AI will not kill software – but vendors must pivot to survive | Transcript


Matt Hedberg

The punch line is Software is not dead. We think the narrative is evolving. We think the definition of software is changing. We think the AI and agentic era is ushering in one of the biggest changes in tech that I've seen in my 20-year career on Wall Street. We think the next several years will be critical for software companies to evolve, to pivot, to change the way they think about delivering tech, change the way they think about value, ROI, pricing. We think there will inevitably be winners, there will certainly be losers. But make no mistake, software is not dead, it's just evolving.

[MUSIC]

Nate Mahrer

Welcome to Strategic Alternatives, the RBC Capital Markets podcast. I'm your host for this episode. Nate Mahrer, Technology, Media and Telecom sector strategist. Today we're discussing our latest Insights report, Thinking Through the AI Software Narrative. We wrote it to unpack a growing disconnect in the market, where AI continues to dominate headlines and drive enthusiasm at the index level, yet much of the software sector has struggled to keep pace. Over the past year, investors have gravitated toward the more tangible parts of the AI story: semiconductors, hardware and infrastructure – areas where demand, earnings and capital intensity are easier to see in value. But by contrast, software has been slower to benefit, as investors wait for clearer signs of monetization and ROI.

Joining me is Matt Hedberg, Global Head of TIMT Research at RBC. Together, we'll look at where sentiment and valuations could turn, what's defining leadership in this platform shift, and where we're seeing deal activity take shape. Matt, welcome to the podcast.

Matt Hedberg

Thanks, Nate. Great to be here as always. 

Nate Mahrer

So, let’s start with the market background. The NASDAQ is near all-time highs, as AI is driving massive market enthusiasm, yet software stocks, broadly, have lagged. Investors have favored picks and shovels benefiting from the AI build, as well as the hyperscalers themselves. And number of these names have seen sizeable estimate revisions, a few have seen re-ratings on favorable narrative shifts. But by contrast, software has been a mixed bag at best. Can you walk us through some of the narrative overhangs, Matt, for the software sector to help frame the discussion?

Matt Hedberg

Sure, yeah. It's something that we talk about every day with investors. And it's been a huge focus recently in particular. Software has really been bifurcated this year. There are certainly categories that are benefiting from AI, like consumption software, specific cybersecurity names, as well as vertical software, while others, most notably, horizontal software and aspects of the infrastructure stack have struggled. 

With the increased spend on picks and shovels of the industry on the semi-cap side and other areas of the infrastructure build-out, we think it's a matter of time before software starts to see benefits from this. And like other definitional changes of software, like on premise to cloud, not all companies will win. Some will be left behind. And we think part of our job is trying to identify some of these categories and individual names that can ultimately benefit from this, and that is, I think, where investors are focused on: is, when will we see some of these tangible benefits, and what ultimately changes the sentiment on software?

Nate Mahrer

So, Matt, you touched on this in the report. Tech does go through these transitions. Within software, how does this remind you of some of the other technological shifts we've seen?

Matt Hedberg

Yeah, we've noted before. We think this is maybe the fourth technological shift in technology. Obviously, cloud being one, internet being one, SAAS being another. We think this is a major tectonic shift in technology, and one that is happening at a very accelerated pace. When we think about which categories or which management teams have the advantage, you can find some really profound answers right now in terms of who's thinking about this aggressively, who's thinking about changing and pricing and packaging. Who's thinking about, ultimately, how do you deliver ROI to the end customers? It's going to be easier in some subcategories, like cyber, like consumption names. But again, I think as some of these, primarily seat-based, horizontal software models need to evolve. That's where we'll see winners and losers, just like we saw in prior definitional changes, like on premise to cloud, there were winners, there were losers, there's new names that come to market. I think that's part of the fun of software, is that the definition is always changing, but the category continues to be one of the biggest value creators in tech over the long term.

Nate Mahrer

So, there's a lot of disruption here. There's a lot of discussion about this current narrative overhang on software. How much do you think of it is justified? You've mentioned the cyber, you mentioned the consumption names, verticals, potential beneficiaries. Do you want to just talk about the risk of falling behind and just sort of how you're thinking about the justification for this narrative?

