Beyond the Fixed Target Operating Model

A company can deliver the operating model it planned and discover that the assumptions behind it have changed.

Leaders still need to make choices about capabilities, responsibilities, processes, technology and controls. Large changes need architecture and coordination. A collection of unrelated pilots is not a strategy.

But a detailed target contains a large number of assumptions about work that has not happened yet.

What will customers need? What will people do? What can be delegated? Where will judgement and review matter? Which information will be available? How will the organisation respond?

When those assumptions change, the route should be allowed to change with them.

The challenge is to preserve direction without turning an early design into an obligation to arrive at the wrong place.

The configuration is a hypothesis

A business goal can remain important while the arrangement required to achieve it changes.

A customer function may need more capacity for useful conversations. The initial design might propose an AI assistant, new roles and a redesigned preparation process. Later evidence could show that fragmented information is the main constraint, or that a different handoff creates more benefit than the proposed tool.

The desired outcome has not disappeared. The current explanation of how to reach it has improved.

That distinction should be visible in the operating model. Which elements reflect durable intent? Which are necessary architectural commitments? Which are assumptions about how future work will be performed?

Treating the configuration as a hypothesis makes those assumptions discussable. It gives leaders a way to change the route without pretending that the original ambition was mistaken.

More than one good future

A preferred operating state need not be one exact arrangement.

A service function might seek faster resolution, sustainable workloads, effective controls and more capacity for complex customer issues. Several combinations of human work, agent delegation and software could satisfy those conditions.

The right level of automation is whatever helps produce that result within the agreed constraints. It is not necessarily the highest level available.

This matters because a single target can distort the assessment. Maximising agent use may increase checking. Minimising handling time may create repeat contacts. Reducing a visible cost may weaken resilience or move effort elsewhere.

The target should describe a set of acceptable conditions, with clear priorities where trade-offs arise. The organisation can then discover which configuration meets them most effectively.

This is a more useful ambition than specifying every future task in advance.

Plan ahead, commit in stages

Model-predictive control offers an engineering analogy for this approach. It considers future actions over a horizon, applies the immediate action and recalculates as the state changes. Its performance depends on the model and constraints; the transferable idea is the discipline of updating the plan. [1]

For a company, three levels of commitment are useful.

The first is the intent: the outcome, principles, constraints and accountable ownership that remain in place until leadership deliberately changes them.

The second is the next intervention: the change the company is ready to make, with a reason, an owner and a way to examine the result.

The third is the later path: the actions that may follow if the next move produces the expected consequences and the assumptions remain credible.

This is still a roadmap. It is simply clearer about which parts are commitments and which depend on what the organisation learns.

That clarity can improve governance. A team should not have to defend an obsolete later step merely because it appeared in the original plan.

Funding the learning cycle

The budgeting process can make this difficult.

An annual investment case often rewards a large, complete account of future benefits. A leader who wants to make a sequence of smaller evidence-led changes may struggle to explain the whole route before learning what the first change reveals.

The result can be more certainty in the funding document than exists in the operation.

A different approach would fund an outcome and a governed sequence of decisions. The initial commitment would support the next worthwhile intervention and the evidence needed to decide what follows. Further commitments would reflect the results, dependencies and remaining uncertainty.

That does not mean unlimited discretion or a permanent experiment budget. Financial accountability, architecture and decision rights still matter. The review would ask whether the latest evidence justifies the next commitment.

It would also make stopping a weak intervention a legitimate outcome. A team that discovers early that a mechanism does not work has improved the investment decision, provided the lesson is credible and used.

Let the operating evidence alter the route

The important moment arrives after the change reaches real work.

If an agent reduces preparation but increases review, the next step may concern review design. If an apparently inefficient control protects a valuable outcome, the design should preserve it.

These are changes to the understanding behind the roadmap, not merely adjustments to its dates.

The company needs a way to connect them to the original intent. What did we expect? What happened? What does it imply about the remaining route?

This is where a living operating picture becomes valuable. It keeps the plan in contact with how the business is actually functioning, rather than relying on a description assembled at the start.

A programme can retain its strategic direction while becoming much more responsive in its execution.

Adaptation should not mean permanent upheaval

People need reliable routines to do good work. They need time to learn, consolidate and understand what is expected.

A company that constantly changes instructions, roles and tools can create the very friction it is trying to remove.

The goal is continuous capacity to improve, not constant disruption. Some parts of the operation should be protected because the evidence shows that they work. Others can support bounded trials. Major shared dependencies may require slower, coordinated changes.

A useful review therefore includes the decision to leave something alone.

It also gives changes enough time to produce a meaningful signal. Immediate observations may reveal adoption or a failure in the mechanism; broader outcomes may take longer. The cadence should fit the decision rather than force every effect into the same weekly report.

Better steering includes knowing when to hold a course.

A company capability, not a programme method

Strategy research on dynamic capabilities examines how firms sense opportunities, seize them and reconfigure what they do. [2] The self-improving company builds on that broader ambition.

The opportunity is to connect strategic intent with a deeper understanding of enacted work, especially as humans, agents and software take on different parts of it. Changes can then be chosen against the current situation and evaluated through their consequences.

The company gains a repeatable way to ask what should change, what should remain stable and what it now understands better.

This should serve several kinds of transformation. AI may make the need especially visible, but the logic also applies to restructuring, service improvement, system implementation and changes in control or operating cost.

The direction stays ambitious

The ambition behind Teho is a company that can understand how it works and keep improving deliberately.

That requires more than a target picture. It requires an ongoing relationship between intent, evidence and action. Leaders need to see how the operation is changing, recognise when an assumption has failed and choose the next move with a better understanding of its likely consequences.

A target operating model can remain part of that system. It provides direction, coherence and a way to communicate the current design.

Its value increases when the organisation can revise it intelligently.

The future company should be able to pursue ambitious outcomes without pretending that every step can be known in advance. It can make a clear next commitment, learn from what happens and keep moving towards the business it wants to become.

References

[1] Mayne, D. Q., Rawlings, J. B., Rao, C. V. & Scokaert, P. O. M. (2000). Constrained model predictive control: Stability and optimality. Automatica, 36(6), 789–814. https://doi.org/10.1016/S0005-1098(99)00214-9

[2] Teece, D. J. (2007). Explicating dynamic capabilities: the nature and microfoundations of (sustainable) enterprise performance. Strategic Management Journal, 28(13), 1319–1350. https://doi.org/10.1002/smj.640