I was three hours into a retrospective that had devolved into blame-shifting when the most senior engineer on the team finally spoke up. “Look,” he said, “we can keep pointing fingers at the data model, or we can admit we don’t actually trust each other enough to have an honest conversation about what went wrong.”

The room went quiet. He was right.

That moment stuck with me because it exposed something I’ve seen destroy more data teams than bad architecture ever could: the absence of genuine connection between people who spend forty-plus hours a week depending on each other.

We obsess over tool selection. We debate medallion architectures until our Teams threads reach scroll-fatigue territory. We write detailed runbooks and build elaborate monitoring systems. But we rarely invest in the thing that actually determines whether a data team succeeds or fails—whether the people on it feel like they belong to something worth sacrificing for.

A while back, I stumbled across a deep dive into Gregg Popovich’s leadership approach. If you don’t follow basketball, Pop coached the San Antonio Spurs to five NBA championships across twenty-two consecutive playoff appearances—the most sustained dynasty in modern professional sports. His winning percentage over three decades is unmatched in North American sports.

And his secret? It wasn’t X’s and O’s. It was wine, tears, and showing up at a player’s hotel room at 2 AM to cry with him after his father died.

What I found wasn’t a basketball story at all. It was a masterclass in building teams that actually work. And everything in it applies to leading data engineers—if you’re willing to do the uncomfortable work.



The problem with how we lead data teams

When managing data team’s: we’ve inherited a leadership model built for factory floors and sales quotas, then awkwardly grafted it onto creative, highly collaborative technical work.

The standard playbook goes something like this:

Set clear goals. Define KPIs. Track story points. Measure velocity. Hold quarterly reviews where you show charts that go up and to the right (or explain why they didn’t).

Provide feedback. Annual performance reviews. Maybe some lightweight 360 input. A few uncomfortable conversations about “areas for growth.”

Remove blockers. Clear the technical debt backlog. Negotiate for headcount. Shield the team from stakeholder noise.

None of this is wrong, exactly. But it’s not sufficient.

Data teams live in a peculiar tension. The work is deeply technical—you need people who genuinely understand distributed systems, query optimization, and schema design. But the outcomes are entirely social—dashboards nobody uses, pipelines that don’t answer the questions stakeholders actually have, data quality that erodes trust with every “Why do these numbers look different?” conversation.

Success requires people who can hold both realities simultaneously. And that kind of cognitive flexibility only emerges in environments where people feel safe enough to be vulnerable, connected enough to sacrifice personal metrics for collective success, and invested enough to push back when the team is heading in the wrong direction.

Popovich understood this intuitively. His teams consistently featured players who took less money, accepted smaller roles, and subordinated personal statistics for championships. How? Not through inspirational speeches or performance bonuses. Through dinners, emotional presence, and a relentless refusal to separate the person from the player.



Belonging cues: the invisible architecture of high-performing teams

Daniel Coyle, who studied Popovich extensively for his book The Culture Code, identified three types of signals that create what he calls “belonging cues”:

  • Personal, up-close connection — behavior that communicates “I care about you as a person”
  • Performance feedback — relentless criticism that says “we have high standards here”
  • Big-picture perspective — conversations that say “life is bigger than this immediate task”

Most data leaders nail one of these and completely miss the others.

The technically-focused leader provides excellent feedback on code quality and architecture decisions but treats personal connection as optional—something that happens organically in Teams channels or maybe at the occasional team happy hour.

The people-focused leader builds genuine relationships but struggles to deliver hard feedback, allowing mediocrity to persist because confrontation feels mean.

The visionary leader connects everything to grand purpose—“we’re democratizing data access across the organization!"—but the lofty framing feels hollow when day-to-day work involves debugging pipeline failures at 10 PM on a Friday.

Popovich weaves all three together constantly. The same evening might include crying with a player over a family tragedy, reviewing film and yelling about a missed rotation, and discussing the Civil Rights Movement. As one assistant coach put it: “He’ll tell you the truth, with no bullshit, and then he’ll love you to death.”

This isn’t contradictory. It’s complete.

