bet on the industry. That allowed throughout 2025. for people to start also kind of focusing on some of these S curves. We started to hear a lot
- 27:19
- Unidentified speaker (speaker_2): more, I got more quantum computing calls across 2025 and early 26 than people were calling about, you know, the Z series refresh. There's more people asking about some of their open source strategies, whether it be Confluent and Hashi Corp or whether it be their approach to LLMs with their granite series. And then you know if we fast forward heading into Q2, we had
- 27:38
- Unidentified speaker (speaker_2): the preliminary announcement. And really what this all came down to was it was really far away, I would say, from like what the primary thesis of a lot of new investors into IBM who are kind of playing some of those newer themes. But it went straight to kind of like the core of what IBM has always been. So we saw this CapEx reprioritization. There was a server um push out
- 27:58
- Unidentified speaker (speaker_2): basically of elongated sales cycles as people pulled memory and storage deals forward. And then really similar to what we were talking about with um AppLovin's Q2 miss, is you start to see the uh numbers come out, you see the stock reaction, and then there's this big kind of backward induction problem where people are starting with the conclusion and then working back to
- 28:18
- Unidentified speaker (speaker_2): different answers that might be different than and And not that we should always take management's narrative as like the pure truth, but it was a lot of different kinds. kind of categorizations of how server weakness could have shown up in a way that wasn't just from these deals being pushed out while something else got reprioritized. Fast forward into Q3,
- 28:38
- Unidentified speaker (speaker_2): things got significantly better in terms of the CapEx side. Software continued to outperform, which also outperformed in Q2. Uh, and then we also saw some green sheets of consulting starting to come back.
- 28:49
- Unidentified speaker (UNKNOWN): So
- 28:49
- Unidentified speaker (speaker_2): a more well-rounded quarter, but as we had into Q4 from a deal cycle standpoint, it does look more like Q2. And so I think some of those ideas around this, like broader um questions around mainframe durability maintain. And I think that's what's reflected right now in the stock price. And from our perspective, you get a software story that's better than it has been historically, a
- 29:09
- Unidentified speaker (speaker_2): consulting story that's better than it has been for the last year and a half. And then the infrastructure side really comes down. to whether or not you feel comfortable with management's uh explanation that this is still a value creation, or if you think there's value destruction during this time of the uh push out of these sales cycles.
- 29:26
- Dan Nathan: Yeah. So it was interesting that that pre-announcement, we don't see many of those anymore. We used to see
- 29:31
- Unidentified speaker (UNKNOWN): a lot back in the
- 29:32
- Dan Nathan: day. And we had a lot of companies that did mid-quarter updates, like two months in that sort of thing. And so that was a big a bit jarring. And I think it kind of shook a handful of
- 29:43
- Dan Nathan: other like, you know, names, that sort of thing. And so as we get into, you know, we're what, a month and a half left in the year. Um, you know, there's still there's still a few names in the
- 29:53
- Unidentified speaker (UNKNOWN): city. We've got the whole off order, all the off cycle. Yeah. Oracle close
- 29:57
- Unidentified speaker (speaker_3): a year out.
- 29:57
- Dan Nathan: Yeah, we gotta got to talk about Oracle in a second. But but I'm just curious, you know, when when you think about a name like that, it is a confusing story. There's a lot of moving pieces in IBM. And I think those are the sort, you know, of names that you know investors probably don't have a lot of time for, especially when you're seeing a sort of push out and you have to wait six to nine months, that sort of thing. And I'm
- 30:17
- Dan Nathan: just curious, like when you saw that print, you saw that push out, it it touched a lot of different aspects, I I guess, of the names that uh, you know, Matt, that you covered. cover also.
- 30:27
- Unidentified speaker (speaker_3): Yeah. I mean, y y Dan, you're right. Investors want simple. They don't want complex. They don't want a lot of explaining in this environment. And so, you know, you look at like a name like Cloudflare, like just a huge quarter, huge guide. Um they're uh at our conference uh this week. Uh you know, uh names like those, that continues to resonate. Uh it's it you know, know, 'cause because I think it's it from from my
- 30:47
- Unidentified speaker (speaker_3): perspective, it's part of where we're going in this AI evolution. But yeah, on the other hand, we saw other like uh Dynatrace had some deals push out. Um, you know, Service Now, you know, there's always a question in push and pull and whatnot. And so yeah, I think we're at this point of the AI cycle, it kind of getting back to the FDEs where it's like,
- 31:03
- Unidentified speaker (UNKNOWN): you know,
- 31:03
- Unidentified speaker (speaker_3): there's only limited IT dollars to go around. And if you're a CIO, you have to really prioritize what are the most important elements of that I'm doing, especially when you've got token bills. So yeah, it's an interesting time right now. And I think, you know, w as we go into our conference, as we think about the year end, we're gonna think about the opportunity for budget flush. We're gonna th think about the opportunity for how our 27 budgets aligning. And I think the reality
- 31:23
- Unidentified speaker (speaker_3): is that there's gonna be a lot of moving parts in there. But yeah, simple, the you know, we we don't need a lot of like explaining in this tape. And you can see those names get hit.
