Using metric trees during a pivot or crisis

Every measurement system is built on assumptions. Revenue targets assume a certain market size. Conversion rates assume a certain customer behaviour. Growth projections assume a certain competitive landscape. These assumptions are so deeply embedded in the metrics organisations track that they become invisible. Nobody questions whether monthly active users is the right metric to optimise when the product is growing steadily. Nobody debates whether pipeline coverage ratio matters when deals are closing predictably. The assumptions hold, the metrics work, and leadership teams develop an intuitive relationship with their numbers that allows them to steer the business effectively.

A crisis shatters those assumptions. When a pandemic forces an entire customer base to change their purchasing behaviour overnight, conversion rate benchmarks from three months ago become meaningless. When a key competitor collapses and floods the market with discounted inventory, your pricing metrics no longer reflect competitive reality. When a regulatory change bans your primary distribution channel, your customer acquisition cost is not just wrong; it is measuring a mechanism that no longer exists. The metrics are still being calculated, the dashboards are still updating, but the numbers have been severed from the reality they were designed to represent.

This is where a metric tree proves its value. Because a metric tree makes the causal relationships between metrics explicit, it allows leadership to identify exactly which assumptions have broken and which still hold. The tree does not just show you numbers. It shows you the model of how your business creates value. When that model breaks, the tree shows you where it breaks, which branches are severed, which connections are weakened, and which remain intact. That structural visibility is the difference between a panicked scramble to "watch all the numbers" and a disciplined triage that focuses attention on the metrics that matter most in the new reality.

The first action in any crisis is triage. In a medical context, triage means sorting patients by the urgency and nature of their condition so that limited resources are directed where they will have the greatest impact. The same principle applies to metrics. When a crisis hits, leadership teams face an overwhelming volume of signals: metrics spiking, metrics crashing, metrics behaving in ways nobody has seen before. The instinct is to monitor everything more closely. This instinct is wrong. It produces information overload at precisely the moment when the organisation needs information clarity.

Metric tree triage works differently. Instead of examining every metric individually, you walk the tree from the root downward, asking a single question at each node: does the causal relationship between this node and its children still hold? If the root metric is revenue, and revenue decomposes into new customer revenue and existing customer revenue, the question is whether both of those branches still represent the primary mechanisms through which the business generates revenue. If the crisis has eliminated one of those mechanisms entirely, perhaps new customer acquisition has halted because the sales team cannot conduct in-person demos, then that branch of the tree is severed. It does not need monitoring. It needs a fundamentally different response.

1. 1

   Walk the tree from root to leaves

   Begin at the top-level metric and move downward through each branch. At every node, assess whether the causal link to its children remains valid. A causal link is broken when the mechanism it represents has been disrupted by the crisis. A causal link is weakened when the mechanism still operates but at a fundamentally different scale or with different dynamics. A causal link is intact when the crisis has not materially affected the relationship between parent and child metrics.

2. 2

   Classify every branch as intact, weakened, or severed

   Intact branches continue to operate as before and should be monitored at their existing cadence. Weakened branches still function but their historical baselines are no longer valid; recalibrate targets and increase monitoring frequency. Severed branches represent mechanisms the crisis has eliminated; stop tracking them against pre-crisis targets immediately and either suspend them or replace them with crisis-specific alternatives.

3. 3

   Identify new branches the crisis has created

   Crises do not only destroy value-creation mechanisms. They sometimes create new ones. A company forced to close its retail stores may discover that its hastily launched e-commerce channel is generating unexpected demand. A SaaS business whose enterprise pipeline has frozen may find that small-business self-serve signups are accelerating. These emergent mechanisms need to be added to the tree as new branches, even if they are provisional, so they can be tracked and nurtured.

4. 4

   Establish a crisis-mode metric set

   From the triage exercise, distil a focused set of five to ten metrics that leadership will review daily or weekly during the crisis. These should include the root metric, the healthy branches that sustain the business in the short term, the weakened branches that require intervention, and any new branches that represent emerging opportunities. Everything else moves to a lower monitoring cadence. The goal is to reduce noise and increase signal at the moment when the organisation most needs clarity.

