In This Article
Culture is what people learn from watching how decisions get made. Here are six research-backed tests for whether yours holds up.
Workplace culture gets built in the ten minutes after a decision lands. That’s when people work out whether the rule was applied the same way as last time, whether anyone asked them first, and whether the reason they were given is the real one.
A review of 183 studies found that how fairly a workplace runs its decisions tracks with job satisfaction. That relationship was measured across 40 of those studies and nearly 32,000 people1. Here are the six tests a decision has to pass.
What Is Workplace Culture?
Workplace culture is the set of shared values, norms, and expectations that shape how people behave inside an organization. It’s learned mostly by observation, as new hires work out what gets rewarded, tolerated, and punished by watching what happens to everyone else.
That observational part is why culture resists the usual fixes. A company can run a full slate of programs and still teach its staff, in a single afternoon, that seniority beats performance, by promoting someone everybody knows was handed the easy accounts. Nobody announces that lesson, and it still gets learned.
Most advice on the subject heads for atmosphere instead: friendlier teams, more fun, a recognition wall by the kitchen. Those things are pleasant and worth doing, and if that’s what you came for, the fuller guides are employee engagement and having fun at work, plus team building questions and icebreakers for larger groups.
There’s a duller thing underneath all of it that has been measured carefully. Across a review of 183 studies, how fairly a decision process runs was one of the strongest correlates of both job satisfaction and organizational commitment, measured across tens of thousands of employees1.
It even edged out the decision itself. How people judged the fairness of the outcome mattered slightly less than how they judged the way it was reached. If you’ve ever had to deliver a decision nobody wanted, that’s the useful half of the finding, because the way it was reached is the part you control.
So if you’re trying to change a culture, the working definition is the accumulated record of how decisions got made and who they got explained to. The six criteria below are what that record gets judged against.
The 6 Tests of a Fair Decision
A theory of procedural justice developed in the 1980s, and carried into that same review, breaks fairness into six criteria a process has to meet before people will call it fair1. Each one is checkable, and it’s common to fail two or three without noticing.
1. Apply the Same Rule Every Time
The first criterion is consistency, across people and across time. The rule that decided last quarter’s bonus should be the rule that decides this quarter’s, and the rule applied to the person you like should be the one applied to the person you don’t.
Inconsistency is the cheapest fairness failure to commit and the most expensive to repair, because it teaches people that the stated rule isn’t the real rule. Once someone believes there’s a second, unstated rule, every future decision gets read through it.
Action Step: Pick one recurring decision you make: who goes to the conference, who gets the new laptop, who covers the holiday shift. Write down the rule you actually used last time. If you can’t, that’s the finding.
2. Keep the Decider Out of the Outcome
Hiring panels include the person whose headcount it is, promotion committees include the manager whose favorite is up, and vendor decisions include whoever the vendor takes to lunch.
Freedom from bias means whoever makes the call has no stake in which way it goes. It’s the criterion that gets nodded at and skipped.
You can’t always remove the interested party, and pretending otherwise makes the criterion useless. What you can do is name the interest out loud before the discussion starts. It’s a smaller move and a more honest one.
3. Get the Facts Before You Make the Call
Accuracy is a property of the inputs. It asks what information got gathered before the call and whether it was any good. A decision made on solid information can still turn out badly, and that’s a separate problem.
That distinction is why the process finding outran the outcome finding. A carefully reached decision that lands badly still tells people the next one will be handled properly.
Action Step: Before your next decision that affects someone else, write down the three things you’d need to know to get it right. Then check how many you actually know.
4. Build an Undo Button
Ask someone how a mistaken performance rating gets fixed at their company and watch the pause.
Every process needs a way to correct a decision that turns out to be wrong, and plenty don’t have one.
An appeal route can be completely informal, as long as it exists, people know about it, and using it costs them nothing.
Action Step: Name the correction path for one decision your team makes. If naming it takes more than a sentence, it isn’t one.
5. Ask the People It Lands On
This is the criterion most often confused with consensus, though it asks for something much smaller. The views of the people affected need to be on the table while the decision is still open.
Timing is the whole thing. Consultation after a decision is announced is a briefing, and people can usually tell the difference.
A standing group with rotating membership from across teams is one way to do it, because it gives you somewhere to take a decision while it’s still open. It works best when the membership actually rotates, so it doesn’t quietly become another committee of the same six people.
6. Hold the Line on the Ugly Calls
Layoffs, misconduct findings, a valuable client who treats your staff badly. How those get handled becomes the rule for everything smaller.