- 31:30
- Dan Nathan: So the idea of budget flush, you know, it used to be um a a thing, but now with tokens, it seems like, you know, we've gone from token maxing, token demanding. democracy.
- 31:39
- Unidentified speaker (UNKNOWN): Minimizing.
- 31:40
- Dan Nathan: Yeah, minimize. You know, it seems like if you have budget, you might actually just kind of like, you know, release the hounds as it relates to tokens again. But token pricing is coming down so dramatically. So I'm just curious, like, and maybe we can go around the horn a little bit, like how your companies are thinking you about that. Because, you know, again, if you're ti
- 32:00
- Dan Nathan: trying to demonstrate growth, and maybe we're seeing it more in the private. markets, you know, you're starting to see some companies let their, you know, their engineers ramp up again. You know, they learned a lesson, they ripped through a lot of their budgets, first half of the year for the full year. And we're seeing that, you know, Uber was one of the first ones that talked about it in the public markets. So I'm just curious, like how you're seeing, at least from the company
- 32:20
- Dan Nathan: perspective, you know, token usage as pricing's coming down.
- 32:23
- Unidentified speaker (speaker_3): I mean, it's well, yeah, part of it is I think there was that sort of that shock. some of the price tags that people saw initially this year. I mean the reality is I think everybody understands the benefit of of using um agents and using um, you know, some of these models. But but I think it really comes down to a balancing act of trying to see where the ROI is um at the end of the day. And like like one of the things, Dan, like my
- 32:43
- Unidentified speaker (speaker_3): medium term thesis, uh I I I listened to Matthew Prince of Cloudflare. He
- 32:47
- Dan Nathan: Total stud, by the way.
- 32:48
- Unidentified speaker (UNKNOWN): Yeah, yeah.
- 32:49
- Unidentified speaker (speaker_3): Yeah, he is some of the things one and this sort of it it'll tie into your question, but he he believes with within the next five years, non-human traffic will 1000 X from here. Ur on the Q2 call, they said nonhuman traffic was equivalent to human traffic and they see north of twenty percent of internet traffic. So if that is true, and I believe directionally he's right, uh there's a lot of profound changes, whether
- 33:09
- Unidentified speaker (speaker_3): it's cybersecurity, token use, uh application software, advertising, uh in that paradigm. And so when I think about how companies are thinking about deploying their own capital, I think there's a there's a lens, there's a there's certainly a paradigm shift of that's where we're going. So what are these technologies that can help get these organizations to be able to
- 33:29
- Unidentified speaker (speaker_3): thrive in this new paradigm where non-human traffic is, you know, a thousand X from where it is here? It's a fascinating conversation because I think it's wide-ranging. wide ranging. Um, but I think a lot of their CIO. CIO's decision makings come down to whether you're a company or you're a software vendor, you know, how do you how do you leverage that? Yeah.
- 33:45
- Dan Nathan: So Richie, how
- 33:46
- Unidentified speaker (UNKNOWN): how
- 33:46
- Dan Nathan: do you how do your companies inject?
- 33:47
- Unidentified speaker (UNKNOWN): Yeah, yeah.
- 33:48
- Unidentified speaker (speaker_3): Look, look, I know I I would I would agree with Matt says and I would uh I would add a couple of things, right? Number number one is, you know, even though obviously there's the immediate uh every enterprise needs to figure out how to use uh AI, as we talked about, they're just they haven't been able to demonstrate that ROI. And I think one way that they're trying to figure out how do we really maximize our token budgets, how do we maximize our AI budgets is actually
- 34:07
- Unidentified speaker (speaker_3): turning a little bit more to their incumbent software vendors for the ability of let's let them handle how do you pick the right model for the right job, how do you optimize for tokens, let's let a Salesforce or a ServiceNow or Workday or a Microsoft handle that and we'll just i invest more in there. So I think there's that element of the equation as companies try to figure that out. I think number two
- 34:28
- Unidentified speaker (speaker_3): is there almost is a certain amount of like we may even see more budget flush than normal because to a certain extent there were and and you know we're we're in the middle of doing our um 2027 enterprise CIO survey. So we'll get a little bit of a better sense there. But there might be a little bit of AI budget was a lot of net new. That's something that we've been seeing time and time
- 34:47
- Unidentified speaker (speaker_3): again that CIOs tell us that a lot of the AI spending is actually net new. It's not coming out of software budget. And so if there is a lot of unused software budget, that could be great for these companies where they could benefit from uh from that. And there is a little bit of pre-buying of AI budget and pre-buying of tokens for the following year on top of that. So that creates a little bit
- 35:07
- Unidentified speaker (speaker_3): of a double whammy effect A fact that we're paying very close attention to uh in Q4. And then maybe lastly, I I would just add, you know, how our companies are responding to this. I think there is this little bit of a paradox that companies are in, which is they don't know what to do with their stocks because they're sitting in no man's land right now. You have some of these companies trading at all time low multiples at you know, eight times, nine times
- 35:27
- Unidentified speaker (speaker_3): cash flow. But Mars and expansion doesn't save the stock. And so there is that question of like, do we just reinvest for the word that you use, durability, right? Like maybe we can't accelerate growth by. 500,000 basis points. But if we can just have a long runway of growth at where we're at today over time, maybe that's enough, especially as that AI mix
- 35:47
- Unidentified speaker (speaker_3): shift goes up. So I would just add those to what Matt said.