Matt Hedberg  05:00

It's sort of like guilty until proven innocent, is what we hear from a lot of investors. And candidly, when investors think about the broader TIMT landscape, it's just been easier to invest in picks and shovels right now. Now, there's a few winners. We publish in our report a handful of names that we think are consensus longs within the AI space, but the vast majority of names are still very, very much under pressure. And I think the challenge again is understanding what changes that narrative. And I think you know, increasingly, as end customers continue to increase their own AI or agentic framework, some of this will come into sharper focus over time, as AI budgets are built out, and we don't see the AI crowding effect that we have historically within IT budgets. And so we do think that, the challenge is that this space is moving so rapidly, and I think a lot of end customers are still uncertain about what the AI or agentic future holds. But as these become more evident, and we're seeing that with some of the data warehousing or data lake vendors, we see that as evidence that customers are preparing their data for AI. The next phase becomes, how does this AI data make its way into production? And ultimately, then, how does that benefit the broad software food chain? Again, there's some categories like cyber security. These things are tip of spear, where it's always a hot topic, regardless of it’s AI or agentic or not. And we think in many times as companies think through an AI digital transformation, they often have to think about a cyber transformation as well. And so we're seeing some early benefits from some of these cyber names. We're seeing some of these from some of the vertical names that have more specific data to their names, and they can leverage that data for advanced AI capabilities already. Ultimately, I think the market is broadly searching for what raises the narrative on this. In other words, for spending all this capital on GPUs and chip build out and infrastructure, when does that flow through to the software layer? That's ultimately the million-dollar question. 

Nate Mahrer

So let's double click into some of this. The report outlines three tiers of AI leadership, current leaders, those next in line, and some emerging contenders. Matt, what do you think separates these groups?

Matt Hedberg  08:06

Yeah, so. Thanks, Nate, we, you know, in the report we call out three tiers of names. Tier one would be those that are seeing very, very early tangible benefits from AI, and a lot of that ultimately shows up in multiples for these stocks. Folks that are enabling AI, whether it's either software or services, those are the tier one names. It's kind of this tier two or tier three names that could see even more significant multiple expansion. We outline a number of names in the tier two category that we think have the right ingredients to be AI winners longer term but aren't considered that thus far from either a market sentiment or even a market multiple perspective. There's quite a few large cap names, some mid cap names that we think kind of span cyber infrastructure, vertical software that all have the right ingredients of market leadership, TAM, architecture. And really, I think, leaning into AI to justify moving up to that tier one category. There's dozens that we called out in the tier three category that are hotly debated, submarket multiples, very controversial names. This is the more murky or brackish area of software where it's hard to determine yet. The market's not calling them winners, but we do think that, through a lot of the analysis that we provide in the report, have the right characteristics of meeting the criteria to become AI-relevant. And so I think that's the category, this kind of tier two, tier three, that are really interesting to me. I think the tier one names likely remain compounders in the space, but this is the category that I think as we look to the back half of this year, and 2026 and even 2027 will be super interesting. And I think some of these tier three names could become increasingly interesting from an M&A perspective, right, as bigger companies think about how they want to try to advance or bolt on some features to their broader stack, you could start to see increasing M&A take hold in kind of this tier three camp AI vendors that we talk about.

Nate Mahrer  10:34

Yep. And also we've seen the look the market has penalized companies it deems to be on the wrong side of the narrative with multiple compression, leading investors that think these could be value traps. But have we seen plenty of times, narrative changes can and do happen. So why don't we pivot and talk to a few key themes of this report? Let's talk about monetization, which feels like one of the most important pieces in how software companies can come back in favor with investors. We've said in the report that AI monetization is still in the very early innings, and that while capex and R&D costs are easy to quantify, the narrative around value creation has been much more fluid. How are software companies actually starting to turn AI investment into revenue? And what should investors be watching for next, Matt?