The last criterion is ethicality: the process conforms to prevailing standards of decency. It stays theoretical until following it costs something, which is the only moment it gets tested.
Try This: Think of the last genuinely uncomfortable call your organization made. Whatever it decided, that’s the standard your team now believes in.
Explaining Beats Being Pleasant
Two different things happen when a decision gets delivered, and they usually get treated as one.
The first is how someone is treated while it happens: politeness, dignity, respect. The second is what they get told, meaning why the procedure was used and why the outcome came out the way it did. Researchers separate the two, and they behave differently.
In the review, the explanation was the strongest predictor of how people evaluated the authority involved, whether that meant their individual manager or the organization as a whole1.
Being treated well ran nearly as strong for the individual manager. Where the two came apart was the organization: explanation held up, and politeness dropped off sharply. That last comparison rests on only a handful of studies, so treat it as a lead. But it points somewhere useful. Warmth seems to buy credit with the person in the room, while explaining buys credit for the place.
That’s worth knowing if you’re a well-liked manager. You can be kind, available, and genuinely on your team’s side, and still announce a restructure with no reasoning attached. What people take from that is that the reasoning wasn’t theirs to have.
The fix is cheap and easy to skip. When you announce a decision, spend more words on how you reached it than on what you decided. The decision itself is usually one sentence, and the reasoning is where the trust gets built.
Pro Tip: People usually want the shortlist more than the business case. Tell them what the other options were and why this one won.
How Managers Can Make the Process Visible
Gallup’s analysis of its own client data puts 70% of the variance in team engagement down to the manager2. That’s a lot of weight on one person. Plenty of managers hear a number like that and conclude the job calls for charisma they don’t have. The six criteria point somewhere more boring and more doable. Most of what a manager controls is process, and a lot of that process is invisible to the people it governs.
The work is writing the rule down.
Response times. Every team has an unwritten rule about how fast you’re supposed to reply. Nobody has ever seen it written down, so people infer it from the fastest replier and then resent them. Publishing the actual expectation converts a source of quiet friction into something anyone can point at.
Interruptions. Same shape. Protected focus time works when the rule about who can break it is explicit, and fails when it depends on how senior the interrupter is. That’s criterion one going wrong quietly.
What gets measured. Shifting from hours worked to outcomes delivered is good advice, and it only works if people know which outcomes. An unstated standard is harder to live with than a demanding one, because a demanding standard can at least be met.
There’s also a rule of thumb worth knowing for setting the standard at the right height. It isn’t a research finding, just a working heuristic. Examine your Commit: Complete ratio. Make a list of all the projects your team has committed to in the last year. How many have you accomplished? That’s your commit: complete ratio. It should NOT be 100%. If it is, you are under-challenging yourself. Aim for 70 to 80%. That’s the sweet spot for challenging yourself.
Language and hiring. Bias suppression is criterion two, and it applies to who gets considered as much as who gets chosen. Reviewing hiring criteria before a role opens is the cheap version. So is watching how people describe the groups they’re not part of.
It’s also worth noticing how the language in your job ads and reviews describes people. Here are some terms to look out for, borrowed from Buffer’s diversity guide.
You could also brush up on a few cultural differences in communication.
Recognition. Who gets thanked publicly is a decision like any other, and it tells your team what the organization actually values. There are fuller guides to making employees feel valued and employee incentives, and the fairness question underneath both is whether the criteria are visible.
Schedules. Flexibility is a decision about whose time counts, and it lands differently depending on whether people know the rule behind it. One of our team puts it this way:
I work from 5 p.m.-3 a.m. I know my night owl tendencies are on the end, but I feel incredibly grateful to write when I feel most creative instead of forcing my brain into a schedule that feels unnatural to me (which I have done for many jobs in the past!)
Remote arrangements are the same test at a larger scale. Whatever the policy turns out to be, what people judge is whether the reasoning was shared and whether it gets applied evenly. Remote work suits some teams and not others. There’s a practical guide to building a remote work schedule if you’re setting one up.
Action Step: Pick the most-guessed-at rule on your team and write it in one sentence somewhere people can find it. Response times, interruption etiquette, and how time off gets approved are all good candidates.
What to Do When You Don’t Make the Decisions
Everything above assumes you’re the one deciding. Plenty of the time you aren’t, and the useful question is what’s available from underneath.
The move that changes the most is a change of timing. Asking “why did that happen?” after a decision puts the other person in a defensive crouch, and what comes back is usually a justification. Asking “what’s this going to be weighed on?” before the decision puts the criteria on the table while they can still be influenced.
