City Shift Finance: Where are SaaS leaders still getting growth and financial discipline wrong today?
Ethan Ruby: They are operating on outdated assumptions.
The way software is bought has changed, and so has the way software businesses carry cost, but a lot of SaaS leaders are still managing as if neither has moved.
For years, headcount was the largest expense, and headcount is easy to track and control. AI costs behave differently. They scale with usage, sit inside infrastructure, and often grow without the same visibility as payroll.That is the risk. Not necessarily that the spend is too high, but that many companies are still managing with a financial model built for a world where labor was the dominant cost. When cost of revenue changes faster than leadership’s understanding of it, growth and burn both become harder to control.
On the growth side, buyer behavior has shifted. The old SaaS model was simple: a company had a problem, identified a process, and bought a dedicated system to manage it. Winning meant building the strongest standalone product. AI has changed that, buyers now think more in systems and workflows. Instead of immediately looking for a new vendor, they are asking whether AI can handle the work inside the tools they already use or whether they can connect existing systems to solve it.
That changes what growth requires: it is no longer enough to explain what your product does on its own. Companies have to show how they fit into a broader operating environment and support the way customers are now building around AI. A strong product still matters, but if it is positioned as a standalone solution, it becomes easier to replace. That same lag is showing up financially. A lot of attention is on token spend, but the bigger blind spot is cost of revenue.
Most companies still do not fully understand what their cost of revenue looks like in an AI-driven environment, even though it now affects COGS, payback periods, and burn management.
City Shift Finance: How is AI changing financial visibility, and where are companies overestimating its business impact?
Ethan Ruby: It is forcing companies to rethink both what they track and what they believe AI is actually capable of.
On the financial side, it starts with cost of revenue. Token costs are now a real part of COGS, which means AI has introduced an entirely new cost layer that many SaaS companies are still learning how to measure. But it goes beyond the product itself. There is also internal AI spend, what engineers and teams are consuming across tools every day, creating a cost center that simply did not exist before. And it is growing fast.
In many cases, AI spend is becoming larger and more complex than cloud costs were in the last cycle. The difference is that cloud was mostly infrastructure. AI touches infrastructure, product delivery, and individual usage across teams, which makes the cost far more distributed and harder to track. That changes what financial visibility needs to look like. The real question is not just what is being spent, but what is coming back.
AI is sold on productivity, so companies should be able to see whether that spend is translating into efficiency, output, or reduced labor requirements. In many cases, it is not. They are paying for the AI while carrying the same headcount. Without visibility into both sides, companies cannot tell whether AI is an investment or simply a new operating expense. That same disconnect shows up in how companies think about AI’s business impact. The biggest overestimation is assuming AI makes decisions when it does not.
It can process data, generate recommendations, and accelerate information flow, but that is not the same as making the right call. The harder the decision, the less AI can substitute for human judgment, because the right answer is usually shaped by industry context, product knowledge, and experience. That is where companies are misreading the shift. They assume AI replaces skilled workers whose value is in judgment, when in reality it mostly replaces the movement and processing of information. What AI actually does is make the difference between those two types of work much clearer.
As information handling becomes automated, the people who are strongest at synthesizing it and making strategic calls become more valuable, while the gap between them and everyone else becomes easier to see. That is the real impact, and it is far more structural than most companies realize.
City Shift Finance: What separates durable growth from short-term growth?
Ethan Ruby: So, if you anchor your product to a real business need rather than to a passing technical trend, you get growth that compounds instead of growth that evaporates the moment the hype moves on to the next acronym.
It’s longevity, it’s how long people are actually going to need the thing you built.
And right now there are a lot of fads, especially in AI. Should we buy or build this? Do we use MCP? Do we use RAG, CLIs, which model, and on and on. You’ve seen a lot of companies build a niche product aimed exactly at where the industry happens to be at that moment. They get a burst of growth from it. But then the models evolve, or the surrounding systems change, and that product just quietly goes away. That’s the trap.
If what you’ve built is essentially “the technical AI topic of the month, but as a product,” you might catch some short-term growth, but it won’t turn into durable revenue. The ground shifts underneath you. Durable growth comes from solving real problems, and real problems are the boring, permanent ones. How do I increase revenue? How do I reduce costs? Those questions don’t go away no matter what happens with the technology, every company has to deal with them forever.
City Shift Finance: What defines strong SaaS leadership in the next five years?
Ethan Ruby: It’s the ability to actually see where the industry is going. How are people going to use software five years from now? It might look very different, it might not look that different, but the biggest change is probably going to be in how we build software with AI in the first place.
Strong leadership is being able to look around the corner, form a real point of view about that future, and then have the conviction to start building toward it now, before it’s obvious. That conviction part is the hard part. It’s easy to wait until the direction is clear, but by then the advantage is gone. The leaders who do well are the ones who can hold a view about where things are heading and commit to it while there’s still uncertainty.
And the way you keep that from turning into chasing fads is the same principle as durable growth. Anchor it to real problems, find real businesses that genuinely need to increase revenue or cut costs, and build toward the future in service of those problems, not in service of whatever’s trendy.
So, it’s two things held together. See the future clearly and have the nerve to build for it but stay rooted in the durable problems that will still matter once the current hype cycle is over.