Matt Hedberg  11:16

Sure. This is the really interesting part of the software landscape is, where do we go from here. And candidly, I think we're taking a lot of cues from some of the AI natives, some of the frontier model building companies that – most are private at this point. They've really built and been built in this AI era. And, we'll hear oftentimes from VCs and early stage investors that investing in some of these companies that are a year or two old is easier than investing in companies five or six or even 10 years old because they've been built in this AI era. And that means a lot of things, right? It means there's this extreme land grab right now going on in terms of sentiment with insight of end customers and thinking about who's going to lead the charge here, in this AI landscape. When we think about how these companies are being built, they're often being built not on a seat based model, but more of a consumption or an ROI or a value based pricing mechanism. And it's easier for these smaller companies to make that cross the chasm and think about pricing and really ROI differently. And so we think as these bigger companies think through this evolution, things are going to have to change. We're going to have to think about pricing and packaging differently. We're going to have to think about the rate and pace of innovation differently. And one of the questions that I find interesting asking in management teams is, how do they think about moats in the future. Historically, companies used to think about moats like software and code, but in the future, there's a reality of where all software could be open source. And so, how do these management teams think about differentiating themselves, staying in front of competition, leapfrogging competitors that are slow to adopt AI and also not necessarily losing out on that mind share to some of the early AI natives. It's a complex algorithm that's happening right here. These bigger companies certainly have the advantage of customers, capital, data, scale, longevity. We think a number of these companies will evolve to make this transition to AI, but this land grab phase is certainly, it's evident in many aspects of the tech stack. And like you said, Nate, there's been a lot of head fakes early on, some names were considered AI winners only to become AI losers, and then only back to the winner category, and vice versa. And so, I think that's one of the challenges too, is that, as we're so early in this, the head fakes that investors are seeing is hard for them to think through, which ultimately leads a lot of investors that we talk to to stay very concentrated in their software bets. Some of the mega caps, you know, are very much overweight, and that's where we feel like a lot of people are placing their bets today. But increasingly, we think, as we as we follow the tea leaves and we understand and we think through some of these winners that aren't in the mega cap category, that's really where some interesting value we think could be created.

Nate Mahrer

And so, as pricing and usage models evolve, how do you see AI changing the relationship between software vendors and their customers?

Matt Hedberg

This is the challenge, right? We hear a lot from customers. I'm working with a particular software vendor, and I have 1000 seats, and I'm paying them a million dollars a year. They're coming to me saying that million dollars should turn into a million five, and then two, and then three, that expansion in commitments to some of these vendors, but at the same time, these customers are saying, well, I might not need 1000 seats in the future. I might need 900 because I'm more efficient with the workflow that they're enabling. How these vendors talk to their customers to show that they're delivering tangible ROI that may not be tethered to seats. That is the opportunity. Theoretically, even if seats were to compress, the money being spent with these vendors could go up if the consumption element more than offsets for a reduction in seats. I think it's part of a hybrid strategy. And ultimately, a lot of these vendors that are thinking about a hybrid strategy are thinking even longer term about, how do we make the complete jump to more consumption value, or ROI based pricing, that is a challenge, but I think also it's an opportunity, especially if these vendors can deliver real, tangible ROI. Because it's very much the Wild West. Who owns the origin agent orchestration level? How do you get work done across multiple systems. It could create a tremendous amount of value for some of these software companies. But again, some of these other companies that don't have that forward thinking in terms of how do, they're going to satisfy their customers from a pricing and packaging, they're going to be left behind. And there's a lot of these vendors that ultimately might end up going the way of private equity and consolidation and repackaging outside of the public markets, but it clearly is a very, very early, volatile time, and pricing and packaging is certainly part of the discussion.

Nate Mahrer

So, this then gets us into agentic AI, where systems act autonomously rather than simply assisting users. How do you see that reshaping the software landscape and influencing enterprise budgets or business models going forward, Matt? 

Matt Hedberg

This is the most exciting element, this gen AI or agentic era, is thinking about how agents will interact with humans. And there's a lot of questions and concerns that we get from, you know, just the broad population about, is AI coming from my job? What does the workforce look like in the future? And obviously, there's going to be a lot of changes in the workforce, but we've said for a long time, I think employees that leverage AI will be more disruptive than AI of itself. And I think when we think about agents and knowledge workers, Nate, like you and I, using multiple agents in the future to get our work done. Today, internally at RBC, we're using AI to enable workflow for us in capital markets research, and it's really been a really exciting time in my career to understand how agents can supplement employee workers, where an agent is not just an intern, but more of a colleague or a coworker, and it's this autonomous work with multiple agents that's really, really exciting, I think, for the future. But how do you think about deploying those agents, managing those agents, securing those agents, understanding what these agents are doing, it's going to be a critical component. And, Nate, if you're like me, and you've worked with a lot of enterprise software, a lot of the UIs of this software is not the most intuitive, it's not the greatest, but we think with an agentic front end, it could even improve the end user experience to make it a much more profound experience. I think employees and organizations that embrace this will be the disruptors in the core economy. It's more evident to me in some of the cyber names, I think some more of the agent orchestration, that is still up for grabs. It's a land grab. It's a very valuable piece of real estate. I think some of the hyper scalers, some of the broader platform players, will all benefit from this, but, but this is the area that I think is still evolving and a little bit TBD and causing a lot of investor skepticism.

