We have seen what equilibrium looks like when a technology and its business model finally align, before not with AI, but with Google and AWS. The structural logic is identical, and the financial outcomes give us the clearest available evidence for the impact of business model and technology alignment.
Google’s technology was PageRank, the algorithm that ranked web pages by the authority of links pointing to them rather than by keyword density. It was genuinely superior. But Google did not become a $2 trillion company because of PageRank. It became one because of a pricing pivot. In 2002, Google shifted from CPM advertising — charging per thousand views, the television model applied to the web — to CPC, cost per click, charging only when a user demonstrated intent by clicking.
Revenue grew from $220,000 in 1999 to $29 billion by 2009 (Source: Alphabet historical financials). Yahoo, AltaVista, Excite, Lycos, and Ask Jeeves had comparable or earlier access to web search technology. None of them made the pricing pivot. All of them became irrelevant.
AWS tells the same story from a different angle.
The technology was virtualization, the ability to run multiple operating systems on a single physical server. It was not new. What was new was the aligned pricing model. Built on experience of separating business and technology functions, Amazon extended that to markets by shifting infrastructure from CapEx, enterprises buying servers and depreciating them over five years, to OpEx, renting compute by the second. Now a startup could scale to millions of users without a bank loan. An enterprise could spin up capacity for a seasonal peak and release it when the peak passed.
AWS revenue grew from near zero in 2006 to $107 billion by 2024 (Source: Amazon annual report, 2024). The incumbents had similar technology but did not have the pricing model, and they did not make the pivot in time.
The company that makes that pivot first, in the right layer, with the right accountability model, will not merely win market share. It will define the market.
In 1824, the French physicist Sadi Carnot described the theoretical maximum efficiency of a heat engine, the point at which no energy is wasted in the exchange between a hot reservoir and a cold one.
In the current market, energy is being wasted on both sides. Vendors are extracting revenue from enterprises through seat uplifts and consumption meters that do not correlate with the value delivered. Enterprises are paying for AI they cannot justify to their boards, buying on sticker price without auditing the iceberg beneath it. Both sides are losing heat.
Mindvista edition 39 featured, Service as a Software as a new model for the AI age, primarily as a vendor framework. A year later, in Edition 67, it becomes a two-sided contract for both enterprises and technology companies.
The six principles that define it are the same but it is modified to reflect both vendors and buyers as below
1. Outcome-Driven means revenue is linked to resolutions, not seats means abandoning the comfort of recurring seat revenue for the accountability of outcome pricing. For buyers, it means demanding contracts where the vendor has skin in the game and being willing to pay a premium when the outcome is genuinely delivered.
2.Decoupled Revenue and Cost means vendor growth does not depend on the buyer’s headcount. For vendors, this is about a model where adding the next customer does not require proportional engineering or support cost. For buyers, it means choosing vendors whose pricing scales with the business value created equitably.
3. Autonomous means the system performs the work with minimal human intervention. For vendors, it means investing in reliability and edge case handling before claiming resolution-based pricing. For buyers, it means being honest about where human supervision is genuinely required and pricing it accordingly.
4. Ethical and Secure means the delivery model includes guardrails and liability protection. This is where the Carnot efficiency breaks down most visibly today. The liability gap is the friction. For vendors, closing it means offering contractual indemnification and audit trails as standard features, not premium add-ons. For buyers, it means requiring them before signing.
5.Continuous Intelligence means the system compounds in value with every interaction. For example the more intelligent the context layer becomes, and the more defensible the vendor’s position. For buyers, this is the difference between renting a tool and building an asset.
6.Sustainable Advantage means the value is in the proprietary context, not the wrapper. Open source has established that the wrapper that the orchestration layer, the basic agent logic will be commoditized. The sustainable moat is the continuous increase in intelligence and reliability and the compounding institutional knowledge encoded in the system over time.
Nothing today meets all six of these well but perhaps the principles can help progress for technology and business model alignment that we saw with Google Search and AWS.
