AI rewrites the economics of Amazon’s cloud-consulting business

• AI is pushing radical changes on the consulting industry.
• Clients of Amazon’s big ProServe cloud-consulting group seek AI-driven discounts.
• ProServe is embracing AI agents and aims to make 75% of projects fixed-price, documents show.

Clients are starting to press consulting firms with an awkward question: If AI is doing more of the work, why are we still paying the same fees?

It’s an issue hitting home inside Amazon‘s cloud business, through a consulting arm called ProServe that influences more than $10 billion in annual revenue for the tech giant.

Major customers, including Booking.com and Danske Bank, have “proactively” sought to renegotiate ProServe contracts, pushing for what an internal Amazon document described as “AI-delivered pricing,” along with faster results at lower cost.

“If GenAI enhances or speeds things up, they should get more value faster and shouldn’t pay the same rates,” the document from June stated, summarizing the prevailing sentiment among ProServe clients.

This pressure is reshaping the ProServe unit. Internal documents reviewed by Business Insider show Amazon reorganizing the cloud consulting arm around AI agents it describes as “digital employees,” while pivoting away from a traditional billable-hour model.

The changes reflect a broader reckoning rippling through the consulting industry. The twin forces of pricing pressure and automation are pushing other consulting firms in a similar direction, accelerating a shift toward outcome-based pricing that charges for results rather than time-based human labor.

“Our customers’ buying patterns shifted to favor agile, outcome-based, Al-powered engagements with faster time to value and clear ROI, creating a significant revenue and workload concentration risk in our top revenue cohort,” the ProServe document warned.

AI is challenging the billable-hour model

Consulting has long operated on a straightforward equation: revenue scales with headcount and hours worked. Projects are often priced on an hourly basis, with teams of engineers, architects, and project managers assigned to client engagements.

But generative AI tools are now accelerating coding, documentation, and software modernization work. Tasks that once required teams of consultants can increasingly be automated.

If work gets done faster, with fewer people, clients expect savings.

ProServe planning documents from June and November acknowledged that dynamic, noting customer demands for AI-enabled efficiencies and lower effective pricing.

The June document outlined a “fundamental” strategic shift from “traditional project-based consulting to product-led, outcome-based delivery.” By this year, ProServe aimed to have up to 75% of projects structured as “fixed-price” deals, this document noted.

AWS is not alone in confronting this shift.

Major consulting firms, including Accenture and Deloitte, have made large-scale AI investments and increasingly emphasize outcome-based pricing in client engagements. Executives across the industry have signaled that generative AI will compress traditional labor-based work while forcing firms to compete more on measurable results. Auditing giant KPMG, for example, pressed its own auditor for lower fees, citing AI efficiency gains, the Financial Times reported recently.

Building a ‘digital workforce’

To support this shift, ProServe is formalizing its approach to deploying AI agents, according to the documents.

ProServe outlined plans to establish an “Agentic AI” team and introduce an “Agent Operating Model” to manage autonomous software that can take action with minimal human intervention.

ProServe described AI agents as a “digital workforce that complements human teams” and proposed creating an AI Agent Service and Marketplace to govern and deploy them, according to the documents.

One of the documents from June projects substantial productivity improvements, including automation of up to 90% of certain tasks.

“This positions AWS as a leader in building AI agents, and managing them as a new category of digital employees,” the June document said.

Amazon adapts

An Amazon spokesperson told Business Insider that ProServe identified early that “agentic AI innovations” were driving a shift in the consulting industry and moved swiftly to adapt its business model. Customers were seeking outcome-based, fixed pricing, and AWS responded by rolling out services designed to deliver greater value and measurable results, the spokesperson added.

“This AI-powered consulting model accelerates customers’ business transformation, enabling them to innovate in less time, at a lower cost, and with more value,” the spokesperson said.

Amazon shared some of its plans for an “agent-first consulting approach” in a blog post in November.

Other firms are following suit. McKinsey CEO Bob Sternfels recently said the consulting firm added 25,000 AI agents to its staff in less than 2 years. PwC, EY, and Boston Consulting Group are also doubling down on deploying AI agents internally.

Consulting as a cloud growth engine

The stakes are high for Amazon Web Services, the world’s largest cloud provider, which relies on ProServe to help it win and keep customers.

According to the planning document from June, ProServe’s consulting work helped AWS generate roughly $12.5 billion in AWS revenue in 2024. The unit’s public-sector practice alone was projected to generate more than $500 million last year, the November document estimated. It wants to help AWS become a $500 billion company in less than a decade, the June document stated, up from $129 billion last year.

ProServe has already seen efficiency gains from AI, according to the documents from last year. The unit was on pace to post a positive operating margin for the first time last year and is targeting a 20% year-over-year increase in revenue per employee in 2026 by expanding its use of AI tools. It also expected to drive broader AI adoption among AWS customers without materially increasing headcount.

Even as profitability improves, ProServe salespeople are confronting a revamped pay structure. Starting this year, hundreds of sales employees moved to a more variable compensation model, according to people familiar with the shift. The new plan cuts back on guaranteed stock awards and links a greater portion of compensation to performance-based incentives, aligning ProServe with the pay model used across other AWS sales roles.

“Enabling AWS growth while maintaining a profitable business is our North Star, guiding our bold transformation for 2026,” the June document stated.

Have a tip? Contact this reporter via email at ekim@businessinsider.com or Signal, Telegram, or WhatsApp at 650-942-3061. Use a personal email address, a nonwork WiFi network, and a nonwork device; here’s our guide to sharing information securely.