5. 5

   Communicate the triage to the entire organisation

   Share the results of the triage openly. Tell teams which branches of the tree are intact, which are weakened, and which are severed. Explain which metrics are in the crisis-mode set and why. This communication is not just informational. It is directional. It tells every person in the organisation where their effort matters most and where it does not. Without this communication, teams will continue optimising metrics that the crisis has rendered irrelevant, wasting effort the organisation cannot afford to waste.

Triage is not a one-time exercise. In a fast-moving crisis, the landscape can shift week by week. A branch that was intact last Monday may be severed by Friday. A branch that appeared severed may begin to recover as the crisis evolves. Schedule a brief triage review at the start of each crisis cycle, typically weekly, to update the classification of each branch and adjust the crisis-mode metric set accordingly. The tree gives you the structure to do this quickly; without it, each review would require rebuilding the assessment from scratch.

Some crises are temporary disruptions. The business model is sound; it just needs to weather the storm until conditions normalise. Other crises are permanent shifts that demand a fundamental change in how the business creates value. When the crisis requires a pivot, the metric tree must be restructured, sometimes from the root. This is one of the most consequential and least discussed aspects of organisational change: the moment when the North Star metric itself changes.

Changing the North Star is not a cosmetic exercise. It rewires the entire measurement system. Every branch beneath the old North Star was designed to decompose and drive that specific outcome. When the root changes, many of those branches become irrelevant, new branches must be created, and the remaining branches connect to the new root through different causal logic. The tree is not simply edited. It is rebuilt around a new model of how the business creates value.

Notice how this crisis-mode tree differs from a steady-state growth tree. The root is not revenue or market share. It is survival and recovery, a composite objective that balances short-term cash preservation with longer-term strategic repositioning. The left branch focuses on extending runway, buying the organisation time to execute the pivot. The right branch focuses on validating the new direction, ensuring that the pivot has a viable destination before the organisation commits fully.

This dual structure is essential. Organisations that focus exclusively on survival cut too deeply and destroy the capacity they need to pivot. Organisations that focus exclusively on the pivot burn through their runway before the new model is validated. The tree holds both imperatives in tension, making the trade-offs between them visible and navigable.

Restructuring the tree also forces difficult conversations about what to stop measuring. Pre-crisis metrics that were sacrosanct, the engagement scores, the NPS surveys, the feature adoption funnels, may need to be suspended entirely during the pivot. This is psychologically difficult. Teams have invested years in building these metrics and the processes around them. Letting them go feels like abandoning institutional knowledge. But attempting to maintain the old measurement system alongside the new one creates confusion, divides attention, and sends a mixed signal about whether the pivot is real. The restructured tree should be the organisation's single source of measurement truth during the crisis. Everything that is not on the tree is paused until the crisis resolves.

In steady-state operations, most organisations review their metrics monthly or quarterly. This cadence is appropriate when the business environment is relatively stable and changes unfold gradually. During a crisis, this cadence is fatally slow. A month is an eternity when customer behaviour is shifting weekly, when cash is being consumed faster than expected, or when a new revenue stream is growing in ways that require rapid scaling decisions. The review cadence must compress to match the pace of change.

The metric tree makes this compression practical. Without a tree, accelerating the review cadence means reviewing every metric more frequently, an exhausting and often counterproductive exercise that produces meeting fatigue without producing insight. With a tree, accelerated reviews focus on the crisis-mode metric set: the five to ten metrics identified during triage that represent the highest-leverage signals. The remaining metrics continue at their normal cadence. The organisation reviews more frequently without reviewing more metrics.

The key discipline is not just accelerating the cadence but protecting it. In a crisis, every day brings new fires that demand attention. Meetings get cancelled. Reviews get postponed. "We will look at the numbers tomorrow" becomes "we will look at them next week" becomes "we have not reviewed the tree in a month." This is precisely how organisations lose control during a crisis: not through a single catastrophic decision, but through the gradual erosion of the review discipline that would have surfaced problems while they were still manageable. Protect the cadence. It is the rhythm that keeps the organisation oriented when everything else is in flux.

