In today’s software market – especially in the AI cycle – capital is flowing into an unusually broad set of categories. Many of them are being funded, scaled and valued as if they can produce billion-dollar outcomes.
But a recurring pattern suggests otherwise.
There are entire segments of the software market that, structurally, are unlikely to generate outlier outcomes – regardless of execution quality. The challenge is not identifying these markets in hindsight. It is recognizing them early enough, when they still look indistinguishable from genuinely large opportunities.
This distinction is becoming increasingly relevant in the current environment, where AI has expanded the perceived opportunity set across almost every category.
When Early Signals Mislead
In the first phase of a software company’s life, most categories behave similarly. Companies show strong growth, high retention and clear product-market fit. Investors, understandably, extrapolate from these signals.
However, early traction primarily reflects the strength of the product – not the size or quality of the underlying market.
For example, over the past decade, many companies built software around compliance, reporting and internal audit workflows. These products often scaled efficiently to tens of millions in revenue, with strong retention driven by regulatory necessity. From the outside, they looked like high-quality SaaS businesses.
But over time, growth tended to plateau. The budgets were defensive, owned by non-strategic buyers, and continuously scrutinized. The software was required, but not central to value creation. As a result, expansion was limited, and the market ceiling became visible only after significant scale had already been reached.
This pattern is not specific to compliance. It reflects a broader dynamic: software sold into cost centers rarely compounds into large platforms.
The Limits of Vertical and Fragmented Markets
A similar effect can be observed in parts of the vertical SaaS ecosystem, which has attracted significant investment in recent years.
In many cases, companies successfully dominate highly specific customer segments – for instance, software tailored to independent medical practices, niche construction trades, or localized service providers. These businesses often achieve strong penetration within their defined market and exhibit high retention and customer satisfaction.
However, the total number of potential customers is inherently limited, and pricing is constrained by the economics of the underlying industry. Expansion into adjacent verticals is often more complex than anticipated, as each segment requires different product capabilities, go-to-market strategies and integrations.
As a result, what appears to be category leadership is often leadership within a bounded market, rather than the foundation of a scalable platform.
The Fragility of Non-Core Tools in an AI-Driven Stack
The current AI cycle has amplified another category of structurally constrained opportunities: tools that enhance workflows without owning them.
Examples include AI-powered analytics layers on top of existing systems, meeting intelligence tools, or point solutions that automate specific tasks within broader workflows. These products are often adopted quickly because they are easy to deploy and demonstrate immediate value.
However, their position in the stack is inherently fragile.
Platform owners – such as CRM, ERP and collaboration systems – have strong incentives to internalize these capabilities. At the same time, the cost of building similar functionality is decreasing due to the widespread availability of AI models and infrastructure.
This dynamic leads to a compression of differentiation over time. Products that initially appear differentiated can become features within larger systems, limiting their ability to sustain independent, large-scale outcomes.
When Value Creation Does Not Translate Into Value Capture
Another recurring pattern involves companies that deliver clear operational value but struggle to capture it economically.
This is particularly common in tools focused on productivity, internal collaboration, or marginal efficiency gains. While the aggregate benefit to an organization may be meaningful, it is often difficult to attribute that value to a specific budget owner.
As a result, pricing power remains limited, even when usage is high. These companies can build solid businesses, but the gap between value creation and value capture constrains their ability to scale into large outcomes.
Valuation, TAM and the Assumptions in Between
One of the clearest indicators of this tension emerges in how companies are valued.
It is not uncommon to see companies priced at valuations that exceed their stated total addressable market. In some cases, this reflects a legitimate belief that the market is larger than currently defined. Investors may assume that the company will expand into adjacent categories, redefine the market, or achieve unusually strong levels of dominance and profitability.
This has been true for a number of breakout companies, where the initial TAM proved to be an incomplete representation of the opportunity.
However, this logic depends entirely on what is implicitly assumed. When the market is already well-defined and there is no credible path to expansion, a valuation above TAM is typically a red flag. It suggests that the investment case relies on unrealistic assumptions about market share, pricing power or long-term economics.
The key question is not whether valuation exceeds TAM, but rather what must be true for that valuation to be justified – and whether those conditions are actually plausible.
Why This Pattern Persists
Despite these recurring outcomes, capital continues to flow into structurally constrained segments.
Part of the reason is that early-stage metrics do not reveal market limitations. By the time the constraints become visible, companies may already be well-funded and valued accordingly.
In addition, narratives around market expansion – particularly those tied to AI – can make even well-defined markets appear open-ended. The assumption that technology will fundamentally reshape market boundaries is sometimes correct, but it is often over-applied.
Finally, the cost of missing a true outlier is significantly higher than the cost of backing a company in a constrained market. This asymmetry encourages investors to tolerate a higher number of false positives.
Market Selection as the Primary Lever
Ultimately, the difference between good outcomes and exceptional ones in software is often determined at the level of market selection.
The largest outcomes tend to emerge in markets that are directly tied to revenue generation, where companies can own core workflows, expand over time, and capture a meaningful share of the value they create.
In contrast, markets defined by structural constraints – whether due to buyer type, market size, or position in the technology stack – tend to produce strong companies, but not outliers.
The challenge is that, in the early stages, both types of markets can look equally attractive.
That is why this pattern continues to repeat.
And why, in software investing, the most important question is often not whether a company is good – but whether the market it operates in is capable of supporting the outcome investors expect.
