Why Commercial Leaders Are the Last to Adopt AI

The data is now unambiguous. McKinsey research published across 2023 and 2024 estimates that generative AI could unlock between $0.8 trillion and $1.2 trillion in annual productivity across sales and marketing placing the function among four that capture roughly three-quarters of the technology's total measurable value. Salesforce's most recent State of Sales report shows 83% of sales teams using AI saw revenue growth in the past year versus 66% of teams without it, with overall adoption now at 87% across surveyed organisations. The largest gains are concentrated in research, qualification, proposal generation, and pipeline analysis.

And yet sales is also the function adopting AI most slowly at the leadership level.

Every CRO survey published in the last eighteen months tells the same story. Tooling budgets are approved. Pilots are launched. Vendors are paid. But adoption stalls one or two layers beneath the senior leader, plateaus inside a quarter, and quietly becomes another line item that nobody can fully account for at the next board review.

The standard explanations have been exhausted. Change management. Training. Tool sprawl. Integration debt. Rep resistance. Each of these is real, and none of them is sufficient. Comparable change management challenges in finance, marketing, and operations have not produced the same plateau. Something specific to commercial leadership is causing the lag.

The argument I want to make is uncomfortable, and I'll state it plainly.

Senior commercial leaders are the slowest adopters of AI not because of resistance to technology, not because of skill gaps, and not because of organizational friction. They are the slowest because their personal authority, the thing that got them promoted, the thing they are paid for, the thing that distinguishes them from the people who report to them, was built on the precise category of judgment that AI now replicates at acceptable quality and dramatically lower cost.

This is not a tooling problem. It is an identity problem. And until the field names it, every AI-for-sales rollout will continue to underperform regardless of how good the technology becomes.

What sales leaders actually got paid for

To understand the resistance, you have to understand what senior commercial leaders were actually selling to their organisations for the last thirty years.

Not revenue. Revenue is the output. Revenue is what every function claims credit for in some form.

What sales leaders were paid for was structured judgment under uncertainty. The ability to look at a pipeline of fifty deals and say with conviction which twelve would close this quarter. The ability to read a buying committee and identify the real decision-maker before the third meeting. The ability to walk into an account and assess, within an hour, whether the relationship was salvageable or terminal. The ability to look at a comp plan and predict which behaviours it would actually produce versus the ones it claimed to incentivise.

This judgment was rarely written down. It was rarely teachable. It was the slow accumulation of pattern recognition across hundreds of deals, dozens of teams, multiple economic cycles, and several cultures. It was the asset that justified the title, the compensation, and the seat at the executive table.

It was also, until recently, the only place in the organisation where this kind of judgment lived.

That is the part that has changed.

A senior commercial leader who has spent twenty years developing pattern recognition across enterprise deals can now sit beside a junior analyst with eighteen months of experience and a competent AI agent and the analyst, on most operational sales tasks, will produce decisions of comparable quality in roughly a tenth of the time. Not on every dimension. Not on the relational ones, not yet on the political ones, not on the ones that require physical presence in a room with a buyer who is uncertain. But on a wide enough set of tasks, qualification, account research, proposal structuring, pipeline analysis, forecast probability scoring, win-loss pattern detection that the value of two decades of accumulated judgment has been visibly compressed.

The senior leader sees this. The senior leader, in fact, sees it earlier and more clearly than anyone else, because the senior leader has the longest reference frame for what good judgment used to require.

This is the source of the lag. It is not that senior commercial leaders don't understand AI. It is that they understand exactly what AI implies for the basis of their authority, and they are quietly trying to manage the transition without dismantling the asset their career was built on.

The pattern of resistance

Once you accept this framing, the patterns of resistance start to read differently.

The senior leader who insists that AI tools "aren't accurate enough yet" and who continues to insist this six months after every benchmark has improved is not making a technical claim. They are making an identity claim. The threshold of acceptable accuracy is being unconsciously calibrated to a level that protects their own indispensability.

The senior leader who delegates AI evaluation to a junior team member and treats the output as advisory is not delegating. They are insulating. The structure ensures that AI never produces a decision that competes with their own judgment in front of the people they manage.

The senior leader who champions AI adoption publicly but never uses the tools personally is performing the most common version of this. They earn the political credit for being forward-thinking while protecting the daily practice of their authority from contact with a system that might match it.

None of this is conscious in the way that explicit resistance is conscious. Most of these leaders genuinely believe they are making prudent judgments about technology readiness, organizational capacity, or implementation risk. The identity protection is operating below the level of explicit reasoning, which is precisely why it is so difficult to address through standard change management.

You cannot solve an identity problem by training. You cannot solve it by tooling. You cannot solve it by mandate. The leader has to renegotiate their own answer to a question that has no comfortable resolution: what am I actually paid for, now that the thing I used to be paid for is no longer scarce?

The functions that adopted faster — and why

The contrast with adjacent functions is instructive.

Finance adopted AI faster, despite comparable disruption to analyst work, because the senior finance leader's authority was never primarily based on the analyst-level judgment AI replicates. The CFO's authority is based on judgment about capital allocation, risk posture, and stakeholder communication domains where AI is currently a tool, not a substitute. The CFO can hand AI to the analyst layer without losing personal indispensability.

