A few years ago, cloud computing was the hottest story in tech. Before that, mobile was the buzz.
Now, artificial intelligence is the new frontier. At least, that is the message from every pitch deck, press release and product demo on the market.
For private investors, cutting through the noise is not easy. But a new report from RBC Capital Markets,
The Software Investor’s Handbook to AI, offers a grounded view of what matters, what does not, and where to look for real opportunity in this fast-moving space.
The main message? Generative AI will be transformational, but not immediately. Investors are overestimating its short-term impact while underestimating what it could do over the next decade.
Don’t follow the hype
AI’s impact on software companies is already visible, but the gains so far are incremental. RBC highlights a paradox. Tools like GitHub Copilot show productivity boosts of more than 25% for developers.
Yet those efficiencies are not yet feeding through into company-level performance. Time saved is often not reinvested back into work. As RBC puts it, AI “is good for your golf game and your dog.”
This is why many current AI strategies are falling short. It is not enough to sprinkle AI into existing platforms. The real winners will be companies that fully reimagine their products, workflows and data models around AI from the ground up.
The four groups to watch
RBC breaks AI-driven software stocks into four categories of likely winners.
First come the large incumbents, such as Microsoft Corp (NASDAQ:MSFT), that combine huge data sets with strong distribution. These firms are well-placed but must move faster than their size might suggest.
Second are the vertical software specialists. These companies already dominate niche sectors like healthcare or logistics, and AI could widen their lead. A model trained to understand medical records or supply chain data offers a meaningful advantage.
Third are nimble mid-cap challengers that can use AI to innovate quickly and close the gap with slower-moving giants. In RBC’s view, these firms may outmanoeuvre legacy players by adopting AI more natively.
The fourth group includes enablers: companies that provide the tools, infrastructure or expertise to help others adopt AI. This includes cloud platforms, cybersecurity firms and data specialists.
And the names to avoid
Not everyone will benefit. RBC identifies four types of software companies at risk.
First are legacy on-premise vendors who lack the flexibility to adapt to cloud-based AI.
Second are firms that talk about AI but fail to rebuild their software around it. Third are analytics players that market themselves as AI firms without truly using the technology. And fourth are businesses whose core product can be replaced by generative AI, for example, simple task management tools.
RBC points to UiPath Inc (NYSE:PATH), UiPath and ZoomInfo as firms that could struggle under this lens.
What this means for investors
Private investors should be cautious of companies selling AI features that may soon be free to use. Basic tools like email writing or chatbot search will quickly become standard. These are not long-term differentiators.
Instead, look for companies that use AI to change their product in a fundamental way.
Also, pay attention to how they plan to monetise it. Some will charge directly for AI tools, as Microsoft has with Copilot. Others will use AI to drive indirect gains: higher retention, greater usage or better margins.
One area to watch closely is what RBC calls “agentic AI”, software that can perform tasks autonomously, not just offer suggestions. This could eventually replace today’s SaaS applications in areas like sales, HR or finance. But it will also require entirely new architectures and workflows.
Final thought
AI is not just a passing phase. It is, as RBC puts it, the fourth major technological shift after the internet, the cloud and mobile. But the trick for investors is to separate marketing from strategy, and novelty from utility.
The best opportunities are likely to emerge not where AI feels most impressive, but where it delivers measurable value — quietly, persistently and at scale.