- 35:49
- Dan Nathan: All right. So thematically though, um, and I keep hearing this a bit, I've had lots of you know founders uh in these very seats, and we've had this conversation. But if you're like building an AI native sort of company, are you likely, you know, to kind of you know go all in on a you know a service now or a workday or a Salesforce? Are you going to
- 36:09
- Dan Nathan: start using sort of agentic tools that you build yourself that are that are coming out of the frontier labs? And I'm just curious, is that something, and again, those small companies are not going to make a dent, but if any of them are the next whatever, you know what I mean? Like they they end up being very disruptive. within their sectors. Is that something if you look out two,
- 36:29
- Dan Nathan: three, five years, that you should kind of keep your antennas up a little bit?
- 36:33
- Unidentified speaker (UNKNOWN): Yep.
- 36:33
- Unidentified speaker (speaker_3): Maybe I maybe I can start there and toss it over to Matt Hedberg. Um, but I would say two things on top of that. Number one is go look at what the uh tech stack from all these super high growth AI native companies look like. The ones that are on your newsfeed of like record hitting 100 million or 500 million of ARR, they're actually using a lot of these incumbents. incumbent software solutions with no plans to replace them anytime
- 36:53
- Unidentified speaker (speaker_3): soon. So I would add that, uh I put that as kind of piece number one.
- 36:56
- Dan Nathan: So that's a bit of a false narrative because
- 36:58
- Unidentified speaker (UNKNOWN): I hear
- 36:59
- Dan Nathan: that a bit. Yeah. And I hear that VCs, you know, some of the are are encouraging them to do that. Cause actually, in many ways, it just kind of helps their whole ecosystem, if you will. Yeah,
- 37:09
- Unidentified speaker (speaker_3): yeah. Well, and I I think the second piece though is, you know, I think the instinct might be let's go build everything ourselves. But then when you start to layer in the full cost of actually building and scaling this up, adding in support, security, all of that, it starts to get prohibitively expensive. And spoiler alert, we have a huge panel on this tomorrow morning called, you know, the
- 37:29
- Unidentified speaker (speaker_3): the the true cost of enterprise AI ownership with some VCs and private companies. I think there is increasingly going to be that realization of like, hey, if incumbent software or other company killed can build the best sales and marketing agents or HR agents or service agents, it's better to use that and focus on what makes our company unique rather than to try to build our
- 37:49
- Unidentified speaker (speaker_3): own quote AI native stack from the ground up. The only point at which I think that makes sense to do that is if there are just no good AI solutions in that category out there today.
- 37:57
- Unidentified speaker (UNKNOWN): Yeah.
- 37:57
- Dan Nathan: So Matt, like help me think about that in in your you know subsector, if you will.
- 38:02
- Unidentified speaker (speaker_3): Yeah. I mean uh listen, I think there's oftentimes I think people generalize about software. It's it there's there's a lot of idiosyncratic elements to a lot of these companies. Uh you know,
- 38:12
- Unidentified speaker (UNKNOWN): you
- 38:12
- Unidentified speaker (speaker_3): know, take a CrowdStrike versus a smaller single vent point product and a cyber solution. Well there's probably a consolidation effort there. That's not necessarily like an AI native. That's like, hey CrowdStrike can do more for me. So I'm just going to take more of that. Uh but I do I I agree with Rishi. I think I think there's, and it kind of gets back to the SaaSpocalypse, like the death of software. You know, these system of records have value.
- 38:32
- Unidentified speaker (speaker_3): Uh the question is, is like, you know, in a headless world, you know, it's an API feed, it's a database. Uh, and you know, how how long do companies willing to pay that tax? And it feels like, you know, these AI natives are are focusing on something in particular, trying to solve, not necessarily build the best CRM system that they can support or that they can run. So I I think it's going to be an evolution. Uh,
- 38:52
- Unidentified speaker (speaker_3): but like names like in cyber data infrastructure, which is a lot of my core coverage.
- 38:57
- Unidentified speaker (UNKNOWN): Like
- 38:58
- Unidentified speaker (speaker_3): I think, you know, we see a lot of that, Tarishi's point. Those those companies that I cover, like a Cloudflare, that's they're they're in these AI natives like from the start.
- 39:06
- Dan Nathan: Yeah. And and and Matt, you know, we talked about a few of those names that are down 70% from the recent highs, but they had these massive, massive runs. I mean, you know, it seems like in your space, again, we kind of alluded to this before, there's a lot of dispersion there. And so I'm just curious, like, what does 2027 look like for you? Like when you start, you know, I I know that you
- 39:26
- Dan Nathan: um, you know, you had some good movements out of some buy ratings and and and stuff like that. And you probably avoided um some of the worst of it. What are some of the themes as you head into 2027 that you think would get you more constructive on? some of these names that just you think have been just baby with the bathwater in your in your coverage?
