ADM Agreements Signed Today Must Be Ready for Autonomous Orchestration
For years, the ADM debate has been framed as a productivity story: automation would reduce effort, AI would help developers code faster, and providers would deliver more with fewer resources. This is true, but incomplete.
The larger shift is that applications are becoming active participants in their own management. Embedded AI agents can detect anomalies, resolve incidents, optimize workflows, generate code, configure processes and recommend business actions with limited human intervention.
That shift challenges one of ADM's oldest assumptions: applications are maintained by people, demand is measured through tickets and effort, and value is linked to the resources required to keep systems running. If applications increasingly manage themselves, that logic starts to break down.
For CIOs and sourcing executives, the issue is immediate. ADM contracts signed today may still be active in 2030 – but by 2030, we will likely be living in a world of self-healing operations, AI-generated development, autonomous testing and embedded agents. The risk is not simply cost. It is structural misalignment with how application services are delivered. An ADM agreement can be contractually alive and commercially obsolete.
The question is whether today's ADM model can support tomorrow's autonomous enterprise.
The Economics of ADM Will Shift from Effort to Autonomy
Traditional ADM contracts were built for human-centric delivery. Pricing linked to full-time equivalents, ticket volumes or predefined service activities made sense when human labor was the primary driver of cost and value.
AI weakens that proxy. As routine support, testing, configuration, documentation and development tasks are executed by agents, effort becomes a less reliable measure of value. Resource consumption becomes harder to justify when value comes from prevention, automation maturity, continuity and speed of change.
This is not a one-sided customer argument. Providers will still invest in platforms, orchestration, governance, monitoring, compliance, security and specialist talent. They need a model that rewards innovation and value creation. Organizations need one that reflects how work is performed as autonomy increases.
That is where outcome-oriented and autonomy-aware pricing become central. ISG's Autonomy-Level Pricing framework links price to execution maturity, human oversight, risk ownership and embedded governance. It asks how work is delivered, how much autonomy is used, what controls applied and where accountability sits. In the ADM market of 2030, autonomy will become a commercial variable.
Please fill out this form to continue.
Operations Will Move from Incident Response to Incident Prevention
ADM organizations have traditionally been designed around a reactive model: incidents happen, tickets are raised, support teams investigate and service is restored. AIOps, observability platforms and autonomous agents are changing that paradigm. Application environments can increasingly detect anomalies, predict failures, identify root causes and initiate remediation without manual intervention.
That changes operational excellence. The strongest ADM provider may no longer be the one that resolves the most tickets fastest. It may be the one that prevents incidents from becoming tickets at all. Future ADM contracts should therefore measure availability and repair times – but also prevention effectiveness, automation maturity, resilience, user impact and autonomous remediation quality.
The ADM Workforce Will Become Human-Agent
Traditional ADM organizations are structured around development, testing, support and operations. AI introduces a workforce model in which human experts and intelligent agents operate side by side.
Routine activities will increasingly be executed autonomously. Human resources will focus more on architecture, governance, exception handling, risk decisions, business enablement and continuous improvement.
This makes the familiar phrase “human in the loop” too generic. The better sourcing question is: where must human judgment remain commercially, operationally and legally non-negotiable?
In ADM, those areas may include architectural decisions, release approvals, business-critical workflow changes, security exceptions, regulatory interpretation, production risk acceptance and major incident decisions. Contracts should define these boundaries before the operating model changes, not after accountability becomes disputed. Managing the shift to human-agent collaboration will become a critical sourcing differentiator.
Development Will Move from Coding Capacity to Capability Orchestration
For decades, ADM capability was often associated with coding capacity and specialist technical skills. Those skills remain important, but they will not define the future of ADM alone.
Enterprise applications are increasingly configured, composed and orchestrated rather than programmed from scratch. Low-code platforms, AI-generated software, workflow orchestration and prompt-based interactions are reducing the relative importance of manual coding.
The competitive edge will shift toward translating business requirements into digital capabilities quickly and safely. That means governing AI behavior, orchestrating processes, managing data dependencies and ensuring automation supports the business model. The future ADM provider will orchestrate intelligent business capabilities across platforms, processes, data and agents.
Performance Management Will Move Closer to Business Experience
Traditional ADM performance management has focused on technical health: availability, incident volumes, defect backlogs, response times and mean time to repair. These remain important, but they provide only a partial view of business value.
Business leaders increasingly care whether applications help employees work without friction, customers use digital channels seamlessly, processes flow across systems and changes translate into faster outcomes.
This will accelerate observability, user journey analytics and experience-level agreements. The conversation will move from “did the system meet the SLA?” to “did the application environment enable the business outcome?” A provider measured on experience, resilience, automation and business impact can align with what the organization actually needs.
Sourcing Models Must Be Designed for Continuous Evolution
The sourcing implications are substantial. Many ADM agreements being negotiated still assume human effort, reactive support and static service definitions. That creates a risk that contracts become outdated long before they end.
Commercial models must evolve. ADM contracts should progressively link price to outcomes, autonomy levels, risk ownership and governance maturity. Autonomy-level pricing is particularly relevant because it creates a structured way to commercialize increasing automation without ignoring the cost and risk of autonomous execution.
Benchmarking must expand beyond pricing to technology adoption, automation maturity, self-healing capability, AI-enabled development productivity and resilience. Without that, customers may be locked into yesterday's assumptions while the market moves on.
Governance must become more adaptive. Contracts should address transparency, accountability, human-agent operating models, performance measurement and change mechanisms where an increasing share of operational activity is executed autonomously.
Finally, sourcing strategies should protect long-term flexibility. As ADM becomes more dependent on AI agents, platform configurations, workflow intelligence and data-driven automation, organizations need clarity on portability, interoperability and exit rights. The future exit question is whether the customer can retain the intelligence embedded in the application operating model.
Do Not Treat ADM 2030 as a Cost-Reduction Exercise
The strategic mistake would be to view ADM as another domain for cost reduction. The organizations creating the greatest value from AI will not only automate application delivery. They will redesign the operating model behind it.
By 2030, ADM will be less about maintaining applications as passive technology assets and more about orchestrating intelligent business capabilities across an autonomous technology landscape. That will require new commercial models, new governance disciplines, new performance metrics and a precise understanding of where human accountability must remain.
The call to action is now. ADM contracts signed today may shape application services for the rest of the decade. The greatest risk is not that applications manage themselves. It is signing long-term ADM agreements that cannot adapt when they do.
ISG helps companies negotiate and manage contracts that measure and govern the right sets of values – even in the rapidly changing era of AI. Contact us to find out how we can help you.
About the author
Dr. Daniel Gerster
Dr. Daniel Gerster ist Partner und leitet ISG‘s EMEA Sourcing Capability. Er treibt die kontinuierliche Verbesserung der Sourcing-Methodik, Vorgehensweise und Plattformen voran. Bei ISG hilft er Führungskräften, komplexe Herausforderungen im Zusammenhang mit IT, Technologie und digitaler Transformation erfolgreich zu meistern, um strategische Ziele zu erreichen und die operative Exzellenz zu verbessern. Dr. Daniel Gerster bietet umfassendes Fachwissen in allen beschaffungsbezogenen Fragen und Vertragsverhandlungen. Er ist regelmäßig an großen, komplexen und herausfordernden Sourcing-Projekten in Europa beteiligt.
