Every HR transformation I have worked on has had a moment that no one schedules on the project plan. It arrives a few weeks after go-live, after the dashboards are running and the training videos have been watched and the vendor has rolled off to the next client. Someone in the business — usually a manager, often in operations, rarely senior enough to have been in the steering committee — sends an email that begins, "I know this is probably just how it works now, but..." What follows is a description of a thing the old system let them do with a phone call, and the new system does not let them do at all.
The email gets forwarded. It lands in a backlog. It is not, strictly speaking, a defect. The transformation is still a success by every measure the steering committee agreed to track. But something was in the building that is no longer in the building, and the manager felt it first.
I have been inside these programs, in one role or another, for more than twenty years. I have lived through the waves: the first proper HRIS rollouts, when we finally retired the filing cabinets and learned what "single source of truth" was supposed to mean; the migration to the cloud, which was sold as liberation and delivered, in many cases, a more elegant cage; the people-analytics era, which promised that data would make us wise and mostly made us busier. Each wave arrived with a manifesto and a vendor and a keynote. Each one promised transformation. Each one delivered, to be fair, a great deal. But each one also left behind that manager's email, and the backlog it went into, and the slow accumulation of things that used to be possible and quietly are not.
I have come to believe that the industry, my industry, does not have a vocabulary for what gets lost in these programs. We have a rich vocabulary for what gets gained. We measure experience, usually through a survey that was designed by the vendor whose system we just bought. We measure cost, usually against a baseline that was itself a rough estimate. We measure future readiness, which is the hardest of the three to pin down and therefore the easiest to claim. These three — experience, cost, future readiness — are the triad on which nearly every HR transformation in the market is evaluated. They are not wrong. They are simply incomplete.
What they miss is the organization's culture, which is to say the accumulated set of small agreements and rituals and workarounds through which people actually get their work done, and through which they understand themselves to be seen. Culture is not what the values poster on the wall says. It is what happens when someone needs an exception and has to decide whether to ask for one. It is whether the manager in operations sends the email or decides not to bother. It is the difference between a company where people believe the system, on the whole, is on their side, and a company where they have learned it is not.
I have watched transformations that delivered on all three legs of the triad and nonetheless left the organization measurably colder. The experience scores went up, because the new interface was prettier. The cost came down, because four roles were consolidated into two. The future-readiness narrative wrote itself, because the logos on the architecture slide were now the right logos. And yet, a year later, the people in the building moved more carefully. They asked fewer questions. They stopped believing that raising a concern would lead anywhere. The transformation had succeeded, and something in the culture had been withdrawn from, and no one on the program had a line item for it.
This is the pattern I have learned to watch for. It rarely announces itself. It almost never shows up in the post-implementation review. It reveals itself in the texture of small interactions, six months and twelve months and two years out — in the quality of the questions people ask in town halls, in the tone of the reviews written by people who have now left, in whether the mid-level manager still sends the email or has learned not to.
The reason it happens, I think, is not that anyone intends it. It is that the program was designed against a ledger that does not contain the relevant column. You cannot protect what you are not counting. And when the pressure comes — when the timeline slips, when the budget tightens, when the executive sponsor needs a win before the quarter closes — the things that are not on the ledger are the things that get traded away. The demo gets polished. The change-management workshop gets shortened. The exception the manager was going to request gets routed into a backlog instead of a conversation. Each trade is small. None of them, individually, is a scandal. In aggregate, over the course of an eighteen-month program, they are how a culture gets thinner.
I am writing this, and what will follow it, because I believe the current wave — the one now being sold under the banner of artificial intelligence — is going to make this pattern more acute, not less. The efficiency gains available are real. The temptation to spend them on more volume, more speed, more dashboards, will be enormous. Most of the market will take that path by default. It is the path of least friction, and it is the path the vendors are already pricing for.
There is another path. It involves spending the dividend differently. It involves treating transformation less as a construction project — something you stand up, roll out, deploy — and more as a cultivation, something whose success is measured partly by what grows in the soil afterward. It requires counting things the industry does not currently count, and protecting things the industry does not currently protect. It is slower in the places where slowness matters and faster in the places where speed is cheap. It produces, at its best, organizations that are more capable and also more themselves.
I do not yet know how to write a short answer to the question of what that path looks like in practice. I know it well enough to work from, not yet well enough to summarize. What I can do, in the pieces that follow this one, is trace the question through the places history has already asked it — in the studios that survived the arrival of photography, in the workshops that refused to be absorbed by the factory, in the kitchens that slowed down when the rest of the world was speeding up. Each of those moments contains something useful for the one we are now in.
The manager's email, in the meantime, is still sitting in a backlog somewhere. I think about it often. I have come to think the work of the next decade, in my field, is learning to answer it.
I write from the work I do at Clover and Myrtle, where we practice HR transformation as cultivation rather than construction.