- 39:44
- Unidentified speaker (speaker_2): Yeah, no, absolutely. And I think the two kind of like sub segments of this conversation come to like as every conversation, I guess, but like how fast is the market growing and what are the shifts in market share? And I think there's been some questions throughout this year in in both those camps of people kind of trying to dissect what's going on. So a quarter like this where
- 40:04
- Unidentified speaker (speaker_2): you know Google search is a little bit light, Amazon's advertising is a little bit light relative to consensus. And it's hard for a lot of investors to look at the open internet ad tech stack and say, you know, how can these names, um, the Magnites, Pomatics, trade desks outperform in a world where you know Google and Amazon aren't getting their mouth fed.
- 40:24
- Unidentified speaker (speaker_2): Right. And so and so I think what we're really trying to pay attention to right now is how kind of the changing consumer behavior patterns that we're seeing in in terms of where people want to spend their time, how generative search changes these things, how as Matt mentioned, a thousand times bots compared to people going on websites, really hard to monetize from an
- 40:44
- Unidentified speaker (speaker_2): advertisement to a robot, right? Or bot bot. Yeah. Um so I I think we're we're going gonna to start focusing more and more on formats. and maximizing for these two factors we always talk about, which is attention and viewability. And it seems like that's gonna be CTV and mobile in app, is what's gonna keep drawing eyeballs. And that really is kind of the cornerstone of the
- 41:04
- Unidentified speaker (speaker_2): whole ecosystem, right? You can't sell something to someone who's not watching and listening, podcasts also cornerstone of the industry. Um, but you but you go through all three of those. kind of buckets and I think we're looking for people with that format exposure. And then it really is going to start coming more and more to who owns the data and who can really set up those uh right
- 41:24
- Unidentified speaker (speaker_2): audience targetings because we are gonna see more and more agentic buying happening at the same time.
- 41:29
- Dan Nathan: Yeah, you know, and you just mentioned Google in that quarter they just reported and obviously Amazon, that is a massively, you know, profitable business for them. And so the whole idea of them kind of locking out let's say, Muse or Instinct or some of these agents, um, is this something that you think, you know, is going to be a huge theme uh over the next you know year or two, that sort of
- 41:49
- Dan Nathan: thing? Because it it seems Amazon has to defend that business. You know what I mean? It was
- 41:53
- Unidentified speaker (UNKNOWN): a
- 41:53
- Dan Nathan: fast growing business. And when you think about retail, it is a very low margin business for them. AWS, you know, they've never had more concentration despite, you know, their uh growth rate reaccelerating. But then obviously they've made some great investments in around they're building smaller models. They have training. I know you guys don't cover Amazon, but it seems like they have a bunch of
- 42:13
- Dan Nathan: bases um sort of covered. It just seems like, you know, that was a really great pillar of the bull case. And at some point they're probably gonna have to let the agents in. So I'm just curious, like if you talk about attention, um, you know, Amazon used to have a lot of your attention. If you were a Prime member and you were there, you weren't just going for one product, or if you were,
- 42:33
- Dan Nathan: you were likely to throw a handful of other things in there. And agents aren't clicking on ads.
- 42:39
- Unidentified speaker (speaker_2): Yeah.
- 42:41
- Unidentified speaker (speaker_2): You covered a lot of ground. I'll try to too. I I think part of that whole dynamic is what led to Amazon uh moving off site, right? They needed to find more inventory.
- 42:51
- Unidentified speaker (speaker_2): You've built the most powerful kind of customer-facing performance marketing tool that there was based on their e-commerce platform. And eventually it said, you know, we should start aiming this thing at other places. We'll we'll go into we'll switch prime video into ad supported. We're gonna start going off-site. We're gonna kind of create this um perpetual ecosystem
- 43:11
- Unidentified speaker (speaker_2): of bringing people in. But one of the biggest questions everybody's kind of asking right now is how does it change the way that we search for products? Because obviously the biggest piece of Amazon's ad business was always their essentially sponsored search. You type in lawnmower on Amazon, which one shows up first? And what if that is no longer the way people start that search?
- 43:30
- Unidentified speaker (UNKNOWN): Which was the Google
- 43:31
- Dan Nathan: Zero kind of uh, you know, it's the same thing over there.
- 43:34
- Unidentified speaker (UNKNOWN): Yeah. And so
- 43:34
- Unidentified speaker (speaker_2): there's there's been two really interesting conversations that have emerged from this. One is about does it create this renewed emphasis on brand spend, which had become kind of less popular because it matters so much more that I type in Nike or Adidas than I just type in Tennis Shoe. So they need me to start my search with a brand in mind. And then there's also these
- 43:54
- Unidentified speaker (speaker_2): new kind of performance promptings, which goes back to the mobile in app guys. It goes back to, I'm sure everyone's noticed your C CTV T V advertising is looking a lot differently than it. used to, where they're trying to get you to buy something right in that moment. And what they're really trying to do is upstream that whole shopping process. So before you even get into a generative search engine or anything else,
- 44:14
- Unidentified speaker (speaker_2): they're trying to actually create a moment of buying that you weren't planning on making a purchase in. And so those are two different kind of opposite ends of the spectrum dynamics of how these big brands are having to kind of reshape their. uh budgets and what's really a rapidly evolving landscape right now.
