The Teammate Everyone Quietly Works Around

9:40 on a Tuesday. Maya is half watching a channel while her tea goes cold. Dan needs a data pull for the two o'clock client review. He asks in the channel, waits four minutes, then messages Sofia directly.

Sofia rebuilds the query from scratch. It eats her morning, and she gets one date filter wrong. The client catches it before anyone on the team does.

Rehan wrote that pipeline. He was online the whole time, green dot, answering in two other threads. His name sits at the top of the schema doc Dan had open in another tab.

Maya lets it go. On its own it looks like nothing.

Over the next six weeks she watches it happen four more times. A vendor question that belongs to Rehan goes to a contractor. A bug in Rehan's own service gets triaged by someone who has to read the code first. By the fifth one, Maya stops calling it coincidence and starts calling it a route.

What was actually going on in that channel

People pick who to ask based on how the asking will feel, and they do it fast, under pressure, without deciding anything consciously.

Anita Tucker and Amy Edmondson spent 239 hours observing 26 nurses across nine hospitals, watching what happened when something blocked the work. One pattern showed up over and over: when nurses needed help, they went to whoever was socially close, not to whoever could actually correct the problem. They followed that rule for 42% of the problems observed. One night-shift oncology nurse ran out of clean linen, walked to a different unit, and took theirs. She described it plainly to the researchers: "Working around problems is just part of my job."

Tiziana Casciaro and Miguel Sousa Lobo found the same instinct in a different setting. Studying 10,000 work relationships across five organizations, they found people choose work partners on likability more than on capability, even when they insist otherwise.

So what was Rehan doing? Nothing dramatic. He answers a request with a question. He wants to know what the number is for before he hands it over. He takes a day. And two years ago, in a room with six people in it, he told a junior analyst her join logic was obviously wrong.

None of that is a fireable offense. All of it is enough to make a tired person at 9:40 on a Tuesday pick the easier door.

The detour has a price and nobody sees the bill

Tucker and Edmondson put a number on the workaround. Nurses lost an average of 33 minutes per eight-hour shift coping with failures that could have been removed. For a 200-bed hospital, the researchers estimated the wasted time at $256,000 a year in nursing salary alone.

Your version has no line item. Sofia's morning went somewhere, and it didn't go on anyone's dashboard.

Then there's the quality tax. Sofia got the filter wrong because she was reverse-engineering something Rehan already knew. The detour pays for the person's comfort with the client's confidence.

And Rehan gets quietly de-skilled. He touches less of the interesting work, so his reviews get vaguer, so his next project gets smaller. Six months from now someone will say he's plateaued, and they'll be right, and nobody will connect it back to a Tuesday morning in the channel.

The last cost is the one that stings. Rehan's annoying question, the one about what the number is for, is often the question that catches the error. Route around him and you route around the check.

Try this on Monday

Start by counting. For one week, keep two columns on a sticky note: who did the thing, and who was the obvious owner of the thing? Five requests is enough. If the columns match, you're fine, and you've spent ten minutes proving it.

Then talk to the requester, not the person being skipped. Dan is the one holding the information you need. Ask him something narrow: "You went to Sofia for that pull. What made that the easier call?" Do this with three different people separately, and you'll get the actual answer inside a week. Ask it in a group and you'll get a shrug.

Now tell Rehan. This one's uncomfortable, and vagueness is what makes it worse. Give him the incident, the cost, and one thing to change. Something like, "Dan needed a pull last Tuesday. He went to Sofia, it took her four hours, and the date filter came out wrong. People are reading your clarifying questions as pushback. For the next month, answer first and ask second."

Last, close the door for a fortnight. Put the named owner on the ticket before the work goes out, so the detour takes effort. Two weeks of that will tell you whether the pattern was a habit or something deeper.

Where this goes wrong

Here are a few ways this can go wrong and what you can do about it.

• Doing any of this in standup. Public and pattern-level is how you turn a fixable habit into a reputation.
• Softening it into "some people have mentioned" and leaving. Rehan can't fix a rumor. He can fix Tuesday.
• Assuming character when the answer is capacity. Check his load first. If Rehan is on call, sitting on forty open tickets, and answering at 11pm, then the team routing around him is a sensible response to a real bottleneck, and the conversation you need is about staffing.
• Telling Sofia to stop picking things up. She was being helpful under time pressure. Punishing that gets you a slower team and the same detour.

One thing this week

Look at the last five things your team delivered and ask who actually did each one. Then ask who should have.

Three weeks after Maya's conversation, Dan needed another pull. He went to Rehan. Rehan asked one clarifying question, which was what the comparison period should be, and sent the numbers back in forty minutes with the filters right.

Dan didn't notice anything had changed. That's what a working route looks like.