Incumbent SaaS Companies Are Misreading the AI Moment
Share
Picture this: It's 1998. A mid-tier retail executive sits in a boardroom, staring at quarterly reports. E-commerce startups like Amazon are popping up, but they're dismissed as low-margin experiments. "We'll wait until the economics make sense," the exec says. Fast forward a few years, and that same company is scrambling to catch up as customers flock to online shopping. The lesson? Disruption doesn't wait for perfect margins.
The pace of disruption in software today feels less like a steady evolution and more like a structural shift. Even experienced leadership teams are finding it difficult to separate signal from noise. Every company claims to be "AI-first." Every product announcement promises transformation. Yet the anxiety underneath the surface is real. The reason is simple: this moment is not about feature velocity or incremental innovation. It is about whether existing business models remain defensible at all.
This is not a typical technology upgrade cycle. It is a once-in-a-generation business model transition. And in moments like this, the greatest risk is not moving too slowly. It is optimizing the wrong thing while the battlefield changes.
AI Is Not a New Business, It Is a Reinforcement of the Moat
For incumbent SaaS companies that operate systems of record, AI should not be viewed as a standalone product category. It is far more analogous to reinforcing the foundations of a building during an earthquake than adding a new floor. The goal is not expansion for expansion’s sake. It is ensuring that the structure remains standing when the ground shifts.
AI-native startups are not trying to out-feature incumbents. They are attempting something far more dangerous: they are trying to pull users into entirely new workflows. If they succeed, the system of record becomes the system of regret.
History is unambiguous here. In the late 1990s, many retail incumbents believed e-commerce was a low-margin distraction. The ones that waited for margins to improve before investing didn’t lose because they lacked capital. They lost because customers learned a new default behavior. Once that happened, the game was over.
AI represents a similar behavioral shift in software. Once users become accustomed to software that anticipates intent, removes friction, and collapses workflows, they will not willingly return to systems that feel static or labor-intensive. Incumbents must ensure that this shift happens inside their platforms, not around them. Take Salesforce, for example. By embedding AI deeply into CRM workflows with tools like Einstein, they've kept users engaged without forcing a full platform switch.
The Strategic Advantage Incumbents Are Underusing
What makes the current moment so striking is that incumbent SaaS companies are actually holding the strongest hand. Yet many are playing it conservatively.
Public software companies already possess three advantages that no venture-backed competitor can replicate:
- Deeply embedded customer relationships
- Distribution at global scale
- Highly cash-generative core businesses
This combination creates a right to win. But only if leadership teams are willing to deploy those advantages aggressively.
Too often, AI investments are being evaluated through the lens of traditional SaaS efficiency metrics: near-term margins, linear ROI, and immediate monetization. That is understandable. But it is also the wrong lens. This is not a product optimization decision. It is a capital allocation and risk management decision.
When the core franchise is at stake, protecting it should take precedence over preserving short-term margin optics. Microsoft learned this with Azure. They invested heavily in AI integration early, using cash flow to defend against AWS and Google Cloud, even if initial returns were slim.
Why Low-Margin AI Is a Rational, Defensive Strategy
There is a persistent concern that offering AI capabilities at little or no margin is undisciplined. In reality, for incumbents, it can be the most disciplined move available.
Venture-backed competitors are effectively using investor capital to subsidize customer acquisition and workflow migration. Incumbents can counter this by using operating cash flow. The most stable and lowest-cost capital there is. To defend their installed base.
Think of it like a price war that only one side can afford to fight indefinitely. Startups must continuously raise capital to sustain losses. Public incumbents, by contrast, already generate the cash needed to fund the defense of their own customers.
In this context, low-margin AI is not a concession. It is a strategic shield. One that prevents competitors from establishing a foothold while buying time for monetization models to mature.
The Two-Engine Model Investors Can Actually Understand
The market’s skepticism around AI is not about profitability. It is about clarity. Investors are comfortable with investment phases as long as leadership teams are explicit about intent and measurement. The cleanest way to communicate this is through a two-engine operating model.
The first engine is the core system of record: stable, growing, highly profitable, and responsible for funding the business. The second engine is the AI expansion layer: usage-based, adoption-focused, and intentionally prioritized for growth over margin.
When reported transparently, this structure sends a powerful signal. It shows that the core business remains healthy while the company is deliberately reinvesting its strength to extend its moat. Over time, the market should reward companies that can demonstrate both durability and relevance. Especially when many pure-play AI companies lack the former. HubSpot has adopted a similar dual approach, balancing its CRM core with AI-driven growth tools.
Pricing, Sales, and the Organizational Reality Ahead
AI will not only reshape products. It will reshape operating models. Pricing will inevitably move away from purely seat-based structures toward consumption or value-based models. This shift may feel uncomfortable. But it reflects a simple truth: AI dramatically increases the value delivered per user.
Sales organizations will evolve as well. AI reduces the need for human involvement in early-stage selling and onboarding. Much like e-commerce reduced the need for physical retail staff. That does not eliminate the role of sales. It concentrates it where it matters most.
A useful mental model is this: AI should deliver the first 80% of value automatically. Humans should focus on the final 20%. The complex, strategic, high-trust interactions that machines cannot replicate.
At the same time, AI presents a massive, often overlooked upside on the cost side. Software companies are uniquely positioned to benefit from AI-driven operational efficiency. Over the next several years, margin expansion driven by opex reduction should be not just possible, but likely.
What CEOs Should Do in the Next 12 Months
- Explicitly frame AI as a moat defense strategy, not a standalone growth bet, in board and investor conversations.
- Separate core and AI reporting, even if AI revenue is small or low margin today.
- Prioritize AI adoption and engagement metrics over short-term monetization.
- Use cash flow aggressively but transparently to neutralize competitive threats.
- Begin transitioning pricing and sales models now, before external pressure forces rushed decisions.
- Invest internally in AI-driven efficiency, and prepare the organization for structural cost improvements.
The Question the Market Is Actually Asking
The market is not asking whether incumbent SaaS companies can build AI. It is asking whether they are willing to temporarily under-monetize AI in order to permanently protect their system-of-record position.
Those who answer that question decisively will emerge from this transition stronger, more relevant, and more defensible than before. Those who hesitate may find that by the time margins look attractive, the customer has already moved on. AI is accelerating the natural sorting process that exists in every industry.