Sound familiar? It should. Google’s Project Aristotle research found essentially the same thing—that psychological safety (the belief that you won’t be punished for taking risks or admitting mistakes) was the single strongest predictor of team performance. But psychological safety isn’t warm fuzziness. It’s the confidence that your team has high standards and will support you through the inevitable failures that come from pushing against those standards.



“Get over yourself” — the four-word philosophy that builds dynasties

Popovich’s core belief reduces to four words: “Get over yourself.”

It sounds harsh. But it’s actually an invitation to freedom.

“It’s not about any one person,” he explains. “You’ve got to get over yourself and realize that it takes a group to get this thing done.”

When evaluating potential players, he watched for this quality specifically: “When there’s a guy who talks about himself all day long, you start to get the sense that he doesn’t listen real well… Has this person gotten over himself? If he has then he’s going to accept parameters. He’s going to accept the role.”

This is the hardest cultural shift for data teams to make.

We work in a field that rewards individual technical brilliance. The engineer who builds the elegant solution gets recognition. The one who writes the clever query gets quoted in architecture reviews. We celebrate the heroic debugging session that saved the quarterly report, rarely asking why the pipeline was fragile enough to need heroics in the first place.

Getting over yourself means accepting that the best data architecture might not showcase your technical sophistication. It means implementing the boring, proven solution instead of the interesting one. It means writing documentation instead of starting a new project. It means saying “I don’t know” instead of confidently bullshitting through a stakeholder meeting.

Manu Ginobili—an All-Star caliber player—started only 349 of 1,057 games because the team needed him as a sixth man. When the coaching staff asked him to permanently accept the bench role, they told him honestly: if he wasn’t good with it, he was going to start. Whatever he said, they would do it. He deserved that choice.

Ginobili agreed to come off the bench.

Tim Duncan was “blown away”: “Are you kidding? He’s Manu! He’s a star! He can’t not start.” But Ginobili embraced the sacrifice because he understood something crucial: his ego was getting between him and the team’s success.

How many data engineers would make that trade? How many would accept that their personal visibility, their resume-building features, their chance to work on the exciting project needed to take a back seat to what the team actually needed?

The uncomfortable truth: they’ll only make that sacrifice if they trust you enough to believe it’s genuine. If they feel like the sacrifice is worth it because they belong to something worth sacrificing for.



The dinner table as sacred ground

Popovich spends an estimated seven figures annually on food and wine for team gatherings. After road games, the team frequently stays overnight specifically to share dinner, with Pop personally scouting restaurants weeks in advance.

These aren’t casual meals. He positions himself at tables of exactly six people—“believing this number fosters diversity of conversation without people breaking off into side chats”—and works the room methodically. The dinners serve a strategic purpose: “Dinners help us have a better understanding of each individual person, which brings us closer to each other—and, on the court, understand each other better.”

Now here’s the challenge for modern data leaders: how do you create this in a world where your team might be spread across six time zones?

You can’t replicate the intimacy of shared meals and expensive wine. But you can replicate the intention behind them.

What Popovich’s dinners actually accomplish:

  • Dedicated time for connection that isn’t about work deliverables
  • Small enough groups to have real conversations
  • Physical proximity and eye contact (or as close as you can get)
  • Consistent repetition that builds cumulative trust
  • Leader modeling vulnerability by being genuinely present

Translation for distributed data teams:

Intentional one-on-ones that go beyond status updates. Not “what are you working on?” but “what’s actually going on with you?” Schedule them consistently. Protect them from cancellation. Come prepared with questions about their life outside work.

Small-group video calls for real conversation. Skip the all-hands. Create rotating groups of 4-6 people for unstructured time. No agenda. No recording. Just talking.

Occasional in-person gatherings that prioritize relationship-building. If you bring the team together once a year, don’t fill the schedule with strategy sessions and team-building exercises. Create space for meals, walks, and conversations that would feel weird to have over Zoom.

Consistent rituals that build cumulative connection. A Slack channel where people share weekend plans. A monthly “show and tell” of something non-work related. Anything that repeats often enough to become meaningful.

