How Outcome-Based Contracting Can Enable Enterprise AI Deployments

Ben Blanquera – VP – AI and Sustainability, Rackspace Technology.

In my role as vice-president of sustainability and AI at Rackspace Technology, I’m seeing organizations invest in enterprise artificial intelligence, expecting massive productivity gains, particularly in business process returns. But as companies rush deployments, they often encounter significant roadblocks unrelated to coding or data as they try to build the future of intelligence using contracting models from the past. McKinsey’s latest State of AI report shows broad adoption, persistent scaling challenges and a strong link between workflow redesign and meaningful business impact.

That’s because traditional time-and-materials or fixed-fee software contracts are misaligned with how AI delivers value. One better way to gain value from AI is through outcome-based contracting.

Why Legacy Models Often Miss The Mark

Traditional contract approaches tend to fail because there’s typically no correlation between the amount spent and the value generated by AI projects that are either extremely complex or involve significant manual effort, with multiple integration points.

Time-and-materials contracts or fixed-price software agreements can miss the mark for three reasons: First, there’s no correlation between the input and the output value. Second, these contracts lack accountability, with no easy way to hold a vendor accountable for an outcome when pricing is based on discrete components. Third, traditional models don’t account for the comprehensive cost of ownership. Paying for a software license or a service contract is only a fraction of the investment. A true AI outcome requires investment in change management, product feature management and AI literacy. Legacy contracts also ignore avoidance costs, such as the software, services and human labor that the organization no longer needs to fund once the AI solution is live.

Navigating The Value Prioritization Matrix

One common mistake that enterprises make is failing to properly prioritize, getting caught between two ends of a spectrum. Some focus on problems that are too narrow, such as building a simple agent for procurement spend management rather than considering the entire supply chain. Others try to solve for too much at once, attempting to optimize operations across an entire billion-dollar enterprise simultaneously.

Successful AI deployment requires a clear strategy and a functioning operating model. This involves two flywheels that must work in sync. The first is the process of reimagination. The second is data and technology. What connects these two is change management, organizational strategy and product management.

Many organizations look at these flywheels individually. However, when you view process and technology through a single lens, you can identify use cases that offer the highest return on investment. Given the plethora of choices between agents, large language models, small language models and hyperscaler stacks, taking a step back to define this operating model is the only way to avoid making short-term decisions that fail to optimize the structure for the long term.

A New Dynamic For Vendors And Buyers

Outcome-based pricing changes the relationship with vendors. There’s a misconception that procurement and IT leaders are only looking for the cheapest deal. In reality, they’re looking to implement the right solution and move fast. When a vendor can deliver an AI outcome and charge for that value, they become a true strategic partner and a trusted, outcome-based provider.

For vendors to successfully move to this model, they must invest in several key areas, starting with consultative salespeople who can listen to customers and understand the business outcome, not just the technology. They also need advisory capabilities to help customers articulate and quantify outcomes, as well as strong engineers who can build as consulting progresses. The days of hiring a firm to consult for months and then parachuting in an engineering team to implement are gone. Both must happen in concert to deliver a proof of value at speed.

Finally, vendors must provide the infrastructure, multicloud capabilities and data skills needed to run production AI workloads. This includes establishing governance frameworks, managing FinOps to control inference costs and ensuring model accuracy doesn’t degrade over time. By managing the entire end-to-end continuum, vendors can help ensure that AI at scale becomes a lasting competitive advantage for the enterprise.

Conclusion

By moving away from legacy models that reward hours worked over results achieved, organizations can finally bridge the gap between AI’s conceptual promise and measurable business impact. When incentives are aligned around value, the focus shifts from managing costs to maximizing returns, allowing enterprises to scale intelligence with confidence.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?