- 44:31
- Dan Nathan: All right, let's let's go around the horn. I know we don't have a ton of time left. Um what what do you think? And Richard, I'll start with you. Yeah. Um is something that was pervasive within your space for twenty twenty six. And and again, we're seeing, you know, other than a handful of names in each of your subsectors, you know, have have basically, you know, at least
- 44:51
- Dan Nathan: underperform the major industries. I think the Nasdaq is up, let's call it nine percent and the S P about um the same. What what what is something that you think, you know, um a switch gets flipped uh in twenty twenty seven that investors should be focused on where you think there's some opportunity?
- 45:06
- Unidentified speaker (speaker_3): I think it's the waterfall of AI monetization going to application software. Um the the market's clearly left most of the app names for dead uh on the uh worries on seat based models and business models and pricing and all that kind of stuff. And I think there is going to be a light bulb switch moment where investors start to say, hey, some of these app vendors have actually innovated a
- 45:26
- Unidentified speaker (speaker_3): lot at the AI layer. They're actually able to show real monetization. That's going to lead to that inflection and growth that we're all looking for. And suddenly this idea that software, the only sectors that work are cyber. cybersecurity or certain infrastructure names, I think it extends into applications in 2027.
- 45:43
- Dan Nathan: All
- 45:43
- Unidentified speaker (UNKNOWN): right.
- 45:43
- Dan Nathan: So when you were last on with me and Microsoft was, you know, not trading particularly well and it was showing, you know, relative underperformance at a time where it really felt like, you know, the hyperscalers, this was, you know, um, you know, I I guess the baton was passed a little bit. Google was trading very well, Amazon was trading well. Um what do you think is the thing in 2020?
- 46:02
- Dan Nathan: 2027 that could, you know, kind of, I don't know, knock Microsoft off of that sort of uh defensive pedestal that it's on right now? Yeah,
- 46:11
- Unidentified speaker (speaker_3): look, I mean, number one, I'd say it's probably still gonna be there just because of the duration of there. But if I were to say, like what gets it to kind of hand that baton off, uh, I think it's gonna be this just this shift of from away from just AI infrastructure, which we're already starting to see a little bit of cracks in that trade. and more towards how are we actually going to get those live knowledge worker use cases of AI that are not
- 46:31
- Unidentified speaker (speaker_3): just all custom built. And once we start to see that happen, I think we can see that baton handoff that you talked about.
- 46:36
- Dan Nathan: All right, Matt, give it to me.
- 46:38
- Unidentified speaker (speaker_3): Yeah, I think, you know, for me, I think,
- 46:41
- Unidentified speaker (UNKNOWN): you
- 46:41
- Unidentified speaker (speaker_3): know, I I I've seen clear AI leadership from five or six names that I cover where there's real acceleration, like a data dog or a Cloudflare or a CrowdStrike. or a Palo Alto
- 46:53
- Unidentified speaker (speaker_3): and a Snowflake. I guess but we've also seen a lot of forward looking metrics from uh ServiceNow talking about AI, ACV hitting new records, or uh Dynatrace talking about, you know, consumption growth north of twenty. The the question I have is when will those forward forward-looking looking metrics actually translate to improve
- 47:12
- Unidentified speaker (speaker_3): PL growth, to improve revenue growth? Uh, because it certainly feels like you know, it's taken time to develop some of these products. Uh, we need to get to the monetization. Like investors need to see that it's, you know, these forward-looking metrics are driving real revenue, either stability or acceleration. And, you know, we we've seen a number of these anecdotal data points out there, but
- 47:32
- Unidentified speaker (speaker_3): investors want to see it ultimately the PL, you know? And I think that's where like the rubber hits the road for some of these names. Until then, I think. you're going to see investors continue to want to crowd into some of these tier one AI winners that then I cover a handful of them. But to me that's one of the bigger deltas for next year.
- 47:47
- Dan Nathan: Yeah. And any names in particular that you're kind of most confident that they're going to be able to demonstrate that in, you know, the first couple quarters of uh 2027?
- 47:55
- Unidentified speaker (speaker_3): Yeah. I mean, uh, so we've liked Okta this year from
- 47:57
- Unidentified speaker (UNKNOWN): an
- 47:57
- Unidentified speaker (speaker_3): identity perspective. Uh they're accelerating growth, but interestingly none of it's because of agentic capabilities. And we think that's still coming. So when when when an Okta, who we think is well positioned in that sort of thousand X growth of non-human traffic, understanding how what these identity what these agents are doing is critical. So I I think Okta will be one that I'll that I'll be watching.
- 48:17
- Unidentified speaker (speaker_3): ServiceNow, uh just kind of given some of their ACV targets that they've talked about. Dynatrace, some of the, you know, like an observability vendor. Um, those are probably three that are kind of interesting. Uh but I think you know, as that, as that narrative expands, I think there's gonna be skepticism from investors because it's gonna just take probably multiple quarters to prove some of these things out
- 48:37
- Unidentified speaker (speaker_3): and really kind of make people feel like this is more of a thesis than a trade.
- 48:41
- Dan Nathan: Yeah. And Matt, and some of the names in your space that have been, you know, they just look like they're so oversold, but you know, bounces get sold. It's like a one step forward, two back sort of. situation. Are there some names that you're warming to um or some themes that you're warming to as you get into 27?