Marketing adopted AI faster because senior marketing leaders had spent the previous decade making peace with creative work being augmented or partially automated. The identity transition had been gradual. By the time generative AI arrived, the senior CMO's authority had already migrated upstream into brand strategy, market positioning, and audience architecture work that AI complements rather than competes with.

Sales is structurally different. The senior commercial leader's authority sits much closer to the ground. It is exercised in pipeline reviews, deal coaching, account strategy, and forecast calls and these are exactly the activities where AI is now most capable. The migration upstream that finance and marketing leaders had already partially completed has not yet happened in sales. Most senior commercial leaders are still operating in proximity to the work AI can do, which means the identity collision is direct rather than mediated.

This is why "AI for sales" rollouts plateau in a way that "AI for finance" or "AI for marketing" rollouts do not. The senior layer of the function is in direct competition with the technology, and no amount of training will resolve that competition. Only a redefinition of what the senior layer is for will resolve it.

What senior commercial leadership has to migrate toward

If the operational judgment that built sales leadership careers is no longer the scarce asset, the question becomes: what is now the scarce asset that justifies senior commercial authority?

The answer, I think, is becoming visible in the leaders who are moving fastest. It is not a single skill but a cluster of four:

The first is commercial judgment at the institutional level — the ability to assess not whether a single deal will close, but whether a market segment is structurally viable, whether a category is repricing, whether a competitive set is consolidating, whether a buying behavior shift is permanent or cyclical. This is judgment AI cannot yet replicate because it requires synthesis across information streams that are not in any single dataset, and it carries career risk that AI cannot bear.

The second is trust architecture — the ability to design organizational and customer relationships that compound trust over time. This includes the design of seller-buyer relationships, the design of account governance, the design of escalation paths, and the design of internal incentives that produce trustworthy commercial behavior. AI can transact. AI cannot yet underwrite trust at scale, and the leader who designs the system within which AI operates retains an authority AI cannot reach.

The third is narrative leadership the ability to translate commercial reality into language that aligns boards, investors, employees, and customers around a coherent forward direction. This is the function that has historically been undervalued in sales leadership and is now becoming primary. AI produces forecasts. Boards still need humans to explain why the forecast moved.

The fourth is the public dimension of leadership the ability to represent the function externally, build talent pipelines through visibility, and contribute to industry discourse. For thirty years this was optional for commercial leaders. It is now structural. The senior leader who is invisible outside their own organization has no defense against AI compression, because internal indispensability alone is no longer sufficient.

These four are the migration path. None of them are AI-replicable in the next decade. All of them require senior leaders to consciously distance themselves from the operational work they used to do, even though that work is what their identity was built on. That distance is the thing most senior commercial leaders find hardest, and it is also the thing the function now requires of them.

What boards and CEOs should be asking

For those who oversee commercial functions, the implication is concrete.

The standard adoption metrics for AI in sales tool deployment, license utilization, time-saved estimates are measuring the wrong layer. They measure adoption beneath the senior leader. The variable that actually predicts whether a commercial AI investment will produce returns is whether the senior commercial leader has personally migrated their own work upstream. Until that has happened, the function will adopt only the parts of AI that do not threaten the existing authority structure, which is precisely the parts that produce the smallest returns.

The right diagnostic question is not "is the team using the tools." It is "what does the senior commercial leader spend their time on, and is it the same work AI is now doing." If the answer is that the leader is still spending their time on deal coaching, pipeline review, and forecast adjudication, the AI investment is being absorbed into the existing structure rather than transforming it. The leader has to be doing different work for the function to do different work.

This is uncomfortable to say out loud, because it implies that some senior commercial leaders cannot make the migration. Some cannot. The leaders whose authority was most tightly bound to operational judgment, and who have not built the four upstream capacities, will be displaced over the next thirty-six months not by AI directly, but by the next layer of leaders who made the migration earlier and now look more current to the people who promote.

The honest version

I have spent twenty years in commercial leadership in hospitality, across markets where the gap between leadership generations has compressed faster than in most industries. The pattern is consistent. The leaders who survive structural shifts are not the ones with the strongest grip on operational judgment. They are the ones who recognize, earlier than their peers, that what made them indispensable yesterday is becoming common and who voluntarily walk away from that work toward something less defended.

AI is the largest version of this pattern most commercial leaders will face in their careers. The temptation is to defend the existing authority, slow the adoption, and hope the technology plateaus before it reaches the work the leader still does. It will not plateau. It is already past the work most senior commercial leaders are still doing daily.

The function with the most to gain from AI is adopting it slowest because the people in charge are protecting an asset they should already be selling. Until commercial leadership names that openly until senior leaders can say, in their own organizations, "the work I used to do is now being done by tools, and my job is to do the work that isn't" every AI rollout in the function will continue to under-deliver.

The question every senior commercial leader should be asking themselves this quarter is not which AI tools to deploy. It is which part of their own work they are no longer the cheapest source of. And then what they are going to do, deliberately and visibly, with the time that question opens up.

Sources

McKinsey & Company, The Economic Potential of Generative AI: The Next Productivity Frontier (June 2023).

McKinsey & Company, An Unconstrained Future: How Generative AI Could Reshape B2B Sales (September 2024).

McKinsey & Company, The State of AI in 2025: Agents, Innovation, and Transformation (November 2025).

Salesforce, State of Sales Report, Sixth Edition (2024) and Seventh Edition (2026).