Culture Works When Nobody Has to Step Away From Their Work

The most expensive lesson I’ve learned as a CEO cost less than a mid-level salary and looked, on paper, like a success. The lesson: If a system only stays alive because someone remembered to log in, it’s already dying.

My team adopted goal-tracking software that I still consider well designed. Our objectives were written clearly, assigned properly and reviewed on schedule. The dashboard stayed green for most of the year. What I didn’t see, because the dashboard didn’t show it, was that nobody opened the system between reviews. Goals got updated the evening before the review, right before we would discuss them. In the hours that made up the workweek, the software was describing a company that didn’t exist, and I was accepting the description because it was the color I wanted.

When the initiative quietly collapsed, initially I blamed the product. It took longer to accept that the product had done exactly what we asked. The failure was within the system and not the product itself. When employees need to readjust their work routine to just maintain software, eventually the system will fail.

I now apply that rule to every purchase, program and process at our company. Whatever survives a real week of work is something that appears inside the work itself.

The survey you’re proud of is still a scoreboard.

Our organization recently earned a high score on an anonymous employee trust survey. I’m glad we did it, and the certification that came with it matters in a field where families trust us with the last months of a loved one’s life. Third-party proof is worth having.

But I’ve become skeptical of what surveys are used for. A survey measures the outcome of decisions that were made months earlier. It is a scoreboard, and a scoreboard has never won a game. When a company’s first clear signal that a team is in trouble arrives through survey data, the real finding isn’t the trouble. It’s that the company’s instruments are positioned further from the truth than its own managers are, and a follow-up action plan won’t move them closer.

Nothing about a survey score comes from measurement. It comes from the parts of a week that no survey can see: managers who hold one-on-ones that actually happen, commitments that are followed until they’re either kept or explained and problems that people feel safe raising the day they notice them rather than months later, anonymously, in a comment box.

A survey should confirm what your managers already know. If it’s telling them something new, that’s the finding to act on.

The same rule applies to every tool you buy.

Earlier this year I created a new department with a single employee in it. Her mandate is narrow: Find the hours our teams spend on work a machine could do, and return those hours to the people who lost them. We provide care in patients’ homes, so a recovered hour isn’t an efficiency statistic. It converts directly into another visit.

The standard we set for anything she brings in is the one I think most leaders get backward. A new tool is not allowed to change how our people already work. Our clinicians and coordinators spend their day inside one collaboration platform, so any software we pilot has to appear there, in the flow that already exists. If it asks them to open a separate window, I already know how the story ends. I’ve paid for that ending once.

The order of operations matters as much as the tool. Before you automate a task, ask whether it should exist. Then simplify what’s left. Automation should come last, and only for work that has survived the first two questions. We’re currently running an AI tool and a plain rules-based automation against each other for the same job, and I have no preference about which one wins. An unremarkable tool people actually use beats an impressive one they abandon.

Culture is what happens between the meetings.

For years, I thought of culture as something to be measured and then improved with a plan. I now think it’s the reverse. Culture is the accumulated content of ordinary weeks: whether the one-on-one occurred, whether the commitment was tracked, whether the problem got raised on Tuesday or buried until the survey. Anything that lives outside those hours, whether it’s a values page, an objectives dashboard or a promising new platform, is competing for attention it will not receive.

If your culture strategy consists of a survey and an action plan, what you have is a scoreboard and a press release.

Ask these three questions before you launch anything.

1. 1. Where will it live? If the answer is a separate site or app, expect adoption to fade within a quarter. Put it inside whatever your people already open a hundred times a day, or reconsider launching it.
2. 2. What does it retire? Each new system should remove an old one. Adding without deleting hands your managers another chore, and they’ll pay for it in the trust of their teams.
3. 3. Would you have known without it? If a dashboard or survey is the only place a problem could have surfaced, the reporting isn’t the weakness. The management rhythm is. Repair the one-on-one before you upgrade the report.

The organizations I most respect don’t run better software or better surveys. They simply ask their people to visit fewer things and let more things happen where the work already is. That’s what management looks like when it’s real.