Many leadership teams say they want IT to be more strategic, but the calendar tells a different story. A service interruption, an urgent access request, or a vendor problem can consume the same people who were supposed to improve the customer experience or prepare the business for its next stage of growth.
The answer is not to separate “operations” from “innovation” and hope the conflict disappears. It is to make the different kinds of technology work visible. A three-speed IT model gives leaders a practical way to protect reliable operations, fund focused improvement, and create room for selective innovation without allowing the loudest request to set the entire roadmap.
Why one speed creates a leadership problem
IT work is often managed as one queue. A password reset, a compliance remediation, a Microsoft 365 change, a workflow improvement, and an AI pilot compete for attention as if they had the same urgency, risk, and value. That creates predictable distortions:
- Important maintenance is deferred because it is less visible than a new project.
- Strategic work is repeatedly paused, so every improvement takes longer than planned.
- Innovation becomes a side project that depends on individual enthusiasm rather than a funded business decision.
- Leaders see activity and ticket counts, but not the capacity tradeoffs behind them.
Research and practitioner guidance describe the technology operating model as the way an organization enables the enterprise, engages with it, and delivers value. Gartner frames the outcomes as enabling IT efficiency, enhancing business performance, and transforming the business. The point is not that every company needs an elaborate enterprise framework. The point is that leaders should deliberately choose how technology work contributes to the business instead of letting the default queue make the choice.
The three speeds: run, improve, and reinvent
The model uses three speeds. They are not three departments, three vendors, or a rigid percentage allocation. They are three different promises that technology makes to the business.
1. Run: protect the promise
Run work keeps the business dependable today. It includes service desk support, identity and access, endpoint health, backups, monitoring, patching, vendor coordination, and the routine changes that keep employees productive. Run is not low-value work. Reliability is a business capability, and weak fundamentals eventually tax every other initiative.
Run needs clear service expectations, a visible backlog, owners for recurring problems, and measures that show whether the environment is becoming more stable. Useful signals include repeat incidents, time to restore service, unresolved high-risk findings, backup recovery tests, and the number of manual steps required to complete common requests.
2. Improve: remove friction from the current business
Improve work makes an existing capability faster, safer, easier to use, or less expensive. Examples include simplifying onboarding, consolidating duplicate applications, automating an approval workflow, improving reporting, or turning a recurring incident into a documented and monitored service.
Improvement work should be close enough to an operating pain that the result can be measured in weeks or months. Give each effort a business owner, a baseline, and a definition of done. A useful improvement project changes a workflow and leaves the team with a better standard, not just a new configuration that nobody owns.
3. Reinvent: test a new way to create value
Reinvent work explores a meaningful change in how the business serves customers, makes decisions, or delivers its product. It may involve an AI-enabled workflow, a new digital service, a data capability, or a different way of coordinating work across teams. The risk is higher because the organization is learning, not simply implementing a known pattern.
Reinvent work needs a bounded hypothesis, an executive sponsor, a safe test environment, and an explicit decision date. The goal of a pilot is not to prove that a tool is impressive. It is to learn whether a defined business outcome is achievable at an acceptable level of cost, risk, and operational complexity.
Allocate capacity by promise, not by habit
There is no universal ratio for the three speeds. A business recovering from a ransomware event, a merger, or a failed migration may need to put most of its capacity into Run. A company with stable operations but a major growth target may need more Improve work. The allocation should reflect the business condition and be revisited as evidence changes.
Start by listing the active work in one portfolio and tagging each item Run, Improve, or Reinvent. Then ask four questions:
- What promise does this work protect or create? Connect the item to reliability, employee productivity, customer experience, revenue, risk reduction, or a strategic capability.
- Who owns the outcome? IT can own delivery and service quality, but a business leader should own the result when the work changes how a team operates or serves customers.
- What capacity does it consume? Include implementation, security review, training, vendor management, support, and the follow-on work required after launch.
- What would we stop to make room? A roadmap is credible only when it shows the tradeoff. If everything is a priority, the organization has not made a decision.
Review the portfolio monthly for blocked decisions and emerging risks, then reset the allocation quarterly. When an experiment proves value, move it into Improve and eventually Run with an owner, support model, documentation, and measurable service expectations. When an experiment does not prove value, close it deliberately and record the lesson.
Design the handoffs between speeds
The most common weakness in a three-speed model is the gap between an idea and a dependable service. A pilot may work for a small group, but no one has planned identity controls, data retention, monitoring, training, licensing, or recovery when the capability reaches the rest of the company.
Define the handoff criteria before the pilot starts. A capability moving from Reinvent to Improve should have a named business owner, a documented workflow, a security and privacy review, a support path, adoption measures, and a clear estimate of ongoing cost. A capability moving from Improve to Run should have service documentation, monitoring, escalation rules, backup or rollback procedures, and a review date.
This is especially important for AI-enabled tools. A human-and-agent workflow changes who performs a task, what information a system can access, and how exceptions are handled. Treat the workflow design, approval boundaries, data sources, and audit trail as part of the service. Do not wait until after purchase to decide whether a person must review an output or approve an external action.
Make the model visible to the leadership team
Technology becomes more strategic when leaders can see the choices behind the work. Bring a short three-speed view to the operating meeting:
- Run: What reliability, security, and support signals require attention?
- Improve: Which two or three workflow changes will produce measurable value next?
- Reinvent: Which bounded experiments deserve funding, and what decision will end each experiment?
- Capacity: What is consuming the team, and which tradeoff needs an executive decision?
Deloitte describes strong technology operating models as a changing blend of modes rather than a single target state. That is a useful leadership test: the model should be stable enough to create accountability but flexible enough to change when the business changes. A growing company does not need to imitate a large enterprise. It needs a simple system for deciding what must be dependable, what should get better, and what is worth learning next.
The three-speed model is ultimately a discipline of attention. It keeps operational reliability from becoming invisible, keeps improvement from being crowded out by emergencies, and keeps innovation from becoming a collection of disconnected experiments. When every initiative has a speed, an owner, a promise, and a next decision, IT can help the business move with intention instead of merely reacting to the next request.
PTG helps growing businesses build that operating rhythm through managed IT, cybersecurity, and technology leadership guidance. The goal is not more technology for its own sake. It is a dependable foundation and a decision system that turns technology capacity into business progress.