- 48:57
- Unidentified speaker (speaker_2): Yeah, no, it's it's been a really uh it's always an interesting market because transactional names just kind of inherently can bounce back faster, right? The tide can shift both ways. I I think one area that's become um a little bit more in vogue this year and actually has worked really well with both names being up, you
- 49:14
- Unidentified speaker (UNKNOWN): know,
- 49:14
- Unidentified speaker (speaker_2): a little over eighty. 80%, is PubMatic and Magnite and this idea that the SSP has always kind of been the secondary market within the core ad tech stack. And Trade Desk and the DSP is always kind of because they were next to the money, were kind of seen as the most durable area. I I think as we've started to see some um disruption in the types of buyers and the number of
- 49:34
- Unidentified speaker (speaker_2): buyers as we add in more agentic um buying capabilities as well, that compliance. complexity has actually created an opportunity for SSPs to get a little bit more credit for the position that they have in the market. And I think what we're also going to start to see more this year is we're hearing more about server-side targeting. So basically being closer to the
- 49:53
- Unidentified speaker (speaker_2): publisher, publishers have great data, especially areas of logged in And users. A lot of these buyers have great data and what they want more than um sometimes is is just someone to help them orchestrate it than someone to really come in and and try to do all the decisioning for them. And so I think those are names that both have, you know, a good uh position in
- 50:13
- Unidentified speaker (speaker_2): terms of benefiting from those market themes, but also from starting to see some sentiment improvement of not being kind of uh lumped in as a commodity.
- 50:21
- Dan Nathan: All right. Let's um lastly, let's go around the horn. Are you using a personal agent and are there agents within let's say, your stack at RBC that are being useful to you as far as your just kind of you know day-to-day um professional endeavors?
- 50:36
- Unidentified speaker (speaker_2): A personal agent. I I do use a personal agent. I have not found the killer app use case yet. Uh typically it's with my son to try to figure out how much his sports cards are worth when he opens them to figure out how quickly I have to take them from him. Um I in work though, I I I've really, I think started to find a real sweet spot uh of
- 50:56
- Unidentified speaker (speaker_2): um kind of separating the knowledge work all of our opinion based stuff, but utilizing agents to really help us work through the fact based part of our job, the fact finding it's it feels like a way of kind of getting a you know custom early research report when you're starting on a new theme. And then I think what we hope the value we create is much more on the context and kind
- 51:16
- Unidentified speaker (speaker_2): of weaving it into the institutional knowledge that we have.
- 51:19
- Dan Nathan: So a lot of it is kind of table stakes as far as like all your competitors or lots of folks who are doing very similar sorts of work. I mean, you're going to have these agents, they're going to be doing different processes. And really, um, does it help you, Matt, become um more productive, more pointed, gives you more opportunities to kind of really think about some of the work product? I'm just curious, like how are you using
- 51:39
- Dan Nathan: it internally?
- 51:40
- Unidentified speaker (UNKNOWN): Oh
- 51:40
- Unidentified speaker (speaker_3): well it and it's been a multi-year journey here at RBC and we started to invest early and I think we built out some really, really good technology, AI-based technology that's improving my workflow daily. I I use it, I you it's called Aiden internally and I use it every day, multiple times a day. Our whole team uses it. You know, our token bills continue to continue to go up, but it's a huge productivity tool for us because ultimately at
- 52:00
- Unidentified speaker (speaker_3): the end of the day, you know, what all three of us are doing is is we're trying, you know, we're spread so thin. And if we can find these little productivity gains of, you know, quick summarization of a transcript, you know, questions that I should be asking a private company if I don't have all the time to do diligence on it, that's incredible. Free us up so that we can talk to clients and we can actually do the stuff that agents aren't able to do. And so really
- 52:20
- Unidentified speaker (speaker_3): for us, it's a huge productivity uh tool for me. And And I I I I can't, I couldn't envision going back uh to a pre-agentic world for us.
- 52:30
- Unidentified speaker (UNKNOWN): Yeah,
- 52:30
- Dan Nathan: and and Risha, I just what Matt just said, I mean, it gives you more time to talk to clients, and that's why you guys exist in a way. And I'm just curious, like, you know, I have to assume your team right here, you guys are often traveling, visiting clients. You have, you know, some nights you'll have four companies reporting. Oh, you'll have uh yeah, that that sort of thing. And um I I have to assume it
- 52:50
- Dan Nathan: just. kind of helps streamline things fairly dramatically and i'm just curious how um you're thinking about it and then also um what what do budgets look like for tokens
- 53:00
- Unidentified speaker (speaker_3): yeah uh um maybe i'll start because you you asked uh at the beginning about personal agents and uh obviously can't use it for work but definitely use it in my personal life the one killer app I found is we fly a lot of cell sell site analysts and we all waste time going to the United app or the Delta app and refreshing seats. Having Muse do that for you every 20 minutes and notifying you the second a better seat opens up so you can have a more
- 53:20
- Unidentified speaker (speaker_3): comfortable flight is the immediate killer app that I found. So I'll just give a quick
- 53:24
- Dan Nathan: what percent of the, you know, does it actually work and and help you?
