There are two ways a manager can get a team badly wrong. The first is to ignore the numbers and run on gut feel, which usually means you find out about a problem three months after it started. The second is to stare only at the dashboard and forget that every figure on it was produced by a person who had a week, a mood, a workload and a reason. Both failures look different on the surface. Underneath, they’re the same mistake: trusting one source of information and ignoring the other.
Most of the conversations I have about Garrett O’Rourke leadership topics eventually land here. Managers want to know whether to trust what the report says or what the person in front of them says. My answer is unsatisfying but true: you need both, and the skill is knowing which one is telling you more in a given moment.
What running sales teams and call centers taught me about measurement
Call-center operations are about as measured as business gets. Calls, talk time, conversion rate, abandon rate, revenue per hour, cost per acquisition. If you want to know exactly how a business is performing hour by hour, that’s the environment for it. And measurement earns its place — you cannot manage a sales floor on impressions. Numbers catch things your eyes miss. They catch the quiet drift where a team’s conversion rate slides two points a week for a month and nobody notices because everybody is busy.
But over the years I’ve also watched managers make confident, disastrous decisions using clean data. A rep with weak numbers gets cut, and the following quarter the team’s performance dips further, because that rep was the one training the new hires and nobody counted that. A campaign is scaled up because cost per acquisition looks great, when the real reason is that a handful of experienced closers happened to be on shift. The numbers weren’t wrong. The interpretation was.
That’s the part people underestimate. Metrics are accurate and incomplete at the same time. They tell you what happened. They rarely tell you why it happened, and “why” is the only thing you can actually manage.
Why the two sides get out of balance
Measurement is comfortable. It’s defensible. If you cut someone because their numbers were bottom of the board, no one can accuse you of favoritism. If you promote based on a spreadsheet, you look objective. A lot of managers lean on data not because it’s the best tool for the decision but because it protects them from having to make a judgment call in public. That’s not rigor. That’s cover.
The opposite failure has a friendlier face. Some leaders are genuinely good with people — warm, loyal, well-liked — and they let that stand in for management. They know their team is trying hard, so they excuse results. They confuse effort with output. That works for a while, right up until the business runs out of room, and then those same well-liked managers have to make cuts that a little honesty six months earlier would have prevented. Being kind in the short term is often the cruelest thing you can do. I’ve written more about that tension in how to approach hard business decisions.
There’s also a structural reason things drift. The higher you go in an organization, the more your view of reality is mediated by reports. You stop hearing calls. You stop sitting in on training. Your picture of the business becomes a monthly deck. It happens gradually and it feels like progress, because that’s what senior roles are supposed to look like. But if the only input you have is summarized data, you have quietly given up half your instrument panel.
Practical ways to manage both at once
Use numbers to ask questions, not to hand down verdicts. When a metric moves, the right first response is curiosity. Why did this change? What changed around it? Is this a person problem, a process problem, a product problem or a market problem? The report tells you where to look. It doesn’t tell you what you’ll find.
Pick a small number of metrics and stay with them. Teams that track forty things track nothing. Choose the handful that genuinely predict outcomes for your business and keep them consistent over time, because consistency is what turns a number into a trend. A trend is information. A single reading is noise.
Separate the score from the diagnosis. A rep’s conversion rate is a score. Whether they’re struggling with objection handling, burning out, or working a worse lead source is a diagnosis. You get diagnoses by listening to calls, sitting with people and asking direct questions. No dashboard will do that part for you.
Keep a channel to the floor. Whatever your role, build a routine that puts you in contact with the actual work — the customer conversation, the site visit, the support ticket. Not to micromanage, but to keep your judgment calibrated. When you’re responsible for a team, the fastest way to become a poor decision-maker is to lose touch with what the work actually feels like.
Be explicit about what’s measured and what isn’t. If coaching newer teammates matters to you, say so, recognize it, and factor it into how you evaluate people. Otherwise your team will optimize for what shows up on the board, and they’d be rational to do it. People do what gets counted.
Let people own outcomes, not just activities. Measuring activity produces activity. Measuring outcomes and then genuinely handing over the decision rights produces performance. That’s the heart of real delegation and accountability — you give someone a target, the resources, and the authority, then you hold them to the result rather than the method.
Judgment is the part that can’t be automated
The hard cases are always the ones where the two signals disagree. The numbers say cut; your read of the person says give it one more quarter. Or the numbers say everything is fine and something in the room tells you it isn’t. There’s no formula for that. What you have is judgment, and judgment comes from having been wrong before and remembering what it cost. It’s why I think experience shapes business judgment more than any framework does. Pattern recognition is earned, not downloaded.
That also means admitting the limits of your read. You will misjudge people. You will trust a number that turns out to have been built on a bad assumption. A leader who can say they’re not sure yet, and who goes and finds out, is more useful than one who is always certain — which is roughly the argument in why good leaders don’t need every answer.
As an investor, I find the same discipline applies away from the office. You can model a property or a business all day, but somebody has to judge whether the assumptions holding up the model are reasonable. The math is necessary and it isn’t sufficient. That’s a personal perspective from my own experience in real estate investing and public markets, not individualized financial advice — everyone’s situation and risk tolerance differ.
The broader principle
Numbers measure the business. People produce it. A leader who respects only the first becomes a scorekeeper; one who respects only the second becomes a friend with a title. Neither runs anything well for long.
As President of Commercial Development Group, and across years of sales, operations and business leadership work from Miami Beach, the most reliable pattern I’ve seen is this: good measurement makes you honest, and good listening makes you accurate. Keep both instruments working, and you’ll make fewer decisions you regret — which, in the end, is most of what management is.
Photo by Sebastian Herrmann on Unsplash