A crisis creates an acute communication problem. Leadership is making rapid decisions, priorities are shifting weekly, and the broader organisation often feels left in the dark, uncertain about what has changed, what still matters, and what they should be doing differently. Traditional communication mechanisms, the all-hands meeting, the leadership email, the updated strategy deck, are too slow and too abstract to convey the specificity of change that a crisis demands.

The metric tree provides a communication medium that is both precise and intuitive. When leadership restructures the tree, that restructuring is itself a communication. Showing the organisation the old tree with branches greyed out and the new tree with its crisis-mode structure tells a story that no slide deck can match. It shows what the business was optimising for, what it is now optimising for, and how the two relate. It answers the question that every employee is asking during a crisis: has my work changed, and if so, how?

The tree is especially powerful for communicating with middle management, the layer of the organisation most likely to be caught between leadership decisions and team-level execution. A middle manager looking at the restructured tree can see immediately how her team's metrics connect to the crisis-mode objectives. She does not need to interpret a leadership email or translate a strategy deck. The connection is structural and explicit. This reduces the translation loss that plagues crisis communication and ensures that the pivot reaches the people doing the work, not just the people approving it.

Transparency during a crisis builds trust. When the organisation can see the same tree that leadership sees, with the same data and the same colour coding, it eliminates the suspicion that leadership knows more than they are sharing. The tree makes the state of the business visible to everyone. That visibility is uncomfortable when the numbers are bad, but it is far less damaging than the rumours and speculation that fill the void when information is withheld.

Every crisis ends. Markets stabilise. New business models find their footing. Customer behaviour settles into a new pattern. The temptation at this point is to revert to the pre-crisis measurement system as quickly as possible, to restore the old tree, the old targets, and the old review cadences as though the crisis were an aberration that can be set aside. This temptation should be resisted. The organisation that emerges from a crisis is not the same organisation that entered it. It has learned things about its business, its customers, and its resilience that the pre-crisis tree did not capture. Reverting to the old tree discards those lessons.

The rebuilding process should be deliberate. Start with the crisis-mode tree and evaluate each branch. Some branches were provisional, created to track emergent opportunities during the crisis, and should be formalised if those opportunities have proven sustainable. Some branches from the pre-crisis tree that were paused should be restored, but with updated baselines that reflect the new reality. And some branches, both crisis-mode and pre-crisis, should be retired because the business has genuinely moved past them.

The most valuable artefact of a well-managed crisis is the restructured tree itself. It documents every decision the organisation made under pressure: which metrics were prioritised, which were abandoned, which new mechanisms were discovered, and how the business model evolved. This is institutional knowledge of extraordinary value. Organisations that preserve it have a structural advantage the next time a crisis arrives, and in a volatile world, the next crisis is never far away.

Rebuilding also requires recalibrating targets. Pre-crisis targets are almost certainly invalid, and crisis-mode targets were set under emergency conditions that no longer apply. New targets should be set using the data accumulated during and after the crisis, not by reverting to pre-crisis benchmarks. A business that lost 30% of its revenue during a downturn and has recovered to 85% of pre-crisis levels is not "underperforming by 15%." It is operating in a new context that requires new baselines. The tree should reflect that context honestly rather than creating a false comparison to a world that no longer exists.

Finally, conduct a tree retrospective. Gather the leadership team and walk through the sequence of tree changes: the initial triage, the restructuring, the cadence acceleration, and the rebuilding. Ask three questions at each stage. What did we get right? What did we get wrong? What would we do differently next time? Document the answers not as a narrative post-mortem but as annotations on the tree itself, so that the next crisis response can begin with the accumulated wisdom of every previous one. The metric tree is not just a tool for navigating a crisis. It is the mechanism through which the organisation learns from it.