The point isn’t to replicate Popovich’s exact methods. The point is to be as intentional about relationship-building as you are about sprint planning.



When tragedy strikes, leaders appear

The most powerful evidence of Popovich’s approach emerges during personal crises.

When DeMar DeRozan’s father passed away during the 2021 season, DeRozan called the team’s GM privately—“I didn’t want nobody to know.” Ninety seconds later, Pop was knocking on his hotel room door. “Pop sat in the room with me and cried with me for about two hours. He was like, ‘I’m not leaving until you leave.’”

When Dejounte Murray’s mother was shot during his rookie year, Popovich called her directly—without Murray’s knowledge—offering to move her to San Antonio on his own dime.

When Robert Horry’s daughter was hospitalized at the start of a season, Popovich told him: “Don’t come back until she’s out of the hospital.” Three weeks later, upon returning, Pop asked if she was out. Horry said yes. Pop responded: “She ain’t out of the woods yet. Go back home. We don’t really need you right now.”

“Family is the most important thing.”

These aren’t management techniques. They’re expressions of genuine care. And they’re the reason players sacrifice for Popovich in ways they wouldn’t for other coaches.

Data leaders: when was the last time you showed up for someone beyond work boundaries?

I’m not suggesting you need to cry with your direct reports or offer to pay for their family members’ relocation. But I am suggesting that most of us dramatically underinvest in being present during the moments that actually matter.

The life events where showing up builds lasting trust:

  • Family emergencies and deaths
  • Health crises (theirs or family members')
  • Major life transitions (new babies, divorces, moves)
  • Career anxieties and disappointments
  • Personal struggles they’re brave enough to share

What “showing up” looks like in practice:

  • Proactively reaching out, not waiting for them to come to you
  • Asking what they need, not assuming you know
  • Offering flexibility without making them negotiate for it
  • Following up later to see how they’re doing
  • Remembering the details and checking in over time

The leaders I’ve worked with who inspire genuine loyalty share one trait: they remember the human stuff. They ask about your kid’s soccer tournament. They notice when you seem off. They don’t treat personal life as interference with work productivity.



High standards and deep care aren’t opposites

Here’s the false dichotomy that destroys most leadership: you’re either a demanding, results-oriented leader who pushes for excellence, or you’re a supportive, people-first leader who prioritizes relationships. Pick your lane.

Popovich refuses the choice.

Boris Diaw once visited Tony Parker at Pop’s house for Christmas dinner. After a warm meal, both Parker and Popovich disappeared. Diaw went searching and found Pop reviewing film with Tony about the game the night before—and yelling at him about missed shots and turnovers.

Diaw’s takeaway captures the essence: “On the same night, you could have the family setting, all the love and the care, and at the same time caring about making Tony a better player.”

The warmth makes the feedback land differently. Not softer—still direct, still pointed, still demanding. But the player receiving it knows it comes from someone who genuinely wants them to succeed, who has invested in them as a complete person, who will be there for the hard stuff outside work.

Research backs this up. Amy Edmondson’s psychological safety research shows that high-performing teams combine high standards with high support—“learning zones” where people push themselves precisely because they feel safe enough to fail.

For data leaders, this means:

Being specific and direct about performance issues without softening the message into meaninglessness. “The pipeline you built has reliability problems that are affecting stakeholder trust” is clearer and more respectful than “there might be some opportunities to improve things.”

Providing the context behind criticism so it lands as investment rather than attack. “I’m telling you this because you’re capable of leading our data platform work, and these patterns would prevent that.”

Following hard conversations with continued connection rather than distance. The relationship doesn’t end after difficult feedback—it continues.

Modeling receiving feedback yourself so the team sees that criticism flows in all directions. Pop regularly admitted his own failures publicly. Your team needs to see you do the same.



Connecting work to something larger

Before the 2014 NBA Finals, Popovich skipped X’s and O’s entirely. Instead, he told his team about Eddie Mabo Day—a significant date for Indigenous Australians—to honor point guard Patty Mills’ heritage.

