price to win: ai redefines software economics

It’s a fact: one in five startups fails because of misaligned pricing for its product or service. Clearly, pricing must be a first-order design choice, not an afterthought - it represents a fundamental prerequisite for commercial take-off.

However, the advent of AI is dramatically disrupting traditional SaaS models, making software pricing even more complex. To illustrate how this complexity can become a source of success, this article tells the story of how AI expands the solution space for pricing models, how we recommend navigating this space, and the business model implications that result.

Pricing as a Source of Failure

But first, let’s step back: how can pricing become a source of failure for every fifth startup, in the first place? At redalpine, we have seen two archetypes responsible for pricing failures: the Overengineer and the Leapfrogger.

• The Overengineer tends to push for technical maturity without adequate pricing maturity in place, often resulting in costly and complex solutions that fail commercially.

• The Leapfrogger, conversely, prioritizes go-to-market strategies prematurely, attempting to establish a business model before even having a viable product.

The ideal path lies somewhere in between these two extremes (see in illustration 1).

Illustration 1: Conceptual Model on the Ideal Path of Software Pricing

Disruption of SaaS Pricing

Fortunately, the golden rules for establishing pricing maturity in the SaaS era were clear. SaaS models operated on a simple assumption: software value scales with user count or usage, while variable costs remain minimal. This gave rise to per-user, usage-based, and flat-rate models that comprised about 90% of all software pricing strategies. Thus, while pricing was essential, it wasn’t a major barrier for our portfolio companies.

However, AI fundamentally disrupts software economics and thus SaaS pricing, in three major ways:

• AI disrupts value creation per user: The number of users becomes less relevant as a measure of software value.

• AI disrupts scalability: Software adoption becomes significantly easier with AI, enhancing scalability.

AI disrupts cost structures: AI reintroduces substantial variable costs - specifically, compute costs.

How Sierra is solving the AI pricing riddle - emergence of value-based pricing

While the disruption of SaaS is shaking the foundations of many software companies, others turn it into a factor of success.

One standout is Sierra AI, the Bret-Taylor-backed powerhouse that designs and operates domain-specific customer-service AI agents that plug into a company’s existing support stack. Sierra’s answer to the SaaS disruption: outcome-based / (value-based) pricing. This means charging a fixed fee for each achieved, predefined result (outcome orientation) or adjusting the fee in proportion to the total economic benefit created (value orientation). Or, in other words, customers pay only when the agent delivers. In practice this could be a resolved support conversation, a prevented cancellation, an upsell, or any other clearly defined success metric. If the conversation remains unresolved, no charge is incurred.

This model has two immediate consequences:

1. Higher margins through value capture: Because pricing scales with the business impact the agent creates, Sierra can claim a fair share of that value instead of being capped by cost-plus or token-based schemes.

2. Shrinking the SaaS bill: Clients pay only when measurable outcomes are achieved, aligning incentives and removing the fear of overpaying for unused or ineffective software.

   
   

By anchoring price directly to realised business value, Sierra sets a precedent for a broader shift toward value-oriented pricing in the software market. This shift will challenge vendors wedded to purely usage-based or cost-based models and reward those who can prove, and price for, tangible outcomes.

Yet, this approach might not suit every scenario. For instance, does value-based pricing contradict widely-adopted token pricing? How do we effectively navigate between usage-based and value-based pricing models?

We’ve looked closely at this conundrum and strongly believe pricing in the age of AI isn't a binary decision. Instead, the optimal degree of value orientation depends on two key factors:

• Degree of Competition: With increasing competition, software pricing often shifts from value-based toward cost-based models, driven by rising price pressures and shrinking margins.

• Degree of Service Integration: The more integrated the software - from basic infrastructure to comprehensive, end-to-end service automation - the more it favors value-based pricing due to substantial business value creation.

  
  

Thus, we identified three dominant strategies, each with distinct advantages and drawbacks. For instance, value-based pricing offers the highest alignment with delivered value, while cost-based models effectively offset variable costs. Hybrid models often represent the ideal compromise. These relationships are illustrated in Figure 2.

Illustration 2: Decision Matrix on AI pricing strategies

The AI Pricing Trap

Lucrative margins attract many AI companies to adopt hybrid or value-based pricing. However, we believe only about 5 - 15% will sustain sufficient vertical differentiation to defend these margins. The other 75 - 95% risk engaging in a margin race down to the marginal cost of compute - driven by increasing commoditization of AI. We refer to this risky pursuit as the AI Pricing Trap.

For the top 5% that can successfully implement value-oriented pricing, the primary challenge remains the domain-specific notion of value. We have observed that the solution space typically converges into four conceptual approaches:

• Top-down & Bottom-up True Value-Oriented Pricing: 

  Based on measured business value.

• Per-execution & Per-agent Outcome-Based Pricing: 

  Determined by transaction success.

  
  

This solution space is mapped in Figure 3.

Illustration 3: Solution Space for Value-based Pricing

Business Model Context of AI Pricing

Regardless of how critical pricing strategy becomes in the age of AI, it is never an isolated decision. We’ve established three critical business model implications:

• Aligned Incentives: Shifting vendor-customer relationships from lock-in toward shared value creation.

• TAM Revolution: AI expands total addressable market (TAM) and margins by offering AI-powered vertical services priced as OPEX solutions rather than CAPEX infrastructure.

• Radical Shift to Domain-Specific Business Logic: Success will increasingly depend on domain-specific R&D, proprietary datasets, and vertical expertise.

  
  

Recommendations for Action

As illustrated, value-aligned pricing realigns vendor-customer incentives, expands TAM, and demands deep domain expertise (proprietary data, vertical talent). The implications for both builders and investors are clear:

• Look for a 10× vertical advantage or 10× scale in software economics.

• Make pricing a strategic decision - not merely a margin-optimization exercise, but integral to all business model components.

• Crucially, avoid the AI Pricing Trap: only pursue value-based pricing with a robust conviction of defensibility.