Nate Mahrer

Matt, you and I are not going to host a podcast on poor user interface on enterprise software, although I'm sure we could fill the time with one. But let's turn to the impact on deal making. Do you see M&A likely accelerate as a direct result of AI? 

Matt Hedberg

Totally. I mean, yeah. Nate, you and I talk about this every day with investors, and we're seeing an increased cadence of M&A activity show up in the headlines. Just this week alone, there were a number of instances of potential M&A in our space. I think, M&A is going to be multifaceted. I think there will certainly be elements of large hyperscaler or large software platform vendors looking at smaller AI startups as a way to stay in front of their peers. I think that you'll see another wave of M&A of perhaps even just buying ARR or buying an executive team or buying a customer base, and that could be both strategics and private equity. Private equity could very well be in a position to stitch together a number of software assets and ultimately bring that back to market in the form of an IPO. And so I think one of the challenges in software is there's a lot of software companies that probably don't need to exist in the future and just don't know it. M&A will certainly take care of some of that, both strategic and private equity, but there'll be a lot of software companies that are going to be effectively left without a dance partner. There's not a conversation that goes by with investors that don't focus on M&A, and I think that only will accelerate from here. And candidly, that could crowd out some of the, some of the more legacy software vendors, which could lead to further consolidation. So, yeah, I feel like it's this flywheel of M&A and IPOs and consolidation that's going to be an ongoing narrative for the next several years.

Nate Mahrer

So as we wrap up, we've seen some software names move from being perceived laggards to leadership stories almost overnight. What can we learn from that, Matt, and what does it say about long term software leadership in an AI-first world?

Matt Hedberg

Yeah, this is – we write on this in the report extensively, and I think it will be something that we have to continue to evolve our thinking. And so I think staying nimble is critical. I think staying close to management teams is critical understanding how they think. Because ultimately, the end of the day, where we get a lot of conviction is, is this a management team that we want to bet on? Are they looking to where the proverbial puck will go? And so I think staying nimble is important. I think thinking about tech more broadly, even beyond software, is critical for investors, because we're on this podcast talking about software, but AI is a much more, broader definition than just software. And so I think thinking holistically in terms of picks of shovels, software, services, it's going to be critical for investors think about AI even more holistically. And then there's a whole set of AI names that are outside of tech. We think about, like, cooling and power and data centers and infrastructure, all that stuff, I think is part of kind of how we think about the future. So at the end of the day, I think we've seen tech go through evolutions, and I've been in this seat now for 20 years at RBC, and I've seen many, many transitions. I think this one is one of the most profound that we'll see. I think picking big markets that can get bigger, where AI can be an inflationary element versus a deflationary element is critical. And then I think, finally, betting on the right architecture, an architecture that is built and can scale in an AI era that's not as much tethered to seat-based counts, but more consumption, more usage, more value, that too will be critical. 

Matt Hedberg

The punch line is: Software is not dead. We think the narrative is evolving. We think the definition of software is changing. We think the AI and agentic era is ushering in one of the biggest changes in tech that I've seen in my 20-year career on Wall Street. We think the next several years will be critical for software companies to evolve, to pivot, to change the way they think about delivering tech. Change the way they think about value, ROI pricing. We think there will inevitably winners, there will certainly be losers. But make no mistake, software is not dead, it's just evolving.

Nate Mahrer

Love it, and that's a great point to end on. Thank you, Matt.

Nate Mahrer

Thank you for listening to Strategic Alternatives, the RBC Capital Markets podcast. This episode was recorded on October 17, 2025. Listen and subscribe to Strategic Alternatives on Apple, Spotify, or wherever you get your podcasts. If you'd like to learn more or continue the conversation, please visit RBCCM.com/strategicalternatives or contact your RBC representative. See you next time.

[DISCLAIMER]

This content is based on information available at the time it was recorded, and is for informational purposes only. It is not an offer to buy or sell or a solicitation, and no recommendations are implied. It is outside the scope of this communication to consider whether it is suitable for you and your financial objectives.