- 53:28
- Unidentified speaker (speaker_3): Oh, I mean, I don't fly middle seats, right?
- 53:31
- Unidentified speaker (UNKNOWN): Like
- 53:31
- Unidentified speaker (speaker_3): even if I book my flight late, like I'm not gonna fly Going to middle seats anymore because the moment something opens up because someone else got upgraded, I'm immediately on top of that. So it just really helps there. And that's its own um productivity enhancer. But look, I think the, you know, Matt hit on very well with what we're doing with Aiden AI internally. I'll give like what I've seen as a killer app for my team, which is these
- 53:51
- Unidentified speaker (speaker_3): analyst days that happen, right? A company puts on their user analyst day, and it's just like you're sitting in a room for five hours and you have to spend three hours writing a note that's gonna look identical to the other 40 people who are in the room at the same time. Using AI. to streamline that means that number one, we can spend more time on the ground floor at these conferences and talking to the actual partners because that's the due diligence that
- 54:11
- Unidentified speaker (speaker_3): matters. Number two, we can spend more time with clients, with private companies, with VCs, with our own public company coverage. And number three, and maybe most importantly for my own practice, spend a lot more time on the long form like reports that that I think we're known for. In the past 10 ten days alone, we was able to get a really big deep dive in physical AI that we just published this morning and another one on the
- 54:30
- Unidentified speaker (speaker_3): transformational power of AI and healthcare. I will tell you without having aid in AI to automate a lot of these things, this would have been three months between these, not 10 days.
- 54:39
- Dan Nathan: Yeah, you guys put a lot of thematic pieces out. And I think that's uh one thing that differentiates your your guys' team in general, but not just, you know, from the street. But obviously internally it kind of helps you think uh a big
- 54:51
- Unidentified speaker (UNKNOWN): it
- 54:51
- Dan Nathan: just helps you think about some of these themes that are gonna probably be some of the biggest drivers within technology over the next, you know, five, ten years. I
- 54:57
- Unidentified speaker (UNKNOWN): got a question.
- 54:58
- Unidentified speaker (speaker_3): You asked such great questions, Dan. Can I
- 54:59
- Unidentified speaker (UNKNOWN): ask you a question?
- 55:00
- Dan Nathan: Matt Swanson doesn't seem to think so. Maybe we're just like too close to each other. I don't know. It's like
- 55:06
- Unidentified speaker (speaker_3): uh you you see a lot of companies,
- 55:08
- Unidentified speaker (UNKNOWN): you
- 55:08
- Unidentified speaker (speaker_3): see a lot of entrepreneurs, you see a lot of trends. What what do you have a hot take for 2027 as it relates to tech? technology or AI or whatever it is from your purview?
- 55:18
- Dan Nathan: Yeah, I think you know, some of the defensive areas right now, and it's kind of hard to find too many, you know, like an Apple to me is just a head scratcher, and I'm not like mentioning it to you guys asking you to o opine on it because when I think about in this kind of you know new world that we live in, I mean like hardware to me is is kind of table stakes. You
- 55:37
- Unidentified speaker (UNKNOWN): know,
- 55:37
- Dan Nathan: I went and looked at the duo. um at the Apple store and I walk across the street to the Google store and I look at that flip that they have or whatever you want to call it. And it's great. They're both great. You know what I mean? Like that sort of thing. So what's gonna be the differentiator? Well it's gonna be like the the AI tools within it. And I say to myself, well Google is in the penalty box right now and it's more for like
- 55:57
- Dan Nathan: search and ad sort of reasons. But you know, Gemini embedded across that operating system is amazing. Yeah. You know what I mean? Like so to me, I just think that there's gonna be um I think there's gonna be a sorting out of the way consumers use it. And I think we're probably, you know, consumers right now, like think about how many people in your life that are not in
- 56:17
- Dan Nathan: finance, not in tech, not early adopters, they're just not using these things. For sure. I I had this is a really fascinating sort of thing, or at least is to me, and maybe you guys will um find it equally interesting. So I had 25 seniors in college visiting, you know, the show that I do on CNBC at NASDAQ. It was about a month ago. I know a lot of our listeners are probably
- 56:37
- Dan Nathan: sick of hearing I've used the example on a few occasions. Um, but we're talking about how they are using AI for academics, you know what I mean, but also more importantly, job search, right? They're like getting to the end of their first semester of their life. last year. And I said to them, Yeah, you guys all use LLMs to cheat on your papers and this and that, whatever. We're all doing that,
- 56:57
- Dan Nathan: right? How are you guys using agents to help your workflow as far as looking for a job? This is the next really important thing you're gonna do. And not one of them raised their hand.
- 57:06
- Unidentified speaker (UNKNOWN): Interesting. Not one of them,
- 57:08
- Dan Nathan: and this is a month ago, okay. So this is like um, you know, a month into or month and a half into Muse, they all spent spend two hours a day on Instagram, you know what I mean? That sort of thing. And I was just shocked, not one of them. And then I said to myself, Oh, yeah, well, it's early, you know what I mean? And then I asked a few other kids who I know just specifically in different circumstances, um, have you
- 57:28
- Dan Nathan: used, have you downloaded Muse yet? What's Muse?
