Ticket Taker vs. Problem Solver: Be Irreplaceable | Darrell Alfonso

Commencement speakers are getting booed off stages for telling graduates that AI is their future. Meanwhile, marketing ops teams are getting leaner, job listings are down, and professionals who built careers on technical fluency are quietly wondering whether that currency still has value. The tension between the official narrative and the lived experience has never been sharper, and most industry commentary is being written by people who have not recently sat on the wrong side of a layoff notice.

Darrell Alfonso has. He is the Marketing Operations Lead at the Impellam Group and, before that, led marketing operations at AWS and Indeed. He wrote the MarTech Handbook and publishes the Marketing Operations Leader newsletter. He is also someone who spent time unemployed while actively preaching the gospel of AI adoption, which gives him a credibility most commentators in this space simply do not have.

The companies laying off marketing ops professionals are largely not doing so because a model has automated the work. Darrell is direct about this: during the war for talent, tech companies overhired engineers and software developers specifically so competitors could not get them. There was no real plan for what those people would do once hired. When the bill came due, AI made a convenient scapegoat.

That framing matters for anyone currently job hunting. If you believe you lost your role because a machine replaced you, you will make very different decisions than if you understand you were caught in the fallout from years of inflated headcount and now-failed growth assumptions. The latter is not personal. It is structural, and structures are eventually correct.

The same dynamic plays out within companies that have not yet laid off employees. Darrell points out that when he returned from parental leave at Indeed, he had set up his team to operate without him. He came back looking for problems to solve. Others in equivalent positions were effectively walking the halls. That kind of bloat is not unique to any one company. It is endemic to large organizations where existing processes have outgrown their usefulness but have never been retired, and where it is still possible to hide behind silos.

One of the most practically actionable observations in this conversation is not a strategy tip. It is a diagnosis. Business leaders, with very few exceptions, cannot clearly distinguish between AI and basic automation. If-this-then-that workflow chains that have existed for a decade are now being called AI because leadership never had a working definition of automation in the first place.

Darrell's advice is not to die on that hill. The better move is to deliberately leverage the confusion. Data quality work that could never get budget on its merits now gets funded because executives are terrified of being left behind on AI, and good AI outputs require clean data. The pitch writes itself.

This is not cynicism for its own sake. It is a recognition that the political economy inside most companies rewards perceived alignment with executive priorities. If the actual work you need to do, building data infrastructure, cleaning up lead management, and establishing proper attribution, can be framed as AI readiness, the budget unlocks. The work was always necessary. The framing is just finally working.

The economic case for replacing humans with AI is weaker than the narrative suggests. MIT research found that AI is only cheaper than humans in 23% of tasks. Uber's CTO burned through the company's entire 2026 AI budget in four months on token costs alone. NVIDIA's VP of Applied Deep Learning has stated that compute now costs more than the people using it.

Darrell is an optimist about the long-term trajectory of those costs, and he is probably right that computing will follow the same curve as every other technology. But the current moment is not that future. Right now, the replacement narrative is built on math that does not survive contact with scale.

The more immediate risk is not that AI makes humans obsolete. It is that companies make chaotic, financially unsustainable AI investments while simultaneously cutting the operational staff who would otherwise prevent those investments from becoming expensive mistakes. The people who know which integrations are held together with the equivalent of tape, who understand why the data looks the way it does, who can tell the difference between a model hallucinating and a genuine insight, those people are not a cost center to optimize. They are the error-correction layer.

Darrell draws a clear line between two types of operators, and the line is not about seniority or compensation. It is about what someone actually does with their time. Ticket takers receive requests, fulfill them on the designated platform, and return the output. The work is real, but it is transactional, and it is the category most directly in the path of automation.

Problem solvers use technology, people, and process as instruments toward a business outcome. They notice that the project management tool is being used in five different ways across five teams, slowing everything down, even though fixing it is technically outside their scope. They ask what success looks like in six months and work backward to the technology and process that would produce it. They are the ones who, when laid off, leave a dumpster fire behind because they were the integration user, the admin, the connective tissue.

The distinction is not permanent. Ticket-taking is often a circumstantial state, not a permanent identity. People get stuck in transactional work when they are in environments that do not create space for anything else. The more important question for any individual practitioner right now is whether they are actively building the pattern-recognition and problem-framing skills that place them in the second category, because the first category is genuinely at risk.

Darrell's argument that marketing ops should function as the COO of marketing is not a morale-boosting metaphor. It describes something that is already happening by necessity. As tactical execution is absorbed by automation, the residual value of an ops professional lies in orchestration, judgment, and cross-functional leverage. That is what a COO does.

The practical version of this is not glamorous. It looks like noticing that the team's project management tool is creating friction because no one agreed on conventions, then fixing it without being asked, because you can see the downstream cost, and the CMO will not. It looks like being the person who understands what the marketing leader needs and translates that into operational reality before the gap becomes a problem. It looks like knowing which AI-generated output is actually good and which one will embarrass the company.

The counter-argument worth taking seriously is that when tactical skills are automated, every ops professional's instinct is to claim the strategic high ground, and at some point, the C-suite may decide that AI-generated strategic recommendations are sufficient without the operational complexity associated with a human. Darrell's response is that we are not at that stage, that the actual revenue drivers of any business still require human judgment at the point of commitment, and that what gets automated is internal efficiency work, not the decisions that move the needle. That is probably right for now, though the window for staking out that higher ground is not unlimited.

Darrell Alfonso is the Marketing Operations Lead at the Pelham Group and the author of the Marketing Operations Leader newsletter on Substack.

Special thanks to Claude for helping to summarize this conversation.