If you run EMS in the Hudson Valley, you know the Gunks. The Shawangunk Ridge draws climbers from the city every weekend, and the terrain above the carriage roads is the kind that's accessible, but not easily accessible — you can get a truck to the trailhead. From there, it's a carry.
The dispatch came over in the flat tones it always does: climber fallen thirty feet, unknown injuries. That could mean a bad day — swung out on a rope, banged against the rock, shaken but intact. Or it could mean the worst day of someone's life — a ground fall, and a ground fall needs things you can't get quickly: a medevac helicopter, an engine to set up a landing zone, a second ALS unit, hands to carry a patient out.
So before I had eyes on them — before I knew which kind of fall this was — I got on the radio and ordered all of it. My partner didn't blink. "You can always cancel a helicopter," she said. "You can't un-waste the twenty minutes you spent wishing you had one."
We hiked in. Rope fall — the good version. Wrist fracture, alert, embarrassed. I got back on the radio and canceled everything I'd requested. Nobody questioned it. That's how the system is built: stage the worst plausible scenario, scale back once the picture clarifies. The cost of ordering a helicopter you don't need is close to zero. The cost of needing one you didn't order is measured in time you can't get back.
In my day job, I've been running a high-visibility project for six weeks now, and I did the opposite of that call.
At the start, I did what needed doing for a quick sprint myself — set up my own environment, figured out how to socialize the work, built the tracking. That part was right. But when it came to bringing in more hands, I didn't. I was offered temp help and turned it down.
I wasn't afraid or too proud to ask. I thought I was making the right call — protect my team from being randomized onto something new, spare everyone the ramp-up time a temp resource from another team would cost before they were even useful. Both of those are real costs, and I still believe it. What I didn't weigh was the cost on the other side of the ledger: the IC work that quick sprint needed from me the whole time. That cost doesn't show up in a review. It shows up in what your team doesn't get from you while you're heads-down being useful in the wrong role.
Because I didn't order early or big, I carried it. For six weeks, I was the infrastructure, the designer, the tracking, and the leadership — all in one person. And here's the cost, and it's not the one I expected: the weeks I spent heads-down as an individual contributor were weeks I wasn't available as a leader. My team didn't get less work done. They got less of me — the coaching, the unblocking, the presence that's actually my job. I was so busy being useful that I stopped being the things only I could be.
Nobody handed me a bad outcome for it. The project shipped. That's exactly what makes this worth naming: you can decline the resource, absorb the cost yourself, and still hit the deadline — and still have made the wrong call, because the bill just gets paid somewhere you're not looking.
I had a helicopter on standby. I sent it home because I was solving for the cost I could see — ramp-up, disruption — and missed the one I couldn't: what my team loses when I'm not leading them. Next time, I'm calling for the helicopter early. Not because I don't trust myself to carry it — I just did. Because carrying it alone isn't actually my job.