For all practical purposes, most service delivery models today are still centered around one thing: human effort.
Whether it’s time and material (T&M), fixed fee, or milestone-based, the commercial model revolves around the number of hours someone puts in, not necessarily the outcomes achieved. This can be primarily attributed to ease of convenience associated with the traditional pricing models in the context of services.
Some of the recent comments by CXOs of large firms around putting 70, 80 or 100 hours a week reflect the fascination of linking value created to the number of hours one puts to deliver an outcome. But that has to change!
With the rise of enterprise automation and agentic artificial intelligence (AI), we’re now seeing the emergence of a new, much-needed layer in the enterprise tech stack, Systems of Execution. And this isn’t just another buzzword. It’s a fundamental shift in how services are designed, built, run, and paid for. With the underlying transformation in the tech stack, we can also expect a shift in the way these services are consumed and paid, thus the rise of execution-based pricing.
So, what exactly is a System of Execution?
Think of it as the intelligent brain that sits on top of various execution elements, be it a platform (CRM), task-based bots, agent swarms, or legacy scripts. SoE sits above operational systems to sense signals, reason over context, and trigger coordinated actions across tools. As an example, consider how a SoE might work in sync with Customer Relationship Management (CRM).
- Orchestration layer: This sits above operational systems and takes care of the goals and plan for SoE. It continuously listens for signals across the CRM (e.g., Salesforce), email/calendar platforms (e.g., Outlook, Gmail), and engagement tools (e.g., Salesloft), maintaining memory of deal progression. It detects that buyer activity has gone quiet relative to past patterns. It generates prompts to ask foundational models for help
- Foundational model: This is the Generative AI (gen AI) part of SoE, powered by reasoning frameworks like ReAct or Tree-of-Thought, evaluates historical sales behavior and pipeline dynamics to determine a plan of action. It determines that executive intervention, a tailored incentive, and forecast reclassification are the optimal actions. It sends the planned action back to orchestration layer, which executes them through the tools it has access to
- Tools: These are the systems through which the SoE executes. That could include Application Programming Interfaces (APIs) to schedule outreach, document automation tools (e.g., Conga or DocuSign) to generate a renewal proposal, Slack bots to alert the account team, and CRM APIs to update opportunity status
In this architecture, the CRM is both a data source and a tool, observed and acted upon. The SoE binds it all together, autonomously sensing context, reasoning across enterprise history, and triggering intelligent, multi-system action, without waiting on a human workflow.
The reality of effort-based services
Here’s the issue: for years, service providers have been incentivized based on effort. T&M billing rewards more time spent. Buyers have preferred to work in T&M, especially as they co-source projects. For providers, the benefit is that the meter keeps running.
Fixed-fee deals are scoped around assumptions of effort with certainty around parameters that impact the quantum of that effort. Even milestone-based models are often tied to activity, not outcomes.
And because of this, there’s little reason to fundamentally rethink how work is done.
Sure, automation and AI have been adopted over the years, but the human contribution to delivering services remains fundamentally large.
Why is the conversation shifting?
Now, platform players such as Salesforce, ServiceNow, and others are openly building towards this new model.
- Salesforce has recently proposed action-based pricing that can take decisions across workflows and charges users based on the number and complexity of action taken
- ServiceNow is integrating generative AI not just for support chats, but to drive real task execution in areas like Information Technology Service Management (ITSM) and case management
These aren’t proof-of-concepts anymore. They’re signaling a future where execution becomes intelligent, automated, and scalable.
What does this look like for service providers?
Let’s take a simple use case, sending a file to a bank via a File Transfer Protocol (FTP) server, something many L2/L3 production support teams still do manually.
Right now, a person might:
- Check if the file was generated from the scheduled job
- Open it, see if it has valid content
- If it’s empty, investigate why (maybe a previous job failed)
- If it’s valid, send it via FTP and log completion
Now imagine an autonomous system monitoring, interpreting, and actioning all of this:
- One agent checks the job status
- Another validates the file content
- A third one investigates exceptions and makes decisions
- And another handles the FTP dispatch
The human in the loop? Their role shifts to overseeing the process and handling only true exceptions.
This is what Systems of Execution enable…
What changes for service providers?
As an example, over the years, we have seen providers shifting from pure fixed fee model to having a hybrid model in systems integration deals. In many deals now, we see a hybrid model:
- 40% of the payment is tied to deliverables (Blueprint designed, Code release in test, UAT, Go-live, etc.)
- 30% is linked to time and quality-based Service Level Agreements (SLAs) (e.g., on-time delivery, <x% incidents in production during first 30 days, etc.)
- 30% is based on actual business outcomes
But as Systems of Execution mature, this will tilt even further:
- Outcome-based components will grow forming a higher proportion of the mix
- The emphasis on just “delivering on time” and “delivered within x effort” will reduce
- Providers will be rewarded for actual value delivered, not just for doing tasks. This value will be measured based on efficiency, effectiveness, and experience
Why is this change beneficial for service providers?
If done right, this shift can help providers:
- Reduce exposure to T&M and other effort-heavy models, which are subject to Third Party Advisor (TPA) benchmarking
- Improve margins by cutting dependency on linear labor
- Increase pricing power through differentiated value
Let’s agree to the fact that margins have been under pressure for years, especially as uncertainty continues to be the dominant reality for the past three years. Wage inflation, talent churn, and pyramid compression aren’t going away. Systems of Execution offer a path to de-risk delivery and rebalance the equation.
We can also expect service providers to take a context driven approach for their clients that helps them keep deals sole sourced.
How can service providers charge for systems of execution?
As service providers increasingly embed Systems of Execution (SoE) into enterprise transformation journeys, pricing strategies may evolve to match this shift.
Historical data from gen AI and Integrated automation adoption offers a useful precedent. Nearly 80% of implementations are bundled into broader service engagements or a black box, with nearly 15–25% structured as standalone instance fees.
However, if we compare the distribution and outsourcing lifecycle prior to automation and gen AI, the above figures mark a significant improvement.
A similar trend may emerge in SoE. Service providers are likely to monetize SoE through configuration-heavy agent deployments and customized orchestration layers, with modular fees tied to the complexity and number of agents involved.
These could include pricing for purpose-built context-aware agents, real-time orchestrators, and autonomous exception handlers. However, just as with gen AI, the path of least resistance in the early days will be bundling SoE into broader services. As the production use cases scale, however, providers may delink the platform services and provide SoE pricing separate from the labor to declutter the value created by each one of them.
Regardless of the pricing structure, enterprises will likely assess the value of SoE through familiar outcome-oriented metrics such as Employee Net Promoter Score (e-NPS), Customer Satisfaction Score (CSAT), cost savings, etc. What’s going to be different this time is that these improvements can be directly the value and impact of the decision making and execution.
The future of services delivery is not more people doing more things. Its smarter systems do more of the right things, with humans steering the boat and not rowing it.
Systems of Execution won’t just change how services are delivered; they’ll change how they’re measured, priced, and scaled.
The shift is already underway. The question is: how ready are buyers for this change and what can service providers do accelerate this change?
If you found this blog interesting, check out our From Interface To Orchestration: Why Underwriting Workbenches Must Evolve Into Systems Of Execution | Blog – Everest Group, which delves deeper into another topic relating to Systems of Execution.
If you have any questions or want to discuss the latest evolutions in Systems of Execution in more depth, please contact Achint Arora ([email protected]) and Manish Malik ([email protected]).
