---
id: ins_ramanujam-ai-pricing-against-labor-budgets
operator: Madhavan Ramanujam
operator_role: Senior partner Simon-Kucher; author Monetizing Innovation
source_url: https://www.simon-kucher.com/en/insights/monetizing-innovation
source_type: essay
source_title: "Monetizing AI — Labor Budgets, Not IT Budgets"
source_date: 2024-09-01
captured_date: 2026-05-05
domain: [ai-native, pmm, gtm]
lifecycle: [pricing-packaging, monetisation, product-positioning]
maturity: frontier
artifact_class: framework
score: { originality: 5, specificity: 5, evidence: 4, transferability: 5, source: 5 }
tier: A
related: [ins_simon-single-price-always-suboptimal, ins_intelligence-cost-converging-electricity, ins_value-equation-grand-slam-offer]
raw_ref: raw/expert-content/experts/madhavan-ramanujam.md
---

# AI products should price against labor budgets, 10× larger than IT budgets, and capture 25-50% of value, not the SaaS-typical 10%

## Claim
AI products that replace or augment human labour should price against the buyer's *labour* budget, which is approximately 10× larger than the IT budget that traditional SaaS competes for, and can credibly capture 25-50% of the value created, far exceeding the traditional SaaS 10% capture rate. The pricing model has to evolve: per-seat / per-feature SaaS pricing under-prices labour-substituting AI by an order of magnitude.

## Mechanism
Traditional SaaS prices against the IT budget, a constrained pool the CIO controls. The labour budget is what HR / business-unit leaders control and is typically 10× larger because labour is the dominant operating cost in most companies. AI products that can credibly substitute for or augment human work (customer-support agents, sales-development reps, data analysts, copywriters) compete for that larger pool. Furthermore, because the AI's output is more autonomously attributable than feature-driven SaaS (a 50% reduction in support-ticket volume is measurable; a "productivity gain" from a generic SaaS is not), buyers accept higher value-capture rates. The right pricing model: outcome-based, ROI-attributed, priced as a fraction of saved or augmented labour cost.

## Conditions
Holds when:
- The AI product credibly substitutes or augments specific human labour with measurable outcomes.
- The buyer has a labour budget pool the AI can compete against (most enterprise functions).
- The value created is attributable in the buyer's measurement systems (not embedded inside a larger product where attribution is muddy).

Fails when:
- The AI is a feature within a larger SaaS product without clear labour-substitution claim.
- The labour the AI substitutes for is politically or ethically sensitive (legal compliance, healthcare diagnosis).
- The buyer's measurement systems cannot attribute the outcome, the value-capture argument lands but cannot be priced concretely.

## Evidence
> "AI companies should price against labor budgets (10x larger than IT budgets) and can capture 25-50% of value created, far exceeding the traditional SaaS 10% capture rate, because AI offers higher autonomy and clearer attribution."

· see `raw/expert-content/experts/madhavan-ramanujam.md` line 20.

## Signals
- Pricing model is outcome-based or labour-substitution-priced, not per-seat or per-feature.
- Sales motion targets the line-of-business buyer with labour budget, not the CIO with IT budget.
- ROI attribution is built into the product so the buyer can measure value capture in their own dashboard.

## Counter-evidence
The 25-50% capture rate assumes credible attribution and acknowledged labour substitution. Many AI products struggle to attribute value cleanly (the work would have happened anyway, the human was retained alongside the AI), and capture collapses to traditional SaaS rates. Sam Altman's `ins_intelligence-cost-converging-electricity` argues that the long-run trajectory is for intelligence cost to drop toward electricity cost, which would compress value-capture rates over time as the AI itself becomes commoditised.

## Cross-references
- `ins_intelligence-cost-converging-electricity`, Altman's frontier claim that complicates the long-run capture-rate trajectory.
- `ins_simon-single-price-always-suboptimal`, segment-by-WTP pricing applied to AI.
- `ins_value-equation-grand-slam-offer`, Hormozi's value equation; AI products score very high on the numerator (Dream Outcome × Likelihood) when they substitute labour.