This reflects Pop’s broader practice of engaging players on issues far beyond basketball. He’s taken teams to see Hamilton on Broadway. He brought civil rights activist John Carlos to speak. He gave players Ta-Nehisi Coates’ Between the World and Me to read.

Why does this matter for building teams?

Because it sends a message: we are complete human beings who care about the world beyond our immediate work.

Coyle noted: “Popovich would create similar conversations on the war in Syria, or a change of government in Argentina, gay marriage, institutional racism, terrorism—it doesn’t really matter, as long as it delivers the message he wants to deliver: There are bigger things than basketball to which we are all connected.”

This is the belonging cue that most data leaders miss entirely.

We act as if the only things worth discussing are data models, pipeline reliability, and stakeholder requirements. We treat outside interests as distractions rather than windows into who our team members actually are.

For data teams, connecting to something larger might include:

  • Discussions about the ethical implications of data work (privacy, algorithmic bias, surveillance)
  • Conversations about industry trends and where the field is heading
  • Space for people to share what they care about outside work
  • Recognition of cultural backgrounds and experiences
  • Honest conversation about work-life balance, burnout, and sustainability

None of this requires seven-figure wine budgets. It requires treating your team members as complete humans who exist in a world larger than your Jira board.



The evidence speaks across decades

Popovich’s methods produce measurable results. Five championships. Twenty-two consecutive playoff appearances. Eighteen straight 50-win seasons—unprecedented in NBA history. The best winning percentage of any major professional sports franchise in North America over a three-decade span.

The 2014 championship team became the first since the ABA-NBA merger to win a title without any player averaging 30+ minutes per game. Four different players led the team in scoring during the playoffs. No dominant statistics—just beautiful, unselfish basketball.

Beyond the Spurs, his coaching tree now shapes the entire league. When The Athletic surveyed NBA players on which coach they’d most want to play for, Popovich received 41%—the highest percentage by far.

Kevin Durant captured the appeal: “When you are with somebody like that, a person who cares about all those things on top of wanting you to be the best player you can be, then you want to go to work with that person.”

The same principle applies to data teams. The leaders who build lasting excellence aren’t the ones with the cleverest technical strategies. They’re the ones who create environments where talented people choose to stay, sacrifice, and push each other toward genuine accomplishment.



The uncomfortable question for data leaders

At his Hall of Fame induction, Popovich offered unusual clarity: “Basketball doesn’t love us back, does it? We use it like a bar of soap, right? It pays our bills. It gives us a wonderful life. But I don’t remember it saying, ‘I love you, Pop.’ It’s different. It’s the family.”

Here’s the uncomfortable question for data leaders: what’s the family equivalent for your team?

Data pipelines don’t love you back. Dashboards don’t remember your birthday. The perfectly optimized query won’t show up at your door when your father dies.

The work matters. Technical excellence matters. Delivering value to the organization matters.

But none of it persists if you haven’t built the human foundation that makes sustained performance possible.

The false choice is between being a great technical leader and being a great people leader. Popovich proves you need both—that technical excellence requires the human foundation, not despite it.

So the next question isn’t whether you have the right technology stack or the optimal team structure. It’s whether you’re investing in the belonging cues, the dinners (literal or figurative), the showing up during crisis, the relentless standards delivered with genuine love.

It’s whether you’re willing to do the uncomfortable work of building something worth sacrificing for.

Because here’s the thing: your team knows the difference. They know when connection is performative and when it’s real. They know when high standards come from caring and when they come from ego. They know whether they’re resources to be optimized or humans to be invested in.

And they’ll calibrate their own investment accordingly.

Tim Duncan, accepting his Hall of Fame honor, said what countless Spurs have felt about Popovich: “You showed up after I got drafted. You came to my island. You sat with my friends, my family. You talked with my dad. I thought that was normal. It’s not.”

I’m nowhere near Pop’s level. Most of us aren’t. But stories like this matter because they remind us what good can look like—and give us something to measure ourselves against.

Not to feel inadequate. To feel inspired.

The gap between where we are and where Pop operates isn’t a source of shame. It’s a direction. And the first step is simply asking the question: Would my team say I showed up?