- 57:31
- Unidentified speaker (speaker_3): Wow.
- 57:32
- Dan Nathan: I I was actually like shocked. I'm like, so that's either these kids are not particularly curious and they're not reading about what's going on in technology.
- 57:40
- Unidentified speaker (UNKNOWN): Yeah.
- 57:41
- Dan Nathan: Or Meta is doing a horrible job marketing it to the three billion people that are on, you know what I mean? Like, I just really like to me, I just thought that was
- 57:50
- Unidentified speaker (UNKNOWN): really
- 57:50
- Dan Nathan: surprising. So either agents are gonna be like the metaverse for consumers, or it's gonna just absolutely take off. And when we saw those Muse numbers when Meta reported, it was not particularly exciting, and maybe that reached really feels a little bit about some of that early uptake of co-pilot. You know what I mean? But but again, I I don't know. To me, I think there's gonna be a
- 58:10
- Dan Nathan: great sorting out of a lot of this stuff. Does that make sense?
- 58:13
- Unidentified speaker (UNKNOWN): It sounds like
- 58:13
- Unidentified speaker (speaker_3): universities need FDEs for these students.
- 58:15
- Unidentified speaker (UNKNOWN): Yeah, I mean, to
- 58:16
- Dan Nathan: some degree, you know, and and maybe Apple, you know, the way they kind of made inroads in education, you know, this is going back 30 some years ago, maybe Siri, because you know, like. every kid's got an iPhone. They're not using Android in college for the most part. I think that's their demo. Maybe there is an opportunity for small models and working within the hardware
- 58:36
- Dan Nathan: that you're never going
- 58:37
- Unidentified speaker (UNKNOWN): to get
- 58:37
- Dan Nathan: rid of. I I don't know. But Apple has not demonstrated that to us yet.
- 58:41
- Unidentified speaker (speaker_3): Yeah. It'll be fun. It'll be fun to watch, especially under new leadership.
- 58:44
- Dan Nathan: Yeah. Listen, gents, let's do it again in Q1. We can kind of take stock a little bit of, you know, kind of the end of the year. I know that you guys, you know, when you have the opportunity to have a conference like this, the ability to talk to companies, to talk to investors, kind of meet in the middle and see where folks are as far as sentiment. It feels like right now, this is obviously not going to be one of those
- 59:04
- Dan Nathan: 20% plus years that we've had in the Nasdaq over the last, you know, sort of four years. I think that's kind of what it's been. Um, and it really does feel like there's gonna be a lot of opportunities to find unusual values um on the long side, but maybe avoid some of the big pitfalls, if that makes some sense. Yep. Absolutely.
- 59:21
- Unidentified speaker (UNKNOWN): That's the
- 59:21
- Unidentified speaker (speaker_3): goal. That's the goal. All right,
- 59:22
- Dan Nathan: Jen. Thanks,
- 59:23
- Unidentified speaker (UNKNOWN): I appreciate it. Good
- 59:23
- Dan Nathan: luck in the next couple
- 59:24
- Unidentified speaker (UNKNOWN): of days. I know it's
- 59:25
- Dan Nathan: gonna feel like a sprint, so thanks for being here. Thanks
- 59:27
- Unidentified speaker (UNKNOWN): for having
- 59:28
- Unidentified speaker (speaker_3): us.
- 59:31
- Unidentified speaker (UNKNOWN): This
- 59:32
- Dan Nathan: podcast is for informational purposes only. All opinions expressed by me, Dan Nathan, Guy Dami, Dani, and any other participants are solely our opinions and should not. be relied upon for specific investment decisions.
More from RiskReversal Pod
Nvidia’s Next Problem: An Avalanche of Wafers + FinTech for Foodies with inKind's Johann Moonesinghe
Episode discussion: NVIDIA faces regulatory and margin risks while InKind scales its restaurant credit marketplace
Speakers’ claims: Speakers analyze NVIDIA's post-earnings performance, citing board entrenchment, potential export control violations, and margin compression risks from inventory buildup. They question whether the market is underpricing these geopolitical and operational risks. The episode then shifts to InKind's CEO Johann Moonesinghe, who explains his two-sided marketplace model that finances restaurants by purchasing future food-and-beverage credit. Moonesinghe details his transition from failed restaurant investments to building a profitable fintech platform with a 2.7% loss rate. He highlights InKind's recent $750 million debt facility and projected revenue growth, emphasizing the company's competitive moat and unit economics.
There’s a New Kid in Town: Nvidia, Anthropic & the AI Unwind
Episode discussion: Hosts discuss Nvidia valuation, Anthropic IPO, and broad AI trade unwind amid leadership changes.
Speakers’ claims: Guy Adami and Dan Nathan analyze Nvidia's recent price action and valuation, noting the stock traded near support levels despite high growth expectations. They highlight the leadership change with Jensen Huang's departure and cite a Wall Street Journal report on internal concerns and circular financing involving neo-clouds. The hosts connect the AI sector's weakness to rising Treasury yields and Bitcoin's drop, while discussing Anthropic's massive IPO. They draw parallels to Cisco's 2007 collapse and note a broad unwind in semiconductors, leaving only large-cap tech stocks as safe havens.