Breaking the Box

01

The box is the current model of doing business

The box is not bureaucracy, an organisation chart or a poor process. It is the set of boundaries and assumptions within which the organisation creates value: the customers it serves, the needs it addresses, the products it offers, the capabilities it owns, the activities it buys from others, the markets it reaches and the risks it accepts. Each boundary was once a reasonable answer to a question about scale, cost, control or focus.

Those answers may still be right. A manufacturer that buys its energy from the grid, contracts out transport and concentrates on a few product families may be well designed for its market. The question is not whether the boundaries are old, but whether they still fit the value the organisation intends to create as customer needs, technology, costs and risks change.

02

Improving execution is not the same as changing the model

There are two kinds of improvement. The first works inside existing boundaries: a better price, a different product mix, a new production sequence, a second supplier, faster invoicing or a more reliable forecast. The second changes the boundaries themselves: a new production line, a new market, a revised portfolio, an acquired capability or a different split between what the organisation owns and what it sources.

The two differ in more than scale. Improvements in execution are usually reversible, quick to test and owned by one or two functions. Changes to the model typically commit capital for years, alter the risk profile and require new capabilities and decision rights across the organisation. They deserve a different kind of analysis and a different level of governance.

Improvement within the model should usually be tried first. If a constraint disappears when the schedule, mix or terms change, there is no case for moving the boundary. A case for change begins only when reasonable improvements within the model have been tried and the limit persists.

03

Success can make the box harder to see

Early success is valuable, but it has a side effect. When automation, better planning and tighter processes improve results, they also confirm the assumptions on which the model was built. Metrics, incentives and investment plans align around what is working. The organisation becomes more efficient and more confident inside a box that may be slowly ceasing to fit.

Research on established companies has described this pattern. Clayton Christensen showed how well-managed firms that listen closely to their best customers and invest in improving existing products can lose ground to technologies those customers did not initially want [1]. The lesson is not that improvement is a mistake. It is that better performance inside existing boundaries does not, by itself, prove that those boundaries remain right.

A useful discipline is to ask, alongside every significant efficiency gain, which assumption it reinforces. Does the improvement depend on demand staying in its current segments, on energy remaining within a certain price range, or on a partner continuing to provide a capability on today’s terms? If so, that assumption belongs on the list of things to test.

04

Look for persistent signals, not single events

A large contract, a competitor’s move or a supply disruption can expose a weakness. The event is a reason to test the model, not proof that the model has failed. More telling are patterns that persist: recurring demand that existing capacity or channels cannot serve economically; customers repeatedly asking for outcomes the current offer does not provide; a constraint that remains after reasonable operational fixes; a capability that is essential to the value proposition but unreliable or costly to secure externally; and a lasting shift in the economics of energy, logistics, capital or technology.

No single signal decides the question. Each raises one: is this temporary friction, or evidence that the model no longer fits? The connected view described in the first article helps here, because it shows whether a signal stays local or travels through capacity, quality, delivery and cash.

05

Stress tests show where the model reaches its limits

Scenario analysis and stress tests can address two different questions. The first is how far the organisation can go within its current model. At what volume is capacity exhausted? When does a bottleneck force a change in production priorities? Which product mix preserves quality and contribution? At what point does additional inventory begin to strain liquidity?

The second is what would need to be true before changing the model makes sense. What level of sustained demand would justify a new line? Under which conditions would a new market support the investment? When would a partnership be preferable to building a capability internally? Which assumptions about supply, prices or customer behaviour would have to hold?

Consider a hypothetical manufacturer whose main plant runs two shifts across several product families. A stress test might show that the plant can absorb roughly a fifth more volume through a third shift and better sequencing, but that beyond that point changeovers, maintenance windows and overtime begin to erode quality and contribution. Another scenario might show that a shift in demand towards one product family would make a dedicated line attractive only if volumes hold for several years and energy costs stay within a certain range. Neither result tells the company to build. Both tell it what to watch and which assumptions carry the decision.

That is the right use of a threshold: a signal for review, not an automatic trigger. When sustained demand, utilisation or cost crosses a level agreed in advance, the question returns to the management team with the evidence attached. Scenarios are not forecasts; their value lies in changing how decision-makers see a situation, not in predicting a date [2]. The threshold itself can also be revisited as evidence accumulates. A model that appears to say exactly when to expand usually hides the assumptions that matter most.

Persistent signal → stress test → review threshold → leadership judgement → staged change or no change → learning

06

Boundary decisions take several forms

Once leaders understand a persistent limit and the value they intend to create, the options are broader than ‘expand or not’. The organisation may invest in capacity, through a new line, a new site or a long-term lease. It may enter a new market or segment, which changes customer risk, product requirements, compliance and logistics. It may change the product portfolio, adding an offer customers repeatedly request or withdrawing one that consumes scarce capacity without creating enough value. It may acquire a capability through hiring, a partnership or an acquisition. Or it may change what it owns, controls or sources from partners.

