The COO role is badly misunderstood.
Most people still see it as execution support. Some see it as operational discipline. Almost nobody sees it for what it is becoming: one of the most critical design roles in the AI enterprise.
That is about to change.
For years, the COO’s job was to make growth repeatable: define OKRs, build the playbook, coordinate functions, cut rollout costs, and stop scale from becoming chaos.
At Dukan, that meant reducing the cost of each new country launch until expansion felt less like rebuilding the company and more like extending it. The goal was not to recreate the business market by market, but to keep a strong core and make adaptation repeatable.
The old COO scale model was relatively simple: learn locally, codify centrally, then scale through controlled duplication.
That logic worked — until process started replacing outcomes.
Jeff Bezos gave that logic a name: Day 2.
Day 2 is stasis. Followed by irrelevance. Followed by excruciating, painful decline. Followed by death.
Jeff Bezos
He also warned that in large organizations, the process can silently become the product — and the real question becomes: do we still own our outcomes, or has the process taken over?
The old COO scaled through controlled duplication. The new COO scales through a portfolio of operating choices.
1. AI changes the economics of scale
This is where the economics of scale start to change.
It lowers the cost of documentation, translation, routing, local rule interpretation, structured decision support, and repetitive coordination work.
At Digital Virgo, international scale often depended on telecom constraints, local regulations, and market-access conditions that could make or break expansion. Much of the operating burden came from absorbing highly structured local complexity. Today, a significant share of that normative and documentation-heavy work can be processed faster by AI precisely because it is rule-based and repetitive.
This changes the economics of scale.
The COO no longer has to choose only between hard standardization and exploding local complexity. A stronger core can now coexist with more adaptive edges.
2. The COO now orchestrates distributed intelligence
That matters even more because intelligence is no longer concentrated only at the top.
Employees generate ideas faster. Customers produce more usable feedback. AI agents summarize, compare, prioritize, and translate. Signals from product, tech, sales, support, and operations are easier to capture.
The COO’s new advantage is the ability to turn scattered signals into aligned roadmaps across tech, product, sales, and operations.
At Hack40, one recurring lesson across scale-ups and transformation programs was simple: good ideas were rarely missing. What was missing was the operating layer able to turn distributed signals into synchronized execution. In the AI era, that synchronization can become dramatically faster — if someone designs it.
AI can reduce the coordination tax of scale and increase the usable surface of feedback.
3. International scale becomes a portfolio of integration models
This is especially true in international growth.
Working across AI scale-ups in Europe and Asia made one lesson impossible to ignore: global expansion was rarely just a matter of opening a country and repeating the same go-to-market motion. Security requirements, integration expectations, procurement logic, and local trust conditions often changed the nature of market entry itself.
In some markets, the right path was a prescriber. In others, a strategic partner. In others, a local investor. In others, a security-first integration model.
The COO of the AI era does not only scale operations. The role designs the market-entry architecture through which scale becomes possible.
4. Cross-cultural execution becomes more codifiable
Scaling internationally has never been only about language, regulation, or time zones.
It has also been about interpretation:
- priorities
- hierarchy
- escalation
- autonomy
- deadlines
- feedback
- local reading of the same playbook
The same operating model is never interpreted the same way everywhere.
At Dukan, the challenge was not just country-launch mechanics. It was making sure the same business intent could survive different local contexts without dissolving into improvisation. At Digital Virgo, working across continents meant living with different interpretations of urgency, risk, partnerships, reporting, and execution norms. At Hack40, this became even more visible in cross-border innovation programs, where the same recommendation could land very differently depending on local culture.
AI does not remove that complexity. But it can lower the cost of absorbing it.
AI agents can increasingly support translation, contextualization, clearer handoffs, role-based communication, and more consistent execution across heterogeneous teams.
5. Four operators every COO should learn from
What is often implicit needs to be made explicit.
Claire Hughes Johnson
At Dukan, I learned this directly: if country-launch logic, product constraints, local adaptation rules, and decision criteria stay implicit, international growth turns into reinvention instead of repeatability. At Hack40, the same held true across startup and corporate transformation programs — unclear operating rules create noise, politics, and execution drift.
Leadership is about making others better as a result of your presence and making sure that impact lasts in your absence.
Sheryl Sandberg
That matters because the COO is not only a process owner. The role is also to build leverage through people. At Digital Virgo, scale did not come only from structure. It came from building teams and leaders who could absorb pressure, make local trade-offs, and keep performance moving without waiting for constant central intervention.
Inventory is fundamentally evil.
Tim Cook
This is about more than supply chain discipline. It reflects a deeper operator logic: reduce dead weight, shorten cycles, and keep the system close to real demand. At Dukan, this translated into reducing the structural cost of each new market entry. At Hack40, it often meant cutting away unnecessary steps, duplicated conversations, and badly aligned roadmaps so momentum was not trapped in operational inventory.
No a–hole policy.
Gwynne Shotwell
It sounds simple, but under extreme complexity, culture is not soft. It is operating infrastructure. At Digital Virgo, under multi-country pressure, the wrong interpersonal dynamics could destroy coordination far faster than any process gap. At Hack40, especially in multicultural teams and entrepreneurial environments, trust, speed of feedback, and the ability to disagree without poisoning execution were often decisive.
6. The COO becomes an AI-enabled operator
AI is not only changing the organization around the COO.
It is also changing the COO’s personal operating layer.
A strong COO can now use AI to:
- synthesize feedback faster
- compare signals across markets
- prepare arbitrations
- reformulate playbooks by geography
- challenge roadmap sequencing
- reduce documentation drag
- improve handoffs between product, tech, sales, and operations
This does not replace judgment. But it increases leverage.
At Hack40, I’ve seen how much time leaders lose not because they lack ideas, but because they lack the operating layer that turns signals into synchronized action. AI can compress a meaningful part of that friction.
The COO of the AI era is not only managing an AI-enabled organization. They are becoming an AI-enabled operator.
7. The blind spots
This article would be incomplete without the blind spots.
AI does not replace management. It can reduce coordination costs. It does not eliminate the need for judgment, arbitration, and trust-building.
Not everything should become a playbook. The wrong COO reflex is to turn every new possibility into a new layer of control.
Speed is not absorption. The danger is not only moving too slowly. It is moving faster than the organization can absorb.
More data does not automatically mean more perception.
An AI-enhanced operating model can become more informed and less perceptive at the same time.
Not all scale is codifiable. Some tasks are structured enough for AI to help absorb them. Others still require political judgment, cultural reading, and human intuition.
The best COO of the AI era is not the one who automates everything.
It is the one who knows what must remain human.
Conclusion
The COO is no longer only the guardian of stability.
They become the architect of absorbable transformation.
The companies that will scale best in the AI era will not be those with the most agents. They will be those with the clearest operating logic for turning intelligence into action.
Final question: in your organization, who designs the rules that make distributed execution viable?
Originally published on LinkedIn. Join the discussion →
