In the late 1990s, Gallup ran a workplace survey on more than a million workers, looking for the questions that best predicted whether a team would succeed. They tested hundreds of factors. The single strongest predictors of engagement weren't salary, title, industry, or perks. The two that mattered most were oddly specific. Do you have a manager who cares about you as a person? And, more controversially: do you have a best friend at work? It is the single most underrated input in any happiness tracker, and the thread running through everything we've written about what makes a career genuinely happy.
When Marcus Buckingham, who led much of that work, presented these findings to executive rooms in the early 2000s, the second question made people laugh. "Best friend" sounded too informal, too soft. Buckingham held his ground because the data was unequivocal. The question wasn't sentimental. It was sorting for something deep. Do you have at least one person in this building you can speak to honestly, without political cost? If yes, almost everything else gets easier. If no, almost nothing else fully compensates.
It's almost never the company.
The cliché ("people don't leave jobs, they leave managers") is real, but it's a softer version of the truth. People don't leave companies. They leave specific rooms with specific people in them. They also stay in objectively difficult jobs because of two or three coworkers who make the days bearable. The unit of workplace happiness, looked at honestly, is rarely "the role" or "the company." It's a small handful of specific humans.
Most of us spend eight to ten hours of every weekday with the same small group. That's more hours than most of us spend with the people we say we love. If two or three of those people lift the room, the room is lifted. If they drain it, it's drained, regardless of what's on the screen in front of you.
"The room you spend forty hours a week in is mostly other people. Notice which ones leave it warmer."
The big-shift trap.
When the people are wrong, the temptation is to leave the whole job. New company. New industry. Fresh start. Sometimes that's right. Often it's an overcorrection. The same statistical pattern follows people across companies because the next workplace has its own version of the bad manager and the same number of difficult personalities. The fix that actually works, in the research and in our own data, is almost always more targeted. Change the team, the project, the manager, or the reporting line. Rarely the company.
Amy Edmondson at Harvard has spent two decades on what she calls psychological safety. She originally got interested in it because she was studying medical error rates in children's hospitals, and found a counterintuitive pattern. The teams that reported the most errors weren't the worst teams. They were the best. The good teams felt safe enough to surface mistakes. The struggling teams hid them. The differentiator wasn't competence. It was whether you could speak up in front of your peers without paying a personal cost.
A psychologically safe team, condensed into behaviors:
- You can ask a basic question without feeling stupid.
- You can disagree without it becoming a personal cost.
- You can admit you got something wrong and the room treats it as information, not ammunition.
- Decisions are explained, not just announced.
- Credit is distributed accurately.
Those five behaviors, in a manager or in a team, predict almost everything else about how a workplace feels. They are also rarer than they should be, which is why the same advice keeps getting written.
The Who lens.
The Designing Your Life course at Stanford uses an exercise called AEIOU to break down what's actually moving your energy at work. One of the letters is "U" for users (the people involved in any activity). A compressed daily version uses just one lens for it: Who.
Tag Who on the moments that stood out for a fortnight, good or bad. Don't think too hard about it. By the end, your top givers list and your top drainers list are populated by specific people. Sometimes a manager. Sometimes a colleague you'd never have flagged in advance. Sometimes the surprise is who's missing from either list. People you spend a lot of time with who just don't move the needle at all.
What to do once you can see it.
The pattern that comes out almost always points to a targeted move, not a dramatic one. If your data shows the activities are still giving you energy but the Who column is consistently red, you know what to negotiate (a different team, a new reporting line, a project change) and what not to bother negotiating (the role itself, the pay band). If two specific people on your list are consistently green, that's a relationship worth protecting. More 1:1s. More lunches. More shared work. If two are consistently red, you have a choice you can finally make on data instead of resentment.
Sometimes the workplace won't change. The manager isn't going to evolve in your timeframe. The team is what it is. That's a real situation, and a daily check-in is one of the more practical tools for it, because it lets you separate "the work" from "the people around the work." If the work is also draining you, the question becomes a different one. Either way, you've stopped guessing.
A small thing to try this week.
For the next ten working days, when you log a moment that stood out (good or bad), name one person who was in it. Just one. By the end of the run, you'll have a small, honest list of who lifts you and who doesn't. Most of the time, what to do with that list will be obvious.
The friction is the logging itself. If it takes too long, you'll skip the days when the room was hardest, which are exactly the days you most need on the record. You also need the Who tag to be explicit, not buried in a paragraph you'll never reread.
Lantern is built for that specific problem. Two minutes at the end of the day. Tag Who on the moments that moved your energy. After two weeks, your patterns view ranks people alongside activities and environments. You'll see which manager, teammate, or client is actually moving the needle, separate from whether the work itself is fine. That's the data you need to negotiate a new team, protect a friendship at work, or leave with clarity.