中文Ask a printing or packaging leadership team a simple question, “What has your AI investment returned?”, and watch the temperature in the room change. Not long ago, Amy Servi-Bonner, Vice President of Printing AI, began opening customer conversations with exactly that question, and she learned something uncomfortable: the room often goes quiet, not from thought but from the realization that no one has an answer. That silence, she argues, may be the most expensive sound in the building.
The point is counterintuitive. Most leaders fear the AI project that failed, the one that burned budget and was quietly killed. But failure gets examined. Failure generates questions, postmortems and accountability. Failure, as painful as it is, produces learning. The initiative that should actually keep leaders up at night is the one everyone thinks is working, the demo that ran, the room that applauded, the budget line that was approved, and then nobody ever asked what changed.
Changed, note, not what the tool does. What changed in labor hours, in throughput, in margin, in the actual numbers a CFO is paid to defend. Servi-Bonner insists this is not a technology problem but a leadership problem. Enthusiasm got the initiative started, but enthusiasm does not pay the bills, and momentum without measurement is not strategy, it is motion without a destination.
The printing and packaging industry runs on margins that do not forgive assumptions. When a leader cannot say whether an AI initiative is working, they are not managing risk, they are deferring it. And deferred risk in a margin-thin business has a way of arriving all at once, usually at the worst possible moment. The operators she worries about most are not the ones who say “we haven’t started yet,” those conversations are straightforward. She worries about the ones who say “we’re already doing AI” with full confidence and then cannot cite a single number that proves it.
Uninformed confidence, in her framing, is more damaging to an AI future than honest ignorance. Ignorance at least prompts a question. False certainty closes the door on one. So what is the right question? Not “Are we doing AI?” and not “Do we have a tool running?” The right question is “What has it returned, how do we know, and what are we building toward?” If a leader cannot answer all three parts, they do not have an AI strategy, they have an AI experiment, and experiments without measurement are just expensive habits.
This is not a counsel of pessimism. Servi-Bonner is clear that moving is better than waiting, and she does not want to strip credit from teams that built something real. But moving without measurement is not a plan. The AI Opportunity Sprint she references exists precisely for companies that have started and now need to know whether what they built actually works, what it would take to make it measurable, and where the real return still sits uncaptured.
For print and packaging executives, the practical discipline is straightforward but rarely practiced. Define the baseline before the pilot. Pick the metric that matters, labor hours saved, makeready time cut, spoilage reduced, quote turnaround shortened. Instrument the tool so the number is observable, not asserted. Review it on a cadence the CFO would respect. And be willing to kill what does not move the needle, just as rigorously as the failed pilots were killed.
The irony is that the industry’s strength, operational discipline on the production floor, transfers perfectly to AI if leaders choose to apply it. Shops measure ink, substrate and click counts to the decimal; applying the same rigor to an AI initiative should feel natural, not novel. The shops that do will separate genuine transformation from theater.
The most dangerous AI initiative in print, then, is not the spectacular failure. It is the quiet success story nobody can verify, quietly consuming budget and attention while delivering nothing measurable. Name the return, or name it a hobby.
Source: American Printer (Amy Servi-Bonner, Vice President of Printing AI, July 2026).