That’s criterion five, applied from below. It also works because it’s an easier question to answer. Someone who can’t yet defend a decision can almost always describe how they plan to make it.
Try This: In the meeting before the next decision that affects you, ask what it’ll be weighed on. That question is usually welcome even when the decision itself is contested.
Then run your own small decisions the way you’d like the big ones run. If you allocate work, say what you allocated it on. If you pick who presents, say why. It sounds trivial at the scale of a three-person team, and it’s how a norm gets established by someone with no authority to establish one.
Put your expertise somewhere other people can reach it. A lunch-and-learn where you teach the thing you’re good at looks like a nice-to-have, and it quietly serves criterion three, because it moves accurate information toward the people who need it before they make a call. Offering to host one is a small ask that reaches further than it looks.
And mentor someone new. Culture is learned by watching, which means a new hire’s first read on how this place works comes from whoever happens to be standing nearest. Volunteering to be that person is the most direct influence over culture available to someone without a title.
There’s a relationship dividend here too. Asking a colleague what they weighed, and then telling them what you weighed, is a specific and unusually easy kind of conversation to have. It comes with an agenda, which makes it more comfortable for people who find open-ended workplace socializing draining.
The ordinary version of being good to work with still counts, and there are plenty of ways to be a nicer person if that’s what you’re after.
One limit worth naming. Asking for criteria works in a workplace where criteria exist and someone is willing to say them out loud. Where decisions are genuinely arbitrary, the question surfaces that fact instead, which is useful information of a different kind.
Where Fairness Runs Out
A few limits worth having in mind before acting on any of this.
The evidence is correlational. Everything above describes relationships measured across large numbers of workplaces, which means fair process is associated with satisfaction, commitment, and people staying. It doesn’t establish that fairness causes them, and a company doing well on both could be doing well for a third reason entirely.
Some of the criteria are better evidenced than others, and the authors say so themselves. The findings on how people are treated rest on noticeably fewer studies than the findings on how decisions get explained. The review is also from 2001. It gathers 25 years of research up to that point and remains heavily cited, but a quarter of a century has passed since.
And fair process doesn’t rescue a bad decision. Someone who loses their job in a redundancy round handled impeccably has still lost their job. What the process changes is what everyone still there concludes about the next one.
Frequently Asked Questions About Workplace Culture
What are the key elements of a healthy workplace culture?
Research on procedural fairness points at six: decisions applied consistently across people and time, decision-makers with no stake in the outcome, accurate information gathered before the call, a route to correct a mistake, conduct that holds up under pressure, and input from the people the decision affects. Values statements and perks sit on top of those.
How does workplace culture affect employee satisfaction?
In a review of 183 studies, how fairly a workplace ran its decision processes was among the strongest measured correlates of job satisfaction and organizational commitment, across tens of thousands of employees. The relationship is correlational, which describes an association and stops short of proving a cause.
How can a positive workplace culture benefit a company?
The measured associations are with job satisfaction, organizational commitment, and lower withdrawal, meaning fewer people heading for the door. Those come from peer-reviewed research, and they are strong associations that stop short of guarantees. Separately, Gallup’s analysis of its own client data attributes 70% of the variance in team engagement to the manager, which is industry analytics rather than published research.
What are the challenges in developing a strong workplace culture?
The common failures are procedural. A rule that changes between quarters, a decision-maker with a stake in the result, a call made without the information to make it, and no way to appeal a mistake. Each of those is cheap to commit and slow to repair.
How long does it take to change a workplace culture?
Culture updates whenever a decision gets made, so the clock starts at the next real one. A single well-explained decision won’t shift a reputation built over years, though a run of them is where the change actually accumulates.
Can you improve workplace culture without a budget?
Most of what this research points at costs nothing: writing down the rule you already use, naming your interest before a discussion starts, saying what a decision was weighed on, and creating a way to appeal it. The parts of culture work that cost money are also the parts this research says least about.
Takeaways on Workplace Culture
- Pick one recurring decision you make and write down the rule you actually used last time. If you can’t reconstruct it, you’ve found where to start.
- Before the next call that affects someone, ask one person it lands on what you’re missing.
- When you announce a decision, spend more words on how you reached it than on what you decided.
- Name the one decision your team makes that has no correction path, and give it one.
- If you don’t make the decisions, ask what the next one will be weighed on before it gets made.
If you want the fuller version of what keeps people invested once the process is sound, the guide to employee engagement strategies picks up where this leaves off.