The last question is often the most consequential. Vertical integration can increase control over an earlier or later stage of the value chain, while adding capital needs and management demands. Outsourcing can increase flexibility and access to expertise, while reducing direct control over a capability that may matter to customer outcomes. Neither is inherently better. The useful question is not simply ‘make or buy?’ but which capabilities are central to the value the organisation intends to create, which it needs to control, and which it can access reliably through others. The answer can change as technology, scale and markets change.

The distinction between operating choices and boundary decisions matters here. Involving a logistics partner to fulfil one contract is an operating choice within the current model. Deciding to outsource logistics entirely, or to build an in-house fleet, changes what the organisation controls after the contract ends. That is a boundary decision.

07

Energy, storage, logistics and infrastructure can move the boundary

Some boundary decisions are less visible because they sit in infrastructure rather than in products or markets. Yet energy systems, storage, logistics and physical infrastructure can change both what the organisation does and how its economics behave.

For the same hypothetical manufacturer, on-site generation combined with battery storage could reduce exposure to peak electricity prices and allow energy-intensive production to be scheduled more flexibly. That changes more than the energy bill. It alters the cost of running an extra shift, the attractiveness of certain product mixes and the resilience of the plant. It also adds capital, maintenance, regulatory obligations and a capability the company may not yet have. The decision moves the organisation partly into the business of managing energy.

Logistics and storage work in a similar way. Additional warehousing near key customers can shorten delivery times and support entry into a new market, but it ties up capital in inventory and property. A shared or contracted network keeps capital light but makes service levels depend on a partner’s priorities. Each choice redraws the line between what the company runs and what it relies on, and each shifts the balance between fixed and variable cost.

08

What leaders need to assess before committing

Before redrawing a boundary, leaders need a view that goes beyond a single return figure. Value over time comes first. A move that improves next year’s margin may weaken future flexibility; an investment that depresses near-term profit may create a stronger position over the life of the asset. Capital and cash follow closely: how much is committed, when it is spent, how working capital changes and how the move affects funding headroom.

Risk changes shape as well as size. A new market brings customer and regulatory risk; a new asset brings utilisation and technology risk; a partnership brings dependency risk. Capabilities matter because a boundary change often requires skills the organisation does not yet have, and building them usually takes longer than buying equipment. Integration demands are easy to underestimate: new processes, systems, data definitions and working relationships must connect to the existing decision architecture.

Decision rights need to be settled before the move, not after it: who owns the evidence, who can approve the next step and who answers for the outcome. Reversibility deserves explicit attention. A market test can be stopped at limited cost; a factory, an acquisition or a long-term infrastructure contract cannot. Finally, learning: what the organisation expects to find out, how it will know, and how that knowledge will shape the next decision.

09

Stage the move when uncertainty is high

Breaking the box does not require a single irreversible leap. When uncertainty is high, the organisation can stage the decision: test willingness to pay with a limited offer, pilot a partnership before committing to it, reserve capacity rather than build it, or invest in phases with conditions defined in advance for proceeding, pausing or stopping. Investment theory has shown that when a commitment is irreversible and its future value uncertain, the flexibility to wait, proceed in stages or stop has value of its own [3].

Staging only works if each step produces evidence for the next decision. That requires the conditions to be set in advance, the outcomes to be measured honestly and the authority to stop to be real. Embedded Governance supports this. The management team agrees strategic boundaries, investment limits and risk appetite; accountable owners run tests within them; and results that cross agreed limits return to the level that can reassess them. Authority should follow the scale, impact and reversibility of the decision.

AI can help by synthesising market signals, comparing scenarios and monitoring results against assumptions. It cannot decide what kind of organisation the company should become or which risks it should accept; those remain business judgements.

10

Redraw the field deliberately

The Connected Enterprise helps an organisation see its dependencies and play better within its current boundaries. Breaking the box begins when leaders use that understanding to ask whether those boundaries still serve the organisation’s purpose, and then decide, with evidence and accountability, what to change.

The aim is neither to defend the current model nor to abandon it at the first sign of strain. It is to preserve the capabilities that still create value, change those that no longer do, and prepare the next field of play before circumstances force the choice.

11

References

[1] Christensen, C. M. (1997) The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail. Boston, MA: Harvard Business School Press.

[2] Wack, P. (1985) ‘Scenarios: uncharted waters ahead’, Harvard Business Review, 63(5), September–October.

[3] Dixit, A. K. and Pindyck, R. S. (1994) Investment under Uncertainty. Princeton, NJ: Princeton University Press.