The most useful thing I do as a manager is decide where pressure lands.
Every business under strain generates pressure. A client is unhappy, revenue is behind, someone senior needs a number by Friday. The easy move is to pass it straight down: forward the email, add it to the sprint, tell the team it's urgent. It feels like leadership, because things start moving.
What actually happens is that the team starts making decisions scared. They cut corners they'd normally flag. They stop telling you when something is wrong, because they can see you haven't got room for more bad news. The work gets faster for a few weeks and worse for months.
Getting there first
The best way I know to keep pressure off a team is to get to it before it arrives. Most things that land on a data team as emergencies were visible weeks earlier to anyone paying attention: a contract renewal coming up, a client who is going to want their data cut differently, a new product that will need a source we don't collect yet.
So a big part of my job is sitting close enough to the commercial side to see those things coming and having the platform ready before anyone asks. If the business wants to sell something new and the data is already there, there's no crisis. The company has no excuse to fall short, and my team never feels it.
Absorbing it isn't the same as hiding it
I learned the limit of this the hard way. At one company I found out my team had a reputation about six months after we'd earned it, from a passing comment in a meeting: "The data team are always behind."
The priorities had kept changing, usually for good reasons, and every time I'd absorbed the change. The team dropped what it was doing and started the new thing. Nobody saw the cost, because the work just moved. The bill arrived weeks later, when everything was half-finished and nothing was ready to ship. Then the priorities changed again. After enough of that, the team was holding a pile of unfinished work and the organisation was holding a reputation.
I'd been calling it adaptability. Really it was invisibility. The people asking for the changes never saw what they cost, because I never showed them.
Now, when someone asks for a change, I put the cost on the table as part of the answer: here's what stops, here's what gets thrown away, and here's when the new thing will actually land. I say it plainly, without complaining or digging my heels in. Most of the time they still want the change, and that's fine. It's now their decision with its cost attached, instead of my team's cost with nobody's name on it.
Stopping pressure at the lead means taking it, not hiding it.
When it's genuinely bad
Sometimes you can't absorb it. This year we had a security incident and a revenue problem at the same time, and I had to take my team from six people to two.
I couldn't keep that away from the people it affected. What I could do was make it land as gently as I was able to: time to find the right next role for the people who were leaving, a fixed-term contract for someone who needed a bridge, and the difficult conversations with the business taken on by me rather than passed through the team.
The two of us who were left finished the platform rebuild. I don't think that happens